Quad ESL 57s
Doug Flynn
I used to own a pair of ESL 57s a few years ago, but I sold them because I
didn't have any room. I always regretted doing so, but as fate would have it
a pair has now landed my way (actually, they were the pair I used to own).
Any suggestions for an amplifier which would do them justice? I am open to
any suggestions, as long as the amp glows in the dark :-)
I'd love to build a 300B SET amp for them, but I'm not sure 8 watts would be
enough. Does anyone here use Quads with low power tube amps?
Cheers. Doug
Patrick Turner
The perfect thing would be a pair of amps with 3 x 300B in parallel for
24W max into 8 ohms, with Rout = 1 ohm, like a pair of Quad-II, but with
less NFB and **ALL** class A.
It isn't huge power you want though, its fidelity with the voltage
output. you need something to give up to around 25Vrms max at high Z
loads because the Z of ESL57 varies from over 30 ohms at LF down to
8ohms at 1 kHz to 1.8 ohms at 18kHz.
So where most music energy resides, you need clean voltage at low
current, and for the highs you need low voltages at low current because
the energy in most music declines greatly above 1kHz. Hence the Quad-II
amps were not too bad - when they are ever working properly.
1 x 211 would be nice, but forget 845 because they seen to have become
outrageously expensive if you get the KR Audio types.
The Chinese seem to have stopped making 845.
Frankly, 3 x KT88 in parallel SE triode would also be nice IMHO, but
then you have prejudices about what sounds good and so don't take any
notice of what i might say to the group.
KT88 in a pair of Quad-II will work just fine BTW, and just follow the
leads at my website if you want to know more on modding Quad-II.
Patrick Turner.
33-45
Hi Doug..good to see those babies are back home...where they belong.
I'm sure you'll get a few answers on the best amplifiers to use..In fact
there's a good site with a list of Tube and a couple of SS amps
recommended for the 57's..just can't remember atm.
I've had my 57's for over ten years now..never had to repair or do
anything to them...so the high maintenance myth is just that..
Keep them away from the head banging teenagers and feed them good
amplification and you'll have many years of hassle free audio nirvana...
Same goes for my amp...an old Contan kit job with upgraded wiring, pot's
etc. Originally had 5881 outputs which are a nice tube but I prefer the
fat bottle Sylvania 6L6GC's..as they give much tighter base imo. Also
replaced the EF86's with E80F's (one of the best kept secrets in tube
substitution..try them sometime)
Any good quality tube amp of around 25-35 watts output would be a good
place to start imo...
Good luck with them...Get it right and you'll be hard pressed to find
anything better......let us know how you go
ps..Ain't in nice to see some audio discussion for a change....touch wood
Patrick Turner
33-45 wrote:
>
> Doug Flynn wrote:
> > Hi chaps
> >
> > I used to own a pair of ESL 57s a few years ago, but I sold them because I
> > didn't have any room. I always regretted doing so, but as fate would have it
> > a pair has now landed my way (actually, they were the pair I used to own).
> >
> > Any suggestions for an amplifier which would do them justice? I am open to
> > any suggestions, as long as the amp glows in the dark :-)
> >
> > I'd love to build a 300B SET amp for them, but I'm not sure 8 watts would be
> > enough. Does anyone here use Quads with low power tube amps?
> >
> > Cheers. Doug
> >
> >
>
> Hi Doug..good to see those babies are back home...where they belong.
> I'm sure you'll get a few answers on the best amplifiers to use..In fact
> there's a good site with a list of Tube and a couple of SS amps
> recommended for the 57's..just can't remember atm.
>
> I've had my 57's for over ten years now..never had to repair or do
> anything to them...so the high maintenance myth is just that..
> Keep them away from the head banging teenagers and feed them good
> amplification and you'll have many years of hassle free audio nirvana...
>
> Same goes for my amp...an old Contan kit job with upgraded wiring, pot's
> etc. Originally had 5881 outputs which are a nice tube but I prefer the
> fat bottle Sylvania 6L6GC's..as they give much tighter base imo. Also
> replaced the EF86's with E80F's (one of the best kept secrets in tube
> substitution..try them sometime)
Many Contan amps were PP and based around the Hedge circuit with
cascoded LTP input triodes in unusual formation,and with lots of GNFB.
From the little I saw of them they seemed to have good OPTs and PTs.
All would run best with KT88. Input should be SE 12AU7 or 6CG7 followed
by LTP with 6CG7 or such like and this will be much better than the
Hedge thing, or anything with pentodes such as Dynaco ST70/80, Leak,
Quad-II, Radford, and a host of others.
For the best of PP, see
http://www.turneraudio.com.au/Integrated5050.htm
You are right about the EF80 as replacements for the EF86, except that
one should try to use a lower RL for the EF80 and get more Ia flowing in
the tube than the usual 0.5mA or less in many circuits with EF86. This
means the EF80 ( 6BX6 ) will need to have different dc carrying RL and
screen feed resistors and biasing resistors. The gm of a pentode varies
greatly with Ia and determines the voltage gain and if Ia for EF80 is as
low as it is in EF86 then no great increase in gain occurs. But with say
5mAdc instead of 0.5mAdc, gm is vastly increased over the EF86 and even
with a lower dc carrying RL the gain is higher and may then double. If
there is GNFB, the amount of GNFB will double because the open loop gain
doubles. This means that although you may have reduced Rout by half, and
halved the THD/IMD, the amp may become quite fussy about loads and
perhaps oscillate badly at HF if the load is capacitive, and at LF if
the amp has a high value load at bass F or has no load connected.
Substituting output tubes usually causes less problems to be solved.
6LGC, KT66, 5881 will all have about the same gain in a UL output stage.
KT88/90 can usually be plugged straight in as replacements for 6l6 etc,
because they have been made to bias up the same. The KT88/90 only give a
marginal increase in voltage gain in a UL or triode circuit but the Rout
will be considerably lower because the Ra even in UL mode with '88 is
lower than that for 6L6 etc.
Some say that KT88 in triode are best in Contan amps ( or any others
which originally had 6L6/KT66/5881 ) thus limiting power a bit but 20W
is plenty. If the drive amp is made linear enough the GNFB can be
reduced to maybe 10dB just to get the Rout low enough, but with ESL57,
you could get away with zero GNFB because bass Z of the ESL is high and
the triodes without any GNFB will give Rout of around 2 ohms. Lowest
Rout in triode class A is achieved when the load is always plugged into
the 4 ohms outlet regardless of what the load is. The use of KT90 gives
the lowest triode Ra and hence the lowest Rout at the output.
EL34 can also be used but you must change the biasing to get the same
output tube idle current. And they have more voltage gain for a given
load so stability should be checked.
Patrick Turner.
atec 7 7
33-45
> You are right about the EF80 as replacements for the EF86, except that
> one should try to use a lower RL for the EF80 and get more Ia flowing in
> the tube than the usual 0.5mA or less in many circuits with EF86. This
> means the EF80 ( 6BX6 ) will need to have different dc carrying RL and
> screen feed resistors and biasing resistors.
Patrick, I know it could be easy to do but don't confuse E80F with EF80
totally different tube
Patrick Turner
I got caught again.
I thought you meant EF80, or 6BX6, but the E80F is a special quality
pentode...
http://www.mif.pg.gda.pl/homepages/frank/sheets/030/e/E80F.pdf
and it is similar to EF86 which was supposed to be special anyway
because it had a helical filament to lower hum.
But the EF80/6BX6 could be used anyway to replace EF86 and with some
mods as I described and you might find some sonic gains.
There was E180F and rare E280F also, and these pentodes had frame grids
and had quite high gm, and mainly used for RF/IF amps.
The pentodes have much higher THD than when used as triodes, or compared
to a real triode.
But because the THD is mainly 2H, when they are used in an LTP the THD
is reduced, but then becomes mainly 3H.
At up to 10V output the THD from an LTP ( differential pair ) of
pentodes is usually much lower than the output stage being driven by
them so they become quite acceptable. Their high Ra means the HF can
easily be shunted by stray C on Miller C and you cannot use a choke load
because the L will shunt the high Ra at LF, and the iron THD is
terrible. The choke load will give a response that is like a big arch
with the highest gain at the top of the arch where L and C have least
effect. In RC coupled amps like Quad-II the EF86 work OK, but its a
paraphase phase splitter they have and the input pair work a lot better
if set up as a true differential amp. In Quad-II, you can't just wire
the EF86 as triodes because the triode gain is far lower than the
pentode gain. So you'd need another input triode to use the same NFB and
get the the same 1.4V input sensitivity for full PO. People originally
addopted pentodes so readily for input stages of power amps because
their high gain reduced the tube count, as Quad-II illustrates. Not even
12AX7 could give the wanted gain needed for Quad-II.
Other makers like Leak used the EF36 ahead of a twin triode for their
input stages, Mullard had EF86 input plus 12AX7 LTP driver stage.
These old designs allowed 20dB of GNFB and they still had input
sensitivity down to 0.1Vrms which meant the old sources from AM/FM
radio, tape, phono needed less gain and could be done with fewer tubes.
The price was a poorer SNR. Williamson's amp of 1947 is still a very
good class A design and will sound well today if you have low noise 6CG7
inputs/drivers, (all triodes like 6SN7 ) and KT88 outputs to give 20W
AB1 in triode instead of the 12W with KT66. The input voltage needed
with the 20dB GNFB is 2Vrms for clipping.
In fact, 20dB GNFB is a bit too much and 12 dB is OK which means input
could be 0.8Vrms and you still get a good SNR.
I have no idea how Contan used the EF86; I only ever saw triodes being
used in his input stages.
Patrick Turner.
roughplanet
news:49D34122...@turneraudio.com.au...
33-45 wrote:
Doug Flynn wrote:
As you know Patrick, my amp runs 12AX7's as drivers & EL84's as outputs.
Recently I purchased a new Laney LC-15R 15 watt practice amp, and, you
guessed it, there was the same configuration; 12AX7's (ECC83's) & EL84's.
Was I pleased? I'm still smiling as I have literally dozens of both types &
so
can play around with them until I get just the sound I like.
>> Any good quality tube amp of around 25-35 watts output would be a good
>> place to start imo...
>>
>> Good luck with them...Get it right and you'll be hard pressed to find
>> anything better......let us know how you go
>>
>> ps..Ain't in nice to see some audio discussion for a change....touch wood
ruff
roughplanet
news:gqvhh2$53k$2...@news.motzarella.org...
Doug Flynn wrote:
> wooffffwoffffdownphilthy
Is that him humping Doug's leg?
ruff
Doug Flynn
> The perfect thing would be a pair of amps with 3 x 300B in parallel for
> 24W max into 8 ohms, with Rout = 1 ohm, like a pair of Quad-II, but with
> less NFB and **ALL** class A.
>
> 1 x 211 would be nice, but forget 845 because they seen to have become
> outrageously expensive if you get the KR Audio types.
> The Chinese seem to have stopped making 845.
>
> Frankly, 3 x KT88 in parallel SE triode would also be nice IMHO, but
> then you have prejudices about what sounds good and so don't take any
> notice of what i might say to the group.
Thanks for the reply. I love 300Bs but 3 in parallel would need a serious SE
OPT. I have some fancy 3K SE OPTs which are gapped for 100mA.
I'm not a fan of big transmitting tubes like 211 or 845. I've built serveral
845 amps, and I always found the sound to be coarse (or maybe that was just
the Chinese 845s I used), although they did have tremendous clout and bass
welly.
I'm not prejudiced against KT88s (or KT66s, of which I have many), as long
as they are triode-strapped. I would rather suffer from scrotal
elephantiasis than listen to a pentode or tetrode. I would never build
anything with 6550s - they are not audio tubes. Sorry, I know you love them,
but they suck badly.
Doug
roughplanet
news:%c_Al.333$y61...@news-server.bigpond.net.au...
"Patrick Turner" <in...@turneraudio.com.au> wrote >
>> The perfect thing would be a pair of amps with 3 x 300B in parallel for
>> 24W max into 8 ohms, with Rout = 1 ohm, like a pair of Quad-II, but with
>> less NFB and **ALL** class A.
>>
>> 1 x 211 would be nice, but forget 845 because they seen to have become
>> outrageously expensive if you get the KR Audio types.
>> The Chinese seem to have stopped making 845.
No loss. I have yet to hear an amp running 845's that didn't sound as rough
as bags. Not a pleasant audio tube IMO.
>> Frankly, 3 x KT88 in parallel SE triode would also be nice IMHO, but
>> then you have prejudices about what sounds good and so don't take any
>> notice of what i might say to the group.
> Thanks for the reply. I love 300Bs but 3 in parallel would need a serious
> SE OPT. I have some fancy 3K SE OPTs which are gapped for 100mA.
>
> I'm not a fan of big transmitting tubes like 211 or 845. I've built
> serveral 845 amps, and I always found the sound to be coarse (or maybe
> that was just the Chinese 845s I used), although they did have tremendous
> clout and bass welly.
>
> I'm not prejudiced against KT88s (or KT66s, of which I have many), as long
> as they are triode-strapped. I would rather suffer from scrotal
> elephantiasis than listen to a pentode or tetrode. I would never build
> anything with 6550s - they are not audio tubes. Sorry, I know you love
> them, but they suck badly.
Dunno about that Doug. I have heard a few amps that had their existing tubes
replaced with 6550's & the results were quite audibly pleasing, even though,
as
you say, 6550's are not audio tubes.
ruff
Phil Allison
"Drug Flynn" <dog...@bigpond.com
>
> I would never build anything with 6550s - they are not audio tubes.
** Total BOLLOCKS !!!!
Here is a link to a November 1954 advertisement from Tung-Sol - who
developed the 6550 type that same year.
http://vintagetubeservices.com/page16.html
From the actual ad: " ... designed specifically for audio service".
Also, in the intro: " Tung-Sol's greatest contribution to the world of audio
was the 6550, conceived and developed for Hi-Fi and introduced in 1955 this
tube is still powering many of the world's greatest sound systems 45 years
later! "
This link is to the opening page in the GE data book for the type:
http://www.triodeel.com/6550ap1.gif
Describes the type as:
" Beam Pentode for A-F Power Amplifier Applications "
Could not be any clearer.
.... Phil
bassett
"Phil Allison" <phila...@tpg.com.au> wrote in message
news:73l004F...@mid.individual.net...
designerd a backhoe to remove you as well.
Doug Flynn
"roughplanet" <rough...@optusnet.com.au> wrote
>> I'm not prejudiced against KT88s (or KT66s, of which I have many), as
>> long as they are triode-strapped. I would rather suffer from scrotal
>> elephantiasis than listen to a pentode or tetrode. I would never build
>> anything with 6550s - they are not audio tubes. Sorry, I know you love
>> them, but they suck badly.
>
> Dunno about that Doug. I have heard a few amps that had their existing
> tubes
> replaced with 6550's & the results were quite audibly pleasing, even
> though, as you say, 6550's are not audio tubes.
If you replace those 6550s with KT88s I guarantee you will lose control of
your bodily functions. When I tread in doggy doo-doo I usually say "oops,
I've just trodden on a 6550"......Doug :-)
Doug Flynn
>
>>
>> I would never build anything with 6550s - they are not audio tubes.
>
>
> ** Total BOLLOCKS !!!!
>
>
> Here is a link to a November 1954 advertisement from Tung-Sol - who
> developed the 6550 type that same year.
>
> http://vintagetubeservices.com/page16.html
>
> From the actual ad: " ... designed specifically for audio service".
Dude, that's just the marketers doing what they do best, ie. speaking
through their arses. OK, so you *can* use them for audio (if you want that
"fingers being dragged down a blackboard" sound), but are you seriously
tellling me they are as linear as a KT66 or KT88???
Love and kisses. Doug
Phil Allison
>
>>> I would never build anything with 6550s - they are not audio tubes.
>>
>>
>> ** Total BOLLOCKS !!!!
>>
>>
>> Here is a link to a November 1954 advertisement from Tung-Sol - who
>> developed the 6550 type that same year.
>>
>> http://vintagetubeservices.com/page16.html
>>
>> From the actual ad: " ... designed specifically for audio service".
>
> Dude, that's just the marketers doing what they do best,
** So you fucking lied.
6550s are specifically hi-fi audio tubes in every way.
..... Phil
Patrick Turner
Doug Flynn wrote:
>
> "Patrick Turner" <in...@turneraudio.com.au> wrote >
>
> > The perfect thing would be a pair of amps with 3 x 300B in parallel for
> > 24W max into 8 ohms, with Rout = 1 ohm, like a pair of Quad-II, but with
> > less NFB and **ALL** class A.
> >
> > 1 x 211 would be nice, but forget 845 because they seen to have become
> > outrageously expensive if you get the KR Audio types.
> > The Chinese seem to have stopped making 845.
> >
> > Frankly, 3 x KT88 in parallel SE triode would also be nice IMHO, but
> > then you have prejudices about what sounds good and so don't take any
> > notice of what i might say to the group.
>
> Thanks for the reply. I love 300Bs but 3 in parallel would need a serious SE
> OPT. I have some fancy 3K SE OPTs which are gapped for 100mA.
That means you can have 70mArms of load current giving nearly 15 watts
into your 3k with Ea at about +350V, Ia 100mAdc.
Good for two SE 300B.
PP 300B can be easily made to make a nice 25W in AB1.
>
> I'm not a fan of big transmitting tubes like 211 or 845. I've built serveral
> 845 amps, and I always found the sound to be coarse (or maybe that was just
> the Chinese 845s I used), although they did have tremendous clout and bass
> welly.
You were a huge fan of such tubes. I always thought maybe you never knew
quite enough to get the best sound. Maybe if you tried again you'd have
better success. Sometimes we have to return to things we failed with and
spend a bigger effort to get the revelation, and the cancelation of our
predjudices. I didn't like making SS amps at first, but after several
tries I found I could get astonishing results.
I found an SE pair of parallel Chinese 845 and KR Audio 845 to sound
almost identical, and to be marginally clearer sounding than a very well
done PP amp using a quad of Sofia electric carbon plate 300B or Sovtek
300B in VAC I had just completly re-engineered.
How you drive the large transmitting tubes is just as important as the
output tubes and exceptional attention needs to be given to power
supplies. Then its easy to get nearly all 50W SE amps to sound better
than most 70W PP amps no matter what they are.
>
> I'm not prejudiced against KT88s (or KT66s, of which I have many), as long
> as they are triode-strapped. I would rather suffer from scrotal
> elephantiasis than listen to a pentode or tetrode.
Maybe you have paralysis of the brain in that area required for you to
expand your consciousness without predjudice.
When I make an amp with triodes or tetrodes it is impossible to easily
tell whether thay are in fact triode or pentode or tetrode connected,
or even if I have used mosfets. And when the triode without NFB versus
UL with NFB issue was tested once amoung a gathered crowd of 30
audiophiliolics in Sydney in 1998, my UL amp with a pair of 6550 and
15dB of GNFB easily beat the guy with his fancy triode thing with twice
the number of 6550 and 6dB of GNFB. I subsequently wowed the crowd with
SE amps based around the 13EI, a big fat *general purpose* beam tetrode
capable of a very refined 32 watts in SE class A.
I think its down to who builds the amps and their attention to detail
and basic issues of linearity, bandwidth, stability, output resistance,
low THD/IMD and low noise etc. Once you have got all those right with a
mimimum of NFB then you worry about the brands of caps and type of
solder and cabling. And the AB tests usually tell people there isn't
much difference in cables, caps or solder. But I know ppl who tell me AB
tests suck, and of course they do, because they are so likely to make a
fool of anyone being involved in a test.
I would never build
> anything with 6550s - they are not audio tubes. Sorry, I know you love them,
> but they suck badly.
I know that you think 6550s suck.
They ARE considered just as much an audio tube as any other is by the
vast majority of those into tubed audio.
The poor press given against beam tetrodes and pentodes stems from
absurd arguments based on short sighted assumptions made by technically
incompetent reviewers with cloth ears in idiototic magazines which
merely fill potential buyers minds full of bullshit. This negative press
is all the more enhanced by the war of watts amoung major players like
ARC and McIntosh who have made their amps ever more salable with higher
and higher amounts of maximum power ratings at the sonic cost of low
amounts of class A power. They set up the tubes in ever more complex
ways but always ending in a virtually class B output stage with all too
little class A and with a load that is way too low a value. The horses
are unhappily flogged to death to make a big impression. The onset of
transistor power after 1960 made matters all the worse. Makers now tend
towards low levels of NFB in PP amps mainly because its so hard to
stabilise them with any more NFB. That's because OPT are poor copies of
bad examples of the 1950s with low bandwidth and high winding losses.
The public has become less critical and don't mind bad sound, bearing in
mind so many use tubes for movie sound which has been processed to be
quite appalling no matter how good your digital gear and amps are.
So the distortion tends to hit you in the ear a bit and the well done SE
amp even if configured in beam tet/pentode mode and with FB will run
rings around many PP amps made now, until the SE amp has to make serious
power, and then the 8W SET is a piece of junk.
With old sensitive speakers the 8W from an SET amp can be quite
captivatingly marvellous. How could something so small do so much!
People, or idiots, without real intellectual style sneer at the 6L6
because its used in guitar amps and what are mainly class B PA amps with
little NFB. These same inconsistent muddle heads often *lerve* the sound
of modern pop music which has so largely depended on the deliberately
created distortions in 1,001 musicians amps using 6L6.
When used with a substantial amount of class A working the 6L6 is just
as marvellous as any other tube depending on who builds an amp and how
they set up the whole thing. Ditto the 6550, KT88, KT90 and all the
others.
The sound you get from tubed audio gear is dependant on the sum of all
the parts and methodology and topology used. Triodes have their own
internal NFB built in and can be used without loop FB, but much care has
to be used for PS and load matching to reduce N&D because you don't use
the loop FB.
I've heard very good music from ESL speakers propelled by 20W SET amps
using 6C33c. The ESL were home brew and better than ESL57.
And they were more sensitive, so the 20W was plenty.
If you really want to reproduce the real sound made by a big band in
front of you then ESL57 may dissapoint. They were designed for mainly
quiet middle class after hours listening at levels which wouldn't wake
the kids. If you stick with low levels then 8W is enough but I like to
turn things up a bit sometimes so a pair of KT88/6550 in PP CFB (like
Quad 40) is not bad form of amp. I quite like trioded multigrids but
there **is nobody** I have ever known who is able to tell any difference
in the sound at all when the trioded KT88/6550/KT66 etc is switched from
UL to triode or back again providing levels are well clear of clipping
which is usually the case.
I've tried the AB comparisons and asked people to pick the differences
and to identify if its triode or UL. Nobody has.
I once was at a comparison between a large SS 100W class A Gryphon and
23W SET amps. The difference was chalk and cheese in favour of the SET
amps. 14 in the room were very happy with the musicality of the SE amp,
but one guy wasn't, for it turned out he owned a Gryphon, and now all
these people around him thought differently about music.
On another occassion at the same shop venue I heard Martin Logans with
an SS amp and the massed strings and singing in the Opera Carmen sounded
quite dreamy. None of the usual SS smear and smirchy transients. Maybe
they had a decent CD player that day. From the limited amount I know, cd
players have generally not got better over the last 10 years, just worse
and cheaper.
Martin Ls are not easy to drive and like Quad ESL63 and later models are
insensitive so 80W capabability is not a bad idea. One customer of mine
in Melbourne runs 3 stacked pairs of ESL57 which have been fully
restored by John Hall, the melbourne guy who specialises in their
restorations. My customer uses 4 x KT90 per channel which can make 100W
into 4 ohms and piles of class A power to cover 99% of what the level
is. Now the level isn't any higher than using one pair of ESL except you
get more bass with the stacking, so the amp doesn't have to make any
more actual power than used when a 33W amp drives one ESL57. But the
operation of more tubes at low power mean the IMD and THD remain very
very low because its all class A power and very honest power. The load
of the 3 paralleled ESL57 does make a challenge to any amp but only if
you push it to deafening levels.
So thus 20W+20W could be used to drive 3 stacked pairs of ESL57, but the
load match becomes more critical. Quad-II amps would need an official 4
ohm outlet which they officially don't have. Unofficially, there *is* a
4 ohm outlet on Quad-II amps but nobody uses it and nobody promoted it
including Quad, mainly because the 4 ohm outlet match increases winding
losses above 25%. Peter Walker made a real mess of his OPT IMHO; much
better performance could have been had with a slightly more complex way
of arrangeing load matching but I think Walker deemed it too complex for
ordinary dumb middle class users to change. They'd always get the change
wrong if there was more than two connections to change.
An accountant using a soldering iron usually leads to a terribly high
repair bill. However, with good design with tubes you have the ability
to have variable secondary winding arrangements on the OPT so the load
matching can be optimised in well made amps and avoid the variation with
winding losses when load matches are changed.
If you have a 35W PP amp then usually that means 35W is possible when 4
ohms is connected to the 4 ohm outlet, or 8 ohms to the 8 ohm outlet.
But usually the 35W of power is AB1 with a small amount of class A. So
its usually always best for fidelity to connect 8 ohm speakers to most
modern 4 ohm outlets. This doubles the load ohms seen by the tubes which
then work with far higher class A % at half the THD/IMD for the same
power. So if its possible, ESL57 should be used with the 4 ohm outlet on
35W rated amps. The maximum possible output voltage with the 4 ohm tap
on a 35W amp will always be less than the maximum allowable into ESL57
without causing arcing in the panels which lead to very expensive repair
costs.
By connecting say 8ohm speakers to 4 ohm outlets the damping factor is
doubled giving tighter bass and usually better mid/treble. But output
powe max reduces from say 35W to 25W. This makes a dissapointing
subconscious feeling of inadequacy and loss in many audiohiles' minds.
If only they would connect a peak and hold voltemeter to the speaker
terminals to capture the maximum peak voltage during an evening's
listening. Then they'd find their amps barely went more than 10 watts
peak if they knew how to work out peak instaneous power from the voltage
readings.
Patrick Turner.
>
> Doug
Patrick Turner
roughplanet wrote:
>
> "Doug Flynn" <doog...@bigpond.com> wrote in message
> news:%c_Al.333$y61...@news-server.bigpond.net.au...
>
> "Patrick Turner" <in...@turneraudio.com.au> wrote >
>
> >> The perfect thing would be a pair of amps with 3 x 300B in parallel for
> >> 24W max into 8 ohms, with Rout = 1 ohm, like a pair of Quad-II, but with
> >> less NFB and **ALL** class A.
> >>
> >> 1 x 211 would be nice, but forget 845 because they seen to have become
> >> outrageously expensive if you get the KR Audio types.
> >> The Chinese seem to have stopped making 845.
>
> No loss. I have yet to hear an amp running 845's that didn't sound as rough
> as bags. Not a pleasant audio tube IMO.
I think your'e wrong, and so did the guy who bought the amps at
http://www.turneraudio.com.au/monobloc845se55.html
>
> >> Frankly, 3 x KT88 in parallel SE triode would also be nice IMHO, but
> >> then you have prejudices about what sounds good and so don't take any
> >> notice of what i might say to the group.
>
> > Thanks for the reply. I love 300Bs but 3 in parallel would need a serious
> > SE OPT. I have some fancy 3K SE OPTs which are gapped for 100mA.
> >
> > I'm not a fan of big transmitting tubes like 211 or 845. I've built
> > serveral 845 amps, and I always found the sound to be coarse (or maybe
> > that was just the Chinese 845s I used), although they did have tremendous
> > clout and bass welly.
> >
> > I'm not prejudiced against KT88s (or KT66s, of which I have many), as long
> > as they are triode-strapped. I would rather suffer from scrotal
> > elephantiasis than listen to a pentode or tetrode. I would never build
> > anything with 6550s - they are not audio tubes. Sorry, I know you love
> > them, but they suck badly.
>
> Dunno about that Doug. I have heard a few amps that had their existing tubes
> replaced with 6550's & the results were quite audibly pleasing, even though,
> as
> you say, 6550's are not audio tubes.
When are you guys going to wake up that 6550 **ARE** meant for audio use
or any general purpose?
Its just another stupid myth that the 6550 is in some way inferior for
audio use compared to tube X or Y favoured by the cognescenti papaazzi
bullshit artists who are technically and mentally deficient and who
remain one eyed, biased, and perpetual resistors to cold hard realities.
I often wondered what category into which I fitted but I found out after
building and listeing to many amps often using the very opposite choice
of tube to those of the die hard rebels who refuse to even subject
themselevs to any kind of AB testing or comparisons with an open mind.
KT66, KT88, KT90, 6L6, 6L6GC, 5881, EL34, 807, EL84, EL36, EL37, 13E1,
12E1, and many many other tubes **are audio tubes**.
They can be used at F between DC and high radio F.
Patrick Turner.
>
> ruff
Patrick Turner
But no matter what the original makers may have had in mind when they
made the first 6550, there will always be people who think the 6550 is
not an audio tube.
Its just as silly a point of view as anyone who dismisses some varieties
of power transistors or mosfets or DH triodes.
There used to be silly arguments about which small signal transistors
sounded the best. That moved to opamps and the jury is still out....
If you bite the top off a 6550, it'll pull your tongue inside because of
the vacuum. This is when they suck.
And it ain't like being kissed by Jennifer Hawking.
Patrick Turner.
Patrick Turner
The 6550 made by Sovtek or Electro Harmonix is **exactly** the same
metal work and electrode work when you examine them or measure them for
ra, Gm. and µ and THD and IMD etc.
There used to be a practical real difference between original KT88 and
original 6550. But those historic differences have become diluted, and
because NOS originals are no longer available at a decent price the
differences are totally irelevant and not worth the persuit.
There is fuck all difference Doug.
Then there are other copies of 6550 by Svetlana and KT88 by quite a few
including Chinese makers and ALL are capable of excellent listening
depending on who makes the amp and the techniques they use. The Chinese
have the worst rep for reliability although I've known many chinses
tubes to last a long time. One guy I knew had 4 different brands of EL34
in his re-engineered ST70, including two chinese tubes. Music was
excellent. You could plug in just one 6CA7 and nobody would notice at
all.
Patrick Turner.
Patrick Turner
Indeed 6550 *ARE* as linear as KT66 or KT88.
We would have you accept that indeed 6550 are just as linear as most
other beam tetrodes and that your stance on the 6550 is not based on any
known facts.
For example, a PP pair of 6550 without any NFB loops can produce 40W at
2% THD into 4k a-a in pure class A, working in beam tetrode mode with a
fixed screen voltage.
This meant that at 1 watt the THD would be less than 0.2%. But 20dB of
NFB is needed to reduce the amp output resistance to get a good damping
factor of at least 10. So THD would then fall to 0.02% at a watt which
is above normal listening levels.
Makers used the 6550 **because** it was so damn linear.
I suggest you check the technical data sheets before condemning the 6550
"because it isn't linear".
In triode mode, they are even more linear and with the advantages of
better reliability than a pair of 300B. KT90 made now by EH in Russia
are like 6550EH or KT88EH but with 10mm longer anode structure and thus
higher pda rating so allowing slightly more class A from a given number
of tubes. And the data curves drawn in the late 1950s when these tubes
were first popular often show poorer linearity than actually existed
because the test gear was all tubed and not highly linear itself. But
the non linearities of all the known tubes used for audio didn't matter
too much because everyone knew that PP class A gave blameless
performance with less artifact production than that produced in the
studio and LP processing.
Its the class A that counts Doug, and all the big octal beam tets are
capable of fabulous sound. You hear what others just don't.
The fingers down the blackboard sound is a myth Doug. I have compared
these things A to B and I know what a myth is.
The only time you get the finger scratching sound is when we might
record the sound Doug makes trying to escape from the argument holes
he's dug himself :-)
Just about all beam tets or pentodes measure similarly. Many specs quote
THD at 10% or more when used in SE mode, usually at Ea = 250V and
appropriate Ia. They make more THD than triodes, and the mix of
artifacts is both odd order and even order and varies in relative amount
and even with relative phase with changes of load value. But they also
make more power and wat for watt the THD difference isn't so much.
They are similar to BJTs and mosfets with their raw open loop
behaviours.
All the devices known make distortions Doug. Just try measuring a few
well known tubes and you'll find out.
If you have not, and have concentrated on making amps with only real
triodes, but have managed to set the loading optimally then you'll get
good sound without much NFB needed, or any need to measure anything
much. Its voltmeter engineering, and how most of my father's generation
coped because they had nothing else but a voltmeter to build something,
except very clear ideas about the basics of what went on inside tubes.
You may well be living with higher measurements of N&D than I like to
know exist, and you'll get the colouring they offer but only if they
measure high enough. If you have sensitive speakers and never ask more
than a watt from an 8 watt capable 300B SET amp then your triode
linearity will be just fine. And so too will be the sound of a well set
up 6550 with just enough NFB.
Given a choice, I'd rather have 5 watts from an SE trioded EL34 than 5
watts from an EL84 in pentode mode. I know, because I have two such amps
I have recently compared with same speaker and FM mono source. The SET
uses 12dB NFB and the SEP uses 20dB NFB. But at low levels there is SFA
difference. When good music plays I turn up this set in my kitchen and
then the '34 gives better perfomance into the 1953 Rola Deluxe speaker
with dome tweeter from 1974.
A 6550 would be better, but I don't need to use one.
However, when used in PP and with the UL or 10% CFB connection as in
Quad-II or 20% CFB as in my amps, the multigrid becomes more linear than
a triode which has internal NFB between the anode-grid-cathode
electrostatic fields.
I have found it possible to make SE amps using beam tets with local CFB
which are better than triodes, without any global NFB.
The local CFB around the OPT works better than the internal
electrostatic NFB of the triode.
I very much like 2a3, and 300B, but I've never wanted an amp with these
tubes because they simply don't sound better than other more modern
cheaper tubes used more easily. And the old DH triodes are all the more
difficult to make silent without hum or microphony.
I believe you don't have to have much microphony to make a tube amp's
microphonic resonant artifacts exceed the THD and IMD levels.
Therefore some of the colur heard in some tubes becomes due to subtle
microphonic effects. If the "ping" in the tube is at a musical note of
the key of the music then the music may be enhanced by the extra
"musician" in the tube. But I'd rather I have the least microphony
possible thankyou, so I stay away from micro tubes.
I witnessed a line level preamp the other week which had 6SN7 which were
so worn out and old they had become loose and the amp could be excited
to accoustic feedback if the tubes were tapped with volume up a little.
The sound was still OK when not on the brink of acoustic howl, but not
as good as it could have been if the tubes really were NOS.
Patrick Turner.
Patrick Turner
Probably Doug believes the marketing and the whole promotion of tubes as
the 6550 being presented as a real audio tube was all a big lie.
For him, the 6550 is a dog turd of a tube and terrible sounding, and the
more we try to reason with him the more determined he will stay with his
beliefs.
Once someone said the Earth was a big ball and it went around the sun.
The popes and bishops didn't agree. He was right this someone.
In a post de-constructionalist post modern world truth is bullshit and
bullshit is truth and *WHATEVER* covers anything and everything.
This means nobodys right or wrong and everyone's right if they believe
they are.
Its brought the world to the mess its in, but so what, whatever..
I don't have the time to unlock Doug from his beliefs and I only have
time to say he may discover his belief may need some review if he built
and tested an amp or two using 6550 with adequate zeal and skill and
compared the results with say something he likes with his favourite real
triode.
I'd like to paddle around the lake here in a plywood canoe I made myself
but the experience of the lake could be just as good in a plastic canoe
I bought at a shop. The ducks wouldn't notice any difference.
Patrick Turner.
roughplanet
roughplanet wrote:
"Doug Flynn" <doog...@bigpond.com> wrote in message
news:%c_Al.333$y61...@news-server.bigpond.net.au...
>> >> The perfect thing would be a pair of amps with 3 x 300B in parallel
>> >> for
>> >> 24W max into 8 ohms, with Rout = 1 ohm, like a pair of Quad-II, but
>> >> with
>> >> less NFB and **ALL** class A.
>> >>
>> >> 1 x 211 would be nice, but forget 845 because they seen to have become
>> >> outrageously expensive if you get the KR Audio types.
>> >> The Chinese seem to have stopped making 845.
>
>> No loss. I have yet to hear an amp running 845's that didn't sound as
>> rough
>> as bags. Not a pleasant audio tube IMO.
> I think your'e wrong, and so did the guy who bought the amps at
>
> http://www.turneraudio.com.au/monobloc845se55.html
>
I don't think I'm wrong, so I went & listened to an amp running 845's last
evening.
They STILL sounded awful - I wouldn't want them in my system, but then
again,
I prefer EL84's, a much more refined and accurate tube, so that probably
explains it.
I am going back at 3.00 PM this afternoon with my CD's to have another
listen.
The guy is the only person I know who runs the same speakers as I do
> themselves to any kind of AB testing or comparisons with an open mind.
>
> KT66, KT88, KT90, 6L6, 6L6GC, 5881, EL34, 807, EL84, EL36, EL37, 13E1,
> 12E1, and many many other tubes **are audio tubes**.
>
> They can be used at F between DC and high radio F.
ruff
roughplanet
"Doug Flynn" commented:
Welcome back Philthy :-).
ruff
Doug Flynn
> When are you guys going to wake up that 6550 **ARE** meant for audio use
> or any general purpose?
>
> Its just another stupid myth that the 6550 is in some way inferior for
> audio use compared to tube X or Y favoured by the cognescenti papaazzi
> bullshit artists who are technically and mentally deficient and who
> remain one eyed, biased, and perpetual resistors to cold hard realities.
> I often wondered what category into which I fitted but I found out after
> building and listeing to many amps often using the very opposite choice
> of tube to those of the die hard rebels who refuse to even subject
> themselevs to any kind of AB testing or comparisons with an open mind.
>
> KT66, KT88, KT90, 6L6, 6L6GC, 5881, EL34, 807, EL84, EL36, EL37, 13E1,
> 12E1, and many many other tubes **are audio tubes**.
807? GMAFB. Perhaps you should upgrade your shitty Dick Smith speakers......
:-)
Doug Flynn
"Patrick Turner" <in...@turneraudio.com.au> wrote
> I very much like 2a3, and 300B, but I've never wanted an amp with these
> tubes because they simply don't sound better than other more modern
> cheaper tubes used more easily. And the old DH triodes are all the more
> difficult to make silent without hum or microphony.
The word "cheaper" pretty much sums you up, doesn't it? You're not Scottish
by any chance are you? (LOL)
If you think that some POS TV or geetar amp tube you found in a dumpster
sounds better than a DHT then you need to visit your doctor to get your ears
examined. Yes, they are finicky to work with, but when properly done a DHT
will jizz all over any other type of tube.
Doug Flynn
>> I'm not a fan of big transmitting tubes like 211 or 845. I've built
>> serveral
>> 845 amps, and I always found the sound to be coarse (or maybe that was
>> just
>> the Chinese 845s I used), although they did have tremendous clout and
>> bass
>> welly.
>
> You were a huge fan of such tubes. I always thought maybe you never knew
> quite enough to get the best sound. Maybe if you tried again you'd have
> better success.
I got "good" sound out of 845s, but they were comprehensively beaten by
lower power triodes like 45s, 2A3s, 10Ys, PX4s and 300Bs
> How you drive the large transmitting tubes is just as important as the
> output tubes and exceptional attention needs to be given to power
> supplies. Then its easy to get nearly all 50W SE amps to sound better
> than most 70W PP amps no matter what they are.
You're on the money there. I found that big tubes take on the flavour of the
driver. Use a 300B and it will sound like a 300B amp on steroids (but it
will lose some of the charm of the lower powered tube in the process). I
have used all sorts of drivers for transmitting tubes - 300Bs, 10Ys,
triode-strapped KT66s/EL34s/6V6s, etc. Direct coupled, cap coupled,
interstage transformer coupled. Massive power supplies. You name it. Been
there, done that.
> I subsequently wowed the crowd with SE amps based around the 13EI, a big
> fat *general purpose* beam tetrode
> capable of a very refined 32 watts in SE class A.
I've heard that amp - it's very nice indeed (there, I'm not prejudiced!).
> I think its down to who builds the amps and their attention to detail
> and basic issues of linearity, bandwidth, stability, output resistance,
> low THD/IMD and low noise etc. Once you have got all those right with a
> mimimum of NFB then you worry about the brands of caps and type of
> solder and cabling. And the AB tests usually tell people there isn't
> much difference in cables, caps or solder.
Here we go; anything to avoid shelling out for decent components. Better nip
down to Jaycar or DSE to pick up a bag of 100 cheap-arse poly caps for
$5.99.
I haven't got time to reply to the rest of your post, which strangely
resembles the last 400 pages of Ulysses...... :-)
Patrick Turner
I have no idea who may have made the system using 845. This affects the
sound. Then you need to know about the speakers and preamp and CD player
etc etc, because **ALL** the items in the line up are contributing to
the sound.
If you were to pull out the 845 amps and insert the EL84/6BQ5 amps, and
leave every single other thing the same and MEASURE the level setting
then you can make some sort of informed perception.
>
> The guy is the only person I know who runs the same speakers as I do
My 845 amps make 55W max and had all the precision and detail anyone
could ever wish for and gave the heart strings a real tug with emotional
music. Real accuracy does that, and IMHO triodes do it best. Its noise
and distortion measurements are much lower than anything with EL84 for
the same power level. So I know the 845 amps **are** more accurate. I
very much like the EL84 and use them routinely as driver tubes for
output stages, and in triode, because one EL84 can do what 3 x 6SN7 can
do and to drive the pair of 845 which require a clean 106Vrms to their
grids I use 3 paralleled EL84 in triode and set up with a choke and R
load for dc and an ac load of 22k. The THD at 106V is 1.4% (55W output)
and partially cancels the THD made by the 845. At 1 W the amps have
utterly negligible noise and distortions. EL84 at 1W would have 5 times
the THD/THD and to say they are more accurate is utter BS. But indeed
you may perceive they sound more precise. That is the paradox of
measures worse but sounds better. I've re-engineered quite a few amps
with PP EL84 which give 9W max in pure class A allowing for the usual
OPT losses in small OPT, or about 15W in AB1. They are not bad
performers until you turn up the wick or want to hear the full drama of
a big orchestra. There was also the Musical Reference amp with PP EL84
which used 700V for the anode supply, 350V for the screens and produced
35W in nearly all class B and with only moderate NFB. That to me sounded
a bit rough, and certainly not better than most ordinary other amps.
Baird used a sixpack of EL84 to make 60W max again with little class A
%.
I would rather have the 8W from a well done lone 300B than 9W of class A
from PP EL84. But then of I had a pair of KT66 in triode for the same 9W
that would be better than EL84.
The amount of NFB also has enormous effect, and without GNFB the EL84 in
UL or pentode is bloody awful.
I suspect you hate NFB as well as 6550 but I have tried to make really
nice triode amps without any GNFB and have achieved good N&D figures and
had Rout as low as 2 ohms but customers **preffered** to have the NFB.
The classic case was where I made a 10/10 amp using 6GW8 as the input
and output tubes for 10W per channel in UL and with 16dB GNFB. The owner
wondered why such a toy could sound so much better than his Cambridge
and Redgum SS amps. His wife could hear the difference when he changed
amps even when she was sitting onm the loo up the hallway.
Then he found a pair of Quad-II in poor condition, so I tried EL34 in
triode with 6dB GNFB. The transients and dynamics were not as good as
the 1010. So out went the EL34 and in went KT88 with fixed bias in
triode and 12dB of GNFB. Then my customer told me the reformed Quads had
the punch of the smaller amp but had a much better ceiling (because 20W
was available in triode AB1).
6GW8 and 6BM8 have small beam tetrodes within rather much like EL84
pentodes and were used in countless 10W amps when speakers were
sensitive enough to allow their use. So whatever ppl say about EL84 can
be aplied to other similar pentodes or beam tetrodes.
When I have made AB comparisons the speakers are switched between amps
being compared. My preamp has two sets of L&R outputs each with its own
gain control so the input levels to both pairs of power amps can be
adjusted while standing between the speakers and switching the speakers
until no level change is perceptable. This usually gets levels within
+/- 0.5dB.
I have then asked listeners to identify which amps they think are being
used. Rarely can anyone pick a difference.
Then I change the level of one power amp up about +1dB and hand the
switch control to the listener. He'll swith back and forth and tell me
which is the more detailed sound. Its ***always*** the slightly *louder*
amp.
I've repeated the test to make the other amp louder and then they pick
that as being "better".
My conclusion is then that amp A = amp B in the goodness stakes. Both
are doing equally well with music.
Speakers can also be tested similarly. I will leap frog the quality
upwards with real tests that mean something. In other words I've built
speaker A, then compared it to speaker B and once B begins to sound
better than A then B becomes the gold standard. Then I make C and if
that's better than A or B then C becomes the gold standard. I would
never sell a pair of speakers then build or buy a new "better" pair
without confirming that the betterment was really real. This has saved
me thousands of dollars and months of creative effort which may have led
knowhere, and perhaps to worse sound than when i started.
I've got my system to a point where I don't know how to make it sound
better. I'd be fooling myself to think much betterment lurks out there.
This leaves me to be creative about systems I make and sell to other
people.
Speakers and rooms are renowned for condemning the amps ppl choose to
drive them. Both speakers and rooms have to be exceptionally good to
make any sort of worthwhile perception about amplifiers. I once built an
elaborate pair of 3 way full range speakers using old 12" electrovoice
woofers, 1992 made Chinese 8" paper cone drivers and Response dome
tweeters. I laboured away fro weeks to get the crossovers to give a flat
response without the crossovers become too complex or offering the amp a
sour load to deal with. I learnt a whole lot more about LCR theory and
built a decent pinknoise source and 1/3 octave switchable band pass
filter with Q of 12 for each band.
Finally, the measurements began to make sense and the peaks in the
response were tamed and the sound improoved.
Then I tried using SEAS drivers for mids and treble. What a revelation!
These beat the pants off all the work with asian made drivers, so I
realised you can't shoe horn drivers to give music unless they have
tight tolerances and decent manufacturing standards. The SEAS were chalk
and cheese compared to the asians mades and the sound FAR better. It
only took two lines of music to find out. But the 12" bass speakers
using the old Electrovoice were still slightly better than the SEAS 8"
with Al cones. The EV bass units have massive magnets and are the real
McCoy. Decent US mades. I'd got the math right for the reflex enclosures
weighing 90Kg and of 135L internally with Fb = 15Hz with Fs = 25Hz and
with cross to MR at 250Hz. Its dreadfully important to NOT let a big old
wooly woofta spoil the midrange by being allowed to go up to 1khz like
many old speakers did to allow makers to use a cheap nasty midrange.
Sound quality overall is mainly determined by the midrange and tweeter
driver quality and the flatness of response. Unless you have carefully
measured your DIY efforts then building DIY speakers is *always* a hit
and miss affair and you just cannot rely on your ears to tell you how to
fine tune a crossover, or whether an amp is really any better or not. If
your room has acres of bare floors with few furnishings with little
reduction of reverberation and reflections then all expense on hi-fi is
in vain.
I'm often amused when I hear of DIY efforts where people really don't
understand the engineering required. If they ever get good sound, its
always a mere accident.
And I'm constantly led to seat in front of expensive speakers only to
find they dissapoint, and one has to be ever so polite.
Perhaps some examination and reform with a soldering iron is needed
along with guidance from the measuring gear but that all costs and takes
time.
The most amusing are the guys with 4 amps per channel for 4 bands. Then
they are always changing something and yet thay have no clue about
filters and attenuation slopes and relative phase or how amps work. Its
The Best recipe for complete self be-fuddlement.
There are some good online sources of programs for pink noise measuring
of DIY speakers and any audio amateur MUST have such a thing and have a
decent microphone and know how to use all it all if he expects to do
better than what is sold in shops at outrageous prices.
Patrick Turner.
Patrick Turner
I don't use or own "shitty DSE speakers". I did buy some asian made
midrange drivers from Jaycar in 1994 which I thought I could tame but
when I went to using SEAS drivers the asian mades now gather dust.
Response tweeters made then are also garbage.
The 807 *is* electronically identical to a 6L6 but with an anode top cap
to reduce capacitance and possibility of arcing at the tube socket. Thus
807 can be used with Ea = 600V and can have an extended F range well up
into the HF band.
Although you may despise the 807, it IS an audio tube becaue it can be
used for audio and it was very popular in the 1950s because it was cheap
and thousands were available from army disposal stores until the late
1960s some 20+ years after WW2 ended. They were made for **general
purpose**, ie, between DC and HF, and often used in audio amps and for
AF modulators in RF transmitters. Their socket config makes them far
more reliable than most octal tubes. It was standard practice by audio
hams in 1960 to use a pair of 807 in UL or triode and there wasn't much
need for anyone to buy pretentious KT66. The 807 is a fabulous tube. In
a retro design it also looks really well.
Later in the 1950s when "TV tubes" were being developed there were many
tubes which coud be used for audio but which were mainly intended for
shifting the scanning lines of a TV set. This group included the 6DQ6
and 6CM5, 6CD6, EL509 etc, etc. Among these groups many were used by Oz
makers of guitar amps. The 6CM5 or EL36 stands out and became a
favourite for radio hams in DIY trans mitters but they excel in triode
and can be used in UL amps quite well. 6CM5 have max gm of 4 times a
KT66, so they very gutsy, and in triode will rival a 2A3. They sound
well in triode PP and can easily make 25W in Ab1 but with a lower Ea
than needed for EL34/KT66 etc. They never became the hi-fi darling of
the month and have an anode top cap to cope better with HV. Bit in the
US there was the 6FW5 which is an octal without a top cap and has
exactly the same innards as the 6CM5, so RCA must have wanted to make a
more audio specific tube more likely to be used by makers. Hardly anyone
did.
But Sansui had a receiver with 2 PP channels with 6CM5 which made 44
watts/ch. This was very high power for the period, but was mainly class
B and not as clean as one could get in PP triode. 6AU5 is another TV
tube capable of audio wonders.
Along with TV came tubes like the 6EJ7 which is a small 9 pin frame grid
pentode with gm up to about 12mA/V. In triode it is a real glory and has
a µ of 60.
Radford used 6EJ7 as drivers in their 100W UL amps with KT88 outputs. I
think the BBC bought piles of Radfords. They were supposed to be far
better than the Leaks and Quads the BBC also bought. Back then they were
used in pentode to get high open loopgain to allow at least 20dB GNFB
and then still get a very sensitive power amp. Some amp makers bragged
they had up to 30dB of GNFB which gave THD levels at 35W = 0.03%, and
below 0.01% at 3.5W. But stability was dodgy. When using 6EJ7 now you'd
ignore the pentode mode and make the LTP driver stage in triode mode and
maybe use one as the input tube. You'd still have too much gain though,
and nobody uses them when 6SN7 or the 9 pin derivative the 6CG7/6FQ7 has
just the right gain and is extremely linear and good sounding.
I use a 6EJ7 as a triode in a phono input stage and cathode driven by a
j-fet that's 10 time quieter than nearly any tube. Sounds just fine.
So Doug, it ain't always wot ya got, its how ya use it that counts.
Patrick Turner.
Patrick Turner
Doug Flynn wrote:
>
> "Patrick Turner" <in...@turneraudio.com.au> wrote
>
> > I very much like 2a3, and 300B, but I've never wanted an amp with these
> > tubes because they simply don't sound better than other more modern
> > cheaper tubes used more easily. And the old DH triodes are all the more
> > difficult to make silent without hum or microphony.
>
> The word "cheaper" pretty much sums you up, doesn't it? You're not Scottish
> by any chance are you? (LOL)
You completely misunderstand me. Try purchasing something from me and
you'll soon find how expensive it is to deal with me.
You are an ignorant little snake to suggest I base my amplifier
productions on what is only the cheapest solution.
You once came by to get some paper in oil caps I had and you wouldn't
pay me a cent. I'd paid $5 each for the 10uF x 1,000V rated caps but
that was all too expensive for you. You were the one to always only use
something if it came free, or nearly free.
You advertised in the Canberra Times to get people to donate tubes.
And what sort of gear did you build? extreme DIYer messes all over your
loungeroom floor and chockoblock with wiring with aligator clips in
temporary circuitry. I'm sorry Doug, but I've never missed not hearing
your systems. I heard from some people who had had a listen that they
were not too impressed.
Audio is all just fun to you Doug, but I expect what I make to appeal to
other people with real money, and they are often very difficult people
and hard to please, and that means serious AB comparisons, and NO
cheapskating.
When did you ever go out of your way to convince anyone other than
yourself that your audio methods would be widely accepted? You've never
asked me around to have a listen.
You have never had any respect for what I've made but lemme tell ya,
what you make with scrounged up bits and pieces doesn't interest me, its
unsaleable junk, and you are going to have to move some mountains before
you gain some respect.
Try building amps you can sell to discerning clients with better hearing
than yours.
I KNOW that paying the stupid high prices being asked for some tubes
does not always buy better music, but If I am asked for the best and the
most expensive then that's what I use, OK.
Take the KR845 for example. I spent $375 about 2 years for each of 4
tubes to put into the 845 amps I made last year for a discerning
customer. $1,500 for 4 output tubes to make only 110W total.
But after 18mths since I made the KR tube purchase the supply of KR845
to Oz seems blocked by the US company in cahoots with KR Audio in
Prague. KR refers me to their US pals and the US pals won't sell them to
me. The 845 was listed at usd $534 which would make it cost about aud
$900 landed with freight and GST and insurance, if I could get any. So
when I get requests for 845 amps I tell ppl they can't have any, and if
they could, the tube price alone would be $3600 for 2 x 55W amps.
The client who purchased the 845 amps I did make can have the amps
changed to suit a quad of 6550 if there are no replacement 845 tubes to
buy when one or two wear out or fail. They can also take 211. I made the
amps so minor alterations will allow a range of tubes.
I'm currently making a pair of 60W SE amps with a six pack of 6550. The
client has insensitive speakers. These will be slightly better than the
845 amps. But the tube cost is aud $480 max for the dozen 6550.
>
> If you think that some POS TV or geetar amp tube you found in a dumpster
> sounds better than a DHT then you need to visit your doctor to get your ears
> examined. Yes, they are finicky to work with, but when properly done a DHT
> will jizz all over any other type of tube.
I've never used any tubes I found in any dumpster and I never inferred
that I thought a budget musicician amp would sound as well as or better
than something really hi-fi with DHTs.
I like DHTs, but they simply are NOT superior in sound compared to other
later tubes. Equal maybe, depending on the app.
If you don't know how to get good music from a 6CM5, 6L6, EL34, etc,
then you know nothing Doug. You may think you've made such a big effort
to find out, but where is the real evidence? where are your amp samples
for people to hear for themselves?
I don't rely on my hearing alone Doug. I don't have fanciful ideas about
myself. I rely on the ears of my customers who I have to please. Their
hearing is often better than mine.
There are people like Trevor Wilson who display they have a concrete
brain when it comes to SET amps, and you are the same with regard to
using multigrid tubes.
OPEN your mind Doug. Try opening both eyes and ears. There's a whole
bunch of ppl who'd disagree with nearly all your beliefs about what
makes a good sound system. They all have ears. Some have better ears
than yours.
I usally don't get on for very long with people who have totally
irrational audio beliefs and who tell me what I make and what i use is
all crap.
They hate it when I ask them to proove what they claim to be true, and
at their expense, and next week if possible, and with a demonstration of
well made gear using samples of tubes they profess to sound the best,
and those they say they hate. And of course the samples of their work
must be made to a high standard, without the trickery of putting a hated
tube into a circuit that ensures high distortions and lousy sound.
Usually they themselves struggle to understand any of the basics about
electronics and they struggle with the science and discipline required
for good music.
They tend to be in a little world of their own where they nourish their
pet theories about what they think they hear.
But other folks hear pretty well you know, and if you are so right, then
convince others around you besides yourself.
I've been there and done it Doug, you need to do a bit more before you
publically rubbish the 6550 and many other renowned techniques I use in
audio gear.
Unlike you Doug, I have built amps using old designs of DH triodes,
later multigrids, bjts, mosfets, and a whole range in fact.
I hate not one single device I have ever used. I was, after considerable
work and study, able to get music from the mosfets and bjts which was
quite acceptable. I've had audio meetings here where I have changed from
a heavy class A tube amp to a mosfet amp and nobody noticed for the next
1/2 hour when I said I'd changed amps. When tube amps are good, they are
extraordinary, and it may be said of some SS amps. Sadly many SS amps
are not, along with many mainly asian made SET amps and PP amps.
I have not built a "digital" PWM amp but I once had a pair here to test
and they sounded quite as well as an average class B SS amp, so all the
class B SS amps will be gradually replaced by digital amps and the
public will not object. Some of the SS amps with high class A might be
better though and like tube amps they'll be in demand.
I spent 3 weeks last year re-designing and reforming a large VAC amp
with 4 x 300B per channel to make up to 2 x 70W in AB1.
This POS from the US had so many design shortcomings which meant N&D was
high and tubes kept dying. I could write a book about the faults.
But I did manage to get it to be reliable, and noise free, and to reduce
N&D and Rout greatly while bettering stability and the amp now seems
quite blameless and as good as anything with 6550, but unable to make
the power you'd get with 6550. At low levels of a watt the N&D is barely
measurable and the sound was very fine, like using 6550.
I now have a pair of Dynaco MkVI amps with 4 x 6550 in each amp for
total revision. But some galoot has altered the crummy Dynaco standard
input circuit to some wacky thing using twice the number of tubes but
achieving virtually no huge THD/IMD reductions. Then they have undone
the supposedly evil ultraliniar screen taps and converted to plain beam
tetrode while using the whole spaker winding to give CFB of about 5% of
the tube voltage, ie, virtually no local NFB, maybe 3dB only. GNFB has
been abandoned. So the amps sure make lots of power but also make lots
more distortions. The owner from Byron Bay was disgusted by the sound of
what had been done by this galoot who clings to a silly theory and
implements it to worsen N&D, raise output resistance, and generally
degrade the performance.
When I am done he will have the best music he will ever have heard from
tubes.
You see, you couldn't rightly conclude that 6550 suck because these
particular amps sounded so bad.
Dynaco amps are a bit primitive to me even in their original form and
like nearly everything from the US UK or China or elsewhere, need major
design revisions and a big make over and then I defy anyone to
demonstrate that 6550 are bad.
Patrick Turner.
Patrick Turner
The flavour is that of the total.
>
> > I subsequently wowed the crowd with SE amps based around the 13EI, a big
> > fat *general purpose* beam tetrode
> > capable of a very refined 32 watts in SE class A.
>
> I've heard that amp - it's very nice indeed (there, I'm not prejudiced!).
Well then you can't honestly say 6550s suck. 13E1 are more non linear
than 6550 in pure beam mode.
>
> > I think its down to who builds the amps and their attention to detail
> > and basic issues of linearity, bandwidth, stability, output resistance,
> > low THD/IMD and low noise etc. Once you have got all those right with a
> > mimimum of NFB then you worry about the brands of caps and type of
> > solder and cabling. And the AB tests usually tell people there isn't
> > much difference in cables, caps or solder.
>
> Here we go; anything to avoid shelling out for decent components. Better nip
> down to Jaycar or DSE to pick up a bag of 100 cheap-arse poly caps for
> $5.99.
>
> I haven't got time to reply to the rest of your post, which strangely
> resembles the last 400 pages of Ulysses...... :-)
Doug, you are a snivelling little snake. You don't like robust
criticisms and you wally around with BS beliefs and you have nothing to
show other folks to prove your points are correct.
Then you insult me by telling lies about what I put into amps.
I sell these to people who I believe may have better hearing than yours.
Patrick Turner.
roughplanet
Doug Flynn wrote:
"Patrick Turner" <in...@turneraudio.com.au> wrote
>> >> I'm not a fan of big transmitting tubes like 211 or 845. I've built
>> >> several
Hey Patrick, no need to hold back. Tell doug what you think of him without
needing to mince your words.
> Then you insult me by telling lies about what I put into amps.
> I sell these to people who I believe may have better hearing than yours.
You may well be right there Patrick. I have had the same power & preamps
since 1990 and my tastes, or rather my hearing doesn't seem to have changed
much since then, i.e. I still enjoy their sound & don't wish to change
ANYTHING
in my audio system at all.
ruff
Doug Flynn
>
>> I haven't got time to reply to the rest of your post, which strangely
>> resembles the last 400 pages of Ulysses...... :-)
>
> Doug, you are a snivelling little snake. You don't like robust
> criticisms and you wally around with BS beliefs and you have nothing to
> show other folks to prove your points are correct.
>
> Then you insult me by telling lies about what I put into amps.
> I sell these to people who I believe may have better hearing than yours.
Dude, lighten up. All the LOLs and smiley faces indicate my tongue is firmly
planted in my cheek. You are in danger of turning into Phil Allison (does it
worry you that you and Philthy agree on this topic?). Or you might give
yourself a stroke. Chill.......
Doug Flynn
"Patrick Turner" <in...@turneraudio.com.au> wrote
> You once came by to get some paper in oil caps I had and you wouldn't
> pay me a cent. I'd paid $5 each for the 10uF x 1,000V rated caps but
> that was all too expensive for you. You were the one to always only use
> something if it came free, or nearly free.
I can't believe you still wet your panties over that episode. I recall I did
in fact pay you for those caps. You were only too happy to part with them.
> You advertised in the Canberra Times to get people to donate tubes.
Erm, no I didn't. It was a wanted add like any other. No-one "donated"
anything. Hang on, you've given me a good idea......
> And what sort of gear did you build? extreme DIYer messes all over your
> loungeroom floor and chockoblock with wiring with aligator clips in
> temporary circuitry. I'm sorry Doug, but I've never missed not hearing
> your systems. I heard from some people who had had a listen that they
> were not too impressed.
Dude, those were test rigs. Sheesh......
> When did you ever go out of your way to convince anyone other than
> yourself that your audio methods would be widely accepted? You've never
> asked me around to have a listen.
That's because you're not one of the SET brethren. We're like a religious
cult. We don't tolerate outsiders :-) :-) :-) (note multiple use of smiley
faces).
> You have never had any respect for what I've made but lemme tell ya,
> what you make with scrounged up bits and pieces doesn't interest me, its
> unsaleable junk, and you are going to have to move some mountains before
> you gain some respect.
Erm, that's not true. I think you make good stuff (that 2A3 SET amp wasn't
half bad, although I wouldn't have used such a lame driver - 12AU7, c'mon,
you can be more adventurous than that). And you *are* a tight-arse when it
comes to caps....
> I like DHTs, but they simply are NOT superior in sound compared to other
> later tubes. Equal maybe, depending on the app.
> If you don't know how to get good music from a 6CM5, 6L6, EL34, etc,
> then you know nothing Doug. You may think you've made such a big effort
> to find out, but where is the real evidence? where are your amp samples
> for people to hear for themselves?
We'll have to disagree on that point. I have done mucho experimentation (see
point above about the test rig - what do you think I used it for, heating in
winter?). What I can tell you is that DHTs will ALWAYS be superior to
multigrid tubes.
> There are people like Trevor Wilson who display they have a concrete
> brain when it comes to SET amps, and you are the same with regard to
> using multigrid tubes.
Guilt as charged your honour! Can't be bothered replying to the rest of your
post as I've now developed RSI from what I've typed so far. Where do you get
the time to write these posts????
Doug Flynn
> And what sort of gear did you build? extreme DIYer messes all over your
> loungeroom floor and chockoblock with wiring with aligator clips in
> temporary circuitry. I'm sorry Doug, but I've never missed not hearing
> your systems. I heard from some people who had had a listen that they
> were not too impressed.
> You have never had any respect for what I've made but lemme tell ya,
> what you make with scrounged up bits and pieces doesn't interest me, its
> unsaleable junk, and you are going to have to move some mountains before
> you gain some respect.
I just have to reply to this again. You have a nerve accusing me of building
"extreme DIYer messes" (I wear that accusation proudly!) when you yourself
create stuff which looks like it was styled by someone from the local
Spastic Centre. Behold the evidence:
http://www.turneraudio.com.au/preamp-10tube-integrated-2006.htm
The labels are written in *texta*, and wonky texta at that. At least I use
stick-on labels. Any my amps are much more user-friendly. Here's a pic to
prove it :-)
http://db.audioasylum.com/cgi/m.mpl?forum=tubediy&n=61053&highlight=pic+Doug+Flynn&r=
Mark Harriss
> Dude, lighten up. All the LOLs and smiley faces indicate my tongue is firmly
> planted in my cheek. You are in danger of turning into Phil Allison (does it
> worry you that you and Philthy agree on this topic?). Or you might give
> yourself a stroke. Chill.......
>
>
Maybe people might listen to your personal opinions more if you
didn't push them as unassailable solid fact and then immediately dive
into the
personal insults more than once, the instant someone has the temerity
and gall to disagree with your "truth".
Then you insult again while telling them to lighten up?!. Sounds like
the last resort of someone who knows they are in the wrong and fighting
dirty.
There are an awful lot of 6550's still being used in amps when there
are plenty of other choices available. Is the rest of the world wrong
while you are the Galileo of valve audio, somehow I don't think so.
keithr
>
>
> Take the KR845 for example. I spent $375 about 2 years for each of 4
> tubes to put into the 845 amps I made last year for a discerning
> customer. $1,500 for 4 output tubes to make only 110W total.
>
> But after 18mths since I made the KR tube purchase the supply of KR845
> to Oz seems blocked by the US company in cahoots with KR Audio in
> Prague. KR refers me to their US pals and the US pals won't sell them to
> me. The 845 was listed at usd $534 which would make it cost about aud
> $900 landed with freight and GST and insurance, if I could get any. So
> when I get requests for 845 amps I tell ppl they can't have any, and if
> they could, the tube price alone would be $3600 for 2 x 55W amps.
>
roughplanet
news:49d879b9$1...@dnews.tpgi.com.au...
Patrick Turner wrote:
They can't really, I used to have almost 2000 tubes, plus a nice AVO MK II
tester whih I sold to a tube reseller, less a few I kept for myself.,
including a set of 8 Mullard gold pin 'Made in Holland' EL84's with the long
black getters which absolutely sing in my TT amp. Price? Around $200 each if
you could get a set, and these are all matched with the gain (Triode) 1 & 2
right on the money and no noise or microphonics either.
Want to buy 'em? Nah, not for sale :-).
ruff
keithr
I suppose that I can answer my own question - because people are silly
enough to pay those prices
> Want to buy 'em? Nah, not for sale :-).
>
> ruff
>
Wouldn't want them at the moment anyway. I might build a tube amp as a
project when I retire (if I can still afford the tubes) but hopefully
that isn't for a couple of years yet. Till then, I'll stick with "Good
enough" solid state gear.
Keith
Patrick Turner
It always smile when guys keep trying different amps and speakers after
they have bought something from me and they find that my gear becomes
the gold standard. And then I am always having to repair or modify what
they've bought that I didn't make to make the N&D lower and the gear
reliable.
If they have a problem with the "way anything sounds", I have to work on
it to get it to their expectations, not mine.
I'm very competive, and I always want my gear to sound better than
anyone else's.
Because I have a competitive nature, it means I sometimes may appear
grossly abrasive and a bully.
And so what do I say to those offended? Well, make better amps than I
do, simple.
When I am not in my work shed or online I am often out cycling around to
enjoy Nature and Fresh Air well away from the work and the Bad Airs of
the Internet.
And I find I often ride my bicycle faster for a longer distance than men
half my age, and they don't much like that either so my advice would be
always "ride more kilometres than I do and change your diet, and you'll
loose weight, go and faster for longer".
Alas that means they'd have to give up chocolate cake and cream, and a
string of sausages for dinner and umpteen cups of coffee, and that's
really difficult for some guys because they are married to ladies who'd
need to cook rather differently, or also be health nutters as they
should be in this absurd age where people mainly die from diseases of
kings, because everyone is enticed to live like a little king and
consume about +40dB more than they ever need to.
There are quite a few guys faster than me and they humble me on a sunday
ride, and most are 10 years younger, and when they pass me I just say a
cheery "Goodbye", because at 62 there is no way more training during the
week will strengthen me more or take 10 years off my age, so I admit I
am beaten by age. One is the veteran time trial champion of Canberra,
and if I spend 2 minutes behind him I nearly die a little bit, even when
going down hill. He doesn't much know how to ride slow, and I for one
refuse to spend all day frothing at the mouth as I have seen others do
to keep up with him. There's only one guy of 62 in my group who is
faster than me and he's a strippling of 65Kg, and he flies uphhill and I
really struggle at 85Kg. But along a flat road I can catch him. He's
never realised it'd be better to ease off a bit up hill so myself and a
few others can continue to rotate as a team on the flat thus raising
average speed and getting A to B sonner than if he tears off ahead up
each and every single fucking hill.
The only thing I don't like about some people who are better at doing
things than I am is when they don't like to be very sociable with anyone
slightly worse or slower than they are.
Meanwhile I quite like slowing down a bit while I ride to remain
sociable and jolly with those who may otherwise bust a gut staying on my
wheel.
I once used to race on bicycles 20 years ago, and over those 6 years it
taught me how to cope with nearly always loosing.
So if anyone has hurt feelings here, then tough titty, because I have
been gobsmacked and robbed of a win a thousand times...
Patrick Turner.
>
> ruff
Patrick Turner
I often agree entirely with Phil Allison's astute knowledge of basic
electronics.
He is less polite than I am, and has a very very short fuse.
This means I have to be very careful about what I type.
I plan to remain myself, happy that I ain't a lot of other folks.
Patrick Turner.
Patrick Turner
Doug Flynn wrote:
>
> "Patrick Turner" <in...@turneraudio.com.au> wrote
>
> > You once came by to get some paper in oil caps I had and you wouldn't
> > pay me a cent. I'd paid $5 each for the 10uF x 1,000V rated caps but
> > that was all too expensive for you. You were the one to always only use
> > something if it came free, or nearly free.
>
> I can't believe you still wet your panties over that episode. I recall I did
> in fact pay you for those caps. You were only too happy to part with them.
My recollection was that you didn't buy anything. That isn't any problem
Doug. I concluded that you seemed quite allergic to spending any money
on your hobby.
>
> > You advertised in the Canberra Times to get people to donate tubes.
>
> Erm, no I didn't. It was a wanted add like any other. No-one "donated"
> anything. Hang on, you've given me a good idea......
I recall you had quite a few parcels of tubes come your way.
Maybe not many that were useful.
Over the years quite a few have phoned me to donate boxes of old tubes
or other electronic gear they cannot sell or give away to a good home.
So I thus have a collection of hundreds of tubes. Most are garbage and
their condition unknown.
But it isn't bad to have a stock for old AM radios which ppl want me to
fix, when and if I ever make a spare week available.
>
> > And what sort of gear did you build? extreme DIYer messes all over your
> > loungeroom floor and chockoblock with wiring with aligator clips in
> > temporary circuitry. I'm sorry Doug, but I've never missed not hearing
> > your systems. I heard from some people who had had a listen that they
> > were not too impressed.
>
> Dude, those were test rigs. Sheesh......
Years and years of test rigs, and forever changing mind about which caps
sounded better.....
>
> > When did you ever go out of your way to convince anyone other than
> > yourself that your audio methods would be widely accepted? You've never
> > asked me around to have a listen.
>
> That's because you're not one of the SET brethren. We're like a religious
> cult. We don't tolerate outsiders :-) :-) :-) (note multiple use of smiley
> faces).
I quite like SET, as you should well know, and I know how to get
excellent music from DH SET amps.
Most of my ideas about how to go about this grate with all the low
priests I meet in your religious cult. I quite like NFB, using large
electrolytic caps, and bypaseed Rks, and IDH tubes 6550 which IMHO are
just as linear or more linear than a 300B in triode.
>
> > You have never had any respect for what I've made but lemme tell ya,
> > what you make with scrounged up bits and pieces doesn't interest me, its
> > unsaleable junk, and you are going to have to move some mountains before
> > you gain some respect.
>
> Erm, that's not true. I think you make good stuff (that 2A3 SET amp wasn't
> half bad, although I wouldn't have used such a lame driver - 12AU7, c'mon,
> you can be more adventurous than that). And you *are* a tight-arse when it
> comes to caps....
The SET amp with 2A3 can be seen at
http://turneraudio.com.au/4+4-set-stereo-2a3.htm
There is a 6SN7 clearly shown which is the driver tube ahead of the 2A3,
and each 1/2 of this driver is in parallel.
The input tube is indeed 1/2 a 12AU7, and it *DOESN'T* need to be
anything else because it has such an easy task of providing voltage and
current to the 6SN7 driver ahead of it.
I just DO NOT know anyone who doesn't like the sound I get with 12AU7 in
preamps or anywhere else. It could have been another 6SN7, or a trioded
pentode, or just about anything else with LOW GAIN and low Ra, but the
12AU7 is and ideal choice with 6SN7 so that the amp isn't too sensitive.
There is some GNFB in those amps, but only to get the Rout low enough
for most low Z speakers.
The gentleman who ordered those amps provided all the iron wound Hammond
parts after I pointed out that if I were to wind all of them the price
would be higher and the time taken far longer.
The gentleman had a VIP job in Canberra but in fact despite his high
social status he probably was paid poorly because of the foreign country
who employed him. So what went into this bargain basement re-engineered
amplifer was determined not by what I'd have liked to use but by what he
could afford, ie, quality was partially determined by finacial bugetary
restraints. He has a wife and children to consider, and I ain't no
hobbyist.
I an a commercial artisan, and I need to eat, a bowl of rice costs more
here than it does in China.
The 4W from the 2A3 is better than the 4W one might get with say 6BQ5
which needs a lot more NFB. But a trioded EL34 or 6550 is better IMHO.
>
> > I like DHTs, but they simply are NOT superior in sound compared to other
> > later tubes. Equal maybe, depending on the app.
> > If you don't know how to get good music from a 6CM5, 6L6, EL34, etc,
> > then you know nothing Doug. You may think you've made such a big effort
> > to find out, but where is the real evidence? where are your amp samples
> > for people to hear for themselves?
>
> We'll have to disagree on that point. I have done mucho experimentation (see
> point above about the test rig - what do you think I used it for, heating in
> winter?). What I can tell you is that DHTs will ALWAYS be superior to
> multigrid tubes.
I've had to work on several amps with 300B over the years. They have
sometimes been attrociously noisy and full of THD and high levels of
IMD, and often set up very crudely in cuicuits at way too high a working
temperature, ie, Pda = rated max Pda or even apve the rated max.
Recent examples of amps with many bothers are VAC, Audion, Sun and
Melody. When we carefully look around under the bonnet of all those
brands wer find just what a horror story they are in terms of basic
engineering. Once the basics have been fixed the used of exotic coupling
caps and paper in oil filter caps do utterly nothing to improve the
sound. My de-fault standard coupling caps are Wima MKP 400v to 1,000V
rated,
ie, normal industry standard very well made German polypropylene caps.
Unless someone can demonstrate to me in an AB test using two identical
amps that cap brand X id better sounding than cap brand Y then I assume
there isn't ever going to be any sonic difference providing the value of
the cap is the same in each test.
I have been able to compare the re-engineered samples of amps with DH
tubes I've worked on with those I've made with IDH triodes or
multigrids. The amps with 300B were never any better than amps of the
same power using half the number of multigrids.
I built a UL amp over last Xmas on an old Luxman chassis using iron
parts I have obtained from a Sydney based amp maker downsizing his
buiness since he retired. I used EL34 with 66% UL taps and 100% pure
class A for all loads above 5 ohms. When you have a high RLa-a for PP
amps there isn't any need to use 43% UL taps when taps can be 66% or
near full triode connection to get a maximum of 25W AB instead of the
lower AB max PO than you'd get with triode connection. So the "slight"
UL connection simply allows the operation of output tubes to remain
unfetted by the onset of grid current on peaks as happens with triodes
at highy levels. This amp was to power some speakers of 95dB efficiency.
So there was always plenty of power. With my own speakers of 88dB
sensitivity the amp sounded magnificent, ie, **could not be heard**,
because only the music could be heard, with all details preserved. I
tried the the same tests when I re-enginered a p[air of Audion amps with
2 x 300B and got the same result. There simply wasn't a huge difference
that could be detected by **anyone**.
Measurements of N&D were similar with both lots of amps.
>
> > There are people like Trevor Wilson who display they have a concrete
> > brain when it comes to SET amps, and you are the same with regard to
> > using multigrid tubes.
>
> Guilt as charged your honour! Can't be bothered replying to the rest of your
> post as I've now developed RSI from what I've typed so far. Where do you get
> the time to write these posts????
I don't watch much crap on TV or spend all day jerking of to porno. I
don't go to pubs, or have a wife and family to slave after.
Why watch other folks doing their lives when you could be doing your
own?
It doesn'y matter what it is i am involved with, I usually try to look
deep and think of 1,001 things, and use huge amounts of time to answer
1,001 questions about what I'm doing. I do all my own bicycle, car, and
house maintenance and cooking, and I only wish I had 480 hours in each
day to get more done.
This makes me one of the slowest amplifier makers in the universe. But
people actually do pay me for the results.
I've rarely ever suffered RSI, but just before I prepared the last
edition of my website in 2005/6, I did have symptoms of weird arm pains
after medium typing work and news group answering before I began
preparing the 17MB for posting. So I went to my accupuncturist, Douglas
Godfrey of Sargood St O'connor and had him use his magical laser on my
arm. Then he showed me which pressure points to press if I ever had
recurring pains and in two days I was fixed miraculously. So I ain't
afraid of RSI or chronic fatigue. You just have to know how to keep at
it. I ride a bicycle 200km a week. My brain didn't fall out, and my
health hasn't been better. Humans were designed to work and be fit
animals right into old age, and although many can't achieve this due to
101 problems like depression and rich folks diseases, they *should* keep
trying. But of course many don't, or can't, and they secumb to the
modern rot of getting unfit and fat and being slow minded.
In summer I also do about 2 kilometres in the pool each week. This
augments what I do on the bicycle, which is an unatural way to move even
though the cardio vascular benefits are second to none. I hate gyms
which are exceedingly unatural places to exercize, IMHO.
In the early 1990s I did have problems with chronic fatigue because in
addition to 200km a week or more on the bike I was doing 30 hours of
building work. A holiday at Byron Bay for a month fixed that. I've never
needed to take a holiday since, and holidays are expensive, and you
spend your time gorking at other folks gorking at you when you could be
at work being creative or spreading a message on the Net.
Patrick Turner.
Patrick Turner
The 10 tube preamp began life in about 1994 and was my first attempt at
using vacuum tubes since I had last mucked around with tubes as a yougan
in 1964.
It is **MY OWN Amplifier**, and NOT for sale.
Its been rebuilt 3 times to increase features and lower N&D and as a
test bed amp for future improvements.
The texta lettering **DOES NOT** reduce the sound quality which is
second to none.
For a look at what is un-saleable junk, using ancient iron wound parts
of dubious quality, all spread out messily over a vast area, go to
http://db.audioasylum.com/cgi/m.mpl?forum=tubediy&n=61053&highlight=pic+Doug+Flynn&r=
This kind of mess is very quick, cheap and easy to assemble, and is a
cat&dog killer if ever I saw one.
When you have done something finely crafted like the amps you see at my
website, let us all know.
I have my reasons to stay well clear of audioasylum.com
Patrick Turner.
Patrick Turner
I believe the best 845 being made 3 years ago was the KR845. Its
availablity in Oz then was via a wealthy Oz agent who owned a huge
trucking company, and who reckoned he may as well buy a distributorship
from KR buy agreeing to buy aud $50,000 worth of stock so that he could
then get some decent 845 amps etc at wholesale costs while being able to
sell off the surplus as a dealer. He sold me the KR845 at $375 each.
Since I bought those, this Oz dealer seems to have vanished; he had a
website about old fashioned tube audio at www.vares.com.au
but its been down for 2 years now and KR have given their
distributorship now to Duratone Hi-Fi in Canberra. They had a 30w+30w
SET stereo amp with solid state driver stage for $18,000, ( I was told
). http://www.kraudiousa.com/va350.php
KR Audio of Prague make whole amplifiers, not just the tubes.
It appears the little KR Audio company does not have its own website at
kraudio.com.cz as it used to, and seems to have become an exclusive
offshoot of the US based enterprize marketting KR gear.
So how does anyone justify the cost of some tubes such as the KRaudio
845, or their slightly better KRT100, similar but better to 845?
Well, if you think its good, you pays your money.
Spare a thought for those buying Halcro amps. These boast that THD at
200W at all F between 10Hz and 20kHz is less than 0.0001%.
Maybe they cost aud $50,000 per pair, and you'll need a preamp.
I know a guy who ditched his SE amps and bought a pair of Halcros. He's
a leading Sydney heart surgeon. The amps maybe cost him only a
fortnight's pay. Quite cheap to him. I could live off $50,000 for 5
years.
The Audio Society of Hong Kong were quoted as saying about the Halcro
amps "Ah, Halcro, it like 300B, but go louder".
I would say the folks at Halcro would have been appalled that their
brandname could be used in such a short meaningful sentence including a
reference to something designed in about 1927. Listening rooms of most
folks in HK are tiny by Oz standards, and asians tend to like sensitive
speakers so what the HK club inferred was that Halcro are about as
useful as tits on a bull.
Meanwhile, I have my doubts about KR audio tube longevity. They have
dull emitter DH cathodes like a 300B, only beefed up to get the required
emissions for the larger amount of tube current so that one 845 can do
about the same work as 2.5 x 300B, except that anode voltage is more
than twice that of that used with 300B.
Now the good old books about tube making say that once you have anode
voltages over about 900V you should have DH cathodes made using thorium
doped tungsten bright filaments which work at about twice the
temperature of the dull emmitters seen in 300B and other similar tubes.
The oxide coated cathodes of the dull emitter DH cathodes are prone to
cathode destruction by positive ion bombardment once the anode voltage
goes high enough to cause the phenomena of very high speeds of cathode
impact due to the high voltage. KR say that the Ea for 845 should be
between +900V to +1,100V for best operation and that's what I found. But
once a tube got a little gassy like they all do get over a time, then
those positive ions are going to cause cathode erosion a lot earlier
than what you'd see i the Chinese 845 and the original samples of 845 by
RCA and others which all had tungsten cathodes. Now it was exceedingly
rare that anyone ever used 845, or similar tubes in SET mode in times
prior to the SET fad which began in earnest in about 1992.
845 and 211 and GM70 and a lot of others were usually used for linear
amps in class AB1 or AB2 when they could make a massive amount of power
from just a pair in PP, up to 300W, so they got used as AF modulators
and stadium amps. They would last well with anode supply voltages of
+1,500V because the cathodes were tungsten and the idle dissipation was
kept quite low as possible, maybe only 20W. The old tungsten cathodes
need 10V x 3.3A or 33 watts just to make them heat up to work and then
with Pad = 20W, the heat produced at idle was only 53W per tube. But
since the SET fad began in 1992, ppl began to use these tubes at Ea of
1,000V and Ia at 75mA or more, ie, Pad = 75W.
An 845 with tungsten cathode was then being asked to run at 108W total
heat production. Bloody Hot, but still only like 3 x 40W lamps. But I
found Chinese tubes would run OK at such a high T. KR audio have these
dull emitter type cathodes needing 10V x only 1A of current to heat them
up to work, or only 10W, so that with Pad = 75W, the total heat in the
tube was 95W, a worthwhile reduction from 108W.
But I'd not be game to use KR audio tubes with a high voltage Ea in a PP
situation.
IMHO, the 211 is probably just as good as the 845 and can be ideally
used in SET mode with ea at +1250V with no worries of cathode erosion.
Such a high Ea minimises the need to drive into grid current to get a
useful amount of output power. And the grid drive voltage requirement is
much less tha 845 so the driver amp can be simpler.
In about 1995, the Japanese Audio Note Ongaku used 211 for 21 watts per
channel and a pair cost about $140,000.
The OPT had silver enameled wire and used 50% nickel and 50% GOSS in the
OPT cores. The circuit was a very cheap looking thing which was easily
improved upon.
I could never understand why anyone would pay so much for such a low
powered amplifier.
211 has a much higher Ra than the 845 so the damping factor obtainable
with 211 without any loop NFB may be poorer compared to 845. For 845
with Ea = 1,050V and Ia = 70mA, the load for 845 should be 12k. Ra is
2k2, so DF = 5.45, not too bad for something without any GNFB.
With 211, you need Ea = +1,250V, and Ia = 60mA, and RL = 20k, and with
Ra = 4k+ then DF = 5 approx and about as good as 845.
But any OPT made for carrying the DC and with a ratio for 12k or 20k to
5 ohms has to be an absolute work of art to keep leakage inductance low
and shunt capacitance low while having high primary inductance and
freedom from core saturation and to keep total winding losses less than
6%. To get the wide bandwidth between say -3dB at 10Hz to 30kHz without
reliance on GNFB takes real doing. Most old iron OPTs one might buy from
ex military apps or made before 1960 had hopeless bandwidth without the
use of GNFB to widen the BW to basic hi-fi standards.
Anyone could become quite predjudiced against using 845 or 211 if they
did not use very good iron wound components.
But the same can be said when making any SE amp; the iron quality has to
simply be there to get basic hi-fi standards.
many ancient old OPTs scrounged up from ex military sales places was of
hopeless specifications for hi-fi use; it was only made to work for a
basic industrial standard which meant bandwidth was rather low, and
saturation would occur at too high a frequency.
The iron quality used in many old OPTs I have come across is not GOSS,
but high loss un-oriented grain low Si content material which was the
cheapest available and this creates more distortions in the signal than
do the tubes used.
When I made amps with 845 last year, I used a pair to get 55W max and a
load of 6k:4.2 ohms, so the Rout without NFB became 0.77 ohms plus the
winding resistance, so DF = 5 without any GNFB. I added some 8dB of GNFB
to reduce Rout to give DF = 10.
OPT was a 72mm stack of 50toungue GOSS, interleaved 4P x 5S sections,
then potted. No need for parafeed. The absense of parafeed meant
stability with NFB was excellent because the extra time constants at LF
and shunt C at HF did not cause excessive open loop phase shifts at each
end of the band.
At a watt or two of average power, THD/IMD was utterly negligible, and
sound was good. Noise at the output with no signal is less than 0.25mV.
http://www.turneraudio.com.au/monobloc845se55.html
Patrick Turner.
Doug Flynn
>
>> The labels are written in *texta*, and wonky texta at that. At least I
>> use
>> stick-on labels. Any my amps are much more user-friendly. Here's a pic to
>> prove it :-)
>
> second to none.
It depends on the colour of the texta. I do believe that green texta will
enhance the sound quality, but red will degrade it. I think we need to do a
double-blind test on this.
> For a look at what is un-saleable junk, using ancient iron wound parts
> of dubious quality, all spread out messily over a vast area, go to
> http://db.audioasylum.com/cgi/m.mpl?forum=tubediy&n=61053&highlight=pic+Doug+Flynn&r=
>
> This kind of mess is very quick, cheap and easy to assemble, and is a
> cat&dog killer if ever I saw one.
It also killed several DMMs I stupidly used to measure 1kV B+ voltages!
Doug Flynn
> Years and years of test rigs, and forever changing mind about which caps
> sounded better.....
Yes, I considered it an apprenticeship. I now know what sounds good and what
doesn't. And crap-arse caps don't sound good. Never did, never will.
> Most of my ideas about how to go about this grate with all the low
> priests I meet in your religious cult. I quite like NFB, using large
> electrolytic caps, and bypaseed Rks, and IDH tubes 6550 which IMHO are
> just as linear or more linear than a 300B in triode.
NFB = evil
Electros = evil
Bypassed Rks = evil (unless quality film caps used)
IDH tubes = evil
6550 more linear than a 300B = utter arse-dribble
> I just DO NOT know anyone who doesn't like the sound I get with 12AU7 in
> preamps or anywhere else. It could have been another 6SN7, or a trioded
> pentode, or just about anything else with LOW GAIN and low Ra, but the
> 12AU7 is and ideal choice with 6SN7 so that the amp isn't too sensitive.
> There is some GNFB in those amps, but only to get the Rout low enough
I don't dispute that it sounds good, but the AU7 is a shitty, high
distortion tube. The SN7 is vastly better, just as an SL7 is vastly better
than a AX7.
> I have been able to compare the re-engineered samples of amps with DH
> tubes I've worked on with those I've made with IDH triodes or
> multigrids. The amps with 300B were never any better than amps of the
> same power using half the number of multigrids.
Maybe you need better speakers.....
keithr
>
> NFB = evil
> Electros = evil
> Bypassed Rks = evil (unless quality film caps used)
> IDH tubes = evil
> 6550 more linear than a 300B = utter arse-dribble
There are two possibilities here:-
a) You are a troll.
b) You are an idiot.
In either case you should be ignored.
Patrick Turner
Doug Flynn wrote:
>
> "Patrick Turner" <in...@turneraudio.com.au> wrote
>
> > Years and years of test rigs, and forever changing mind about which caps
> > sounded better.....
>
> Yes, I considered it an apprenticeship. I now know what sounds good and what
> doesn't. And crap-arse caps don't sound good. Never did, never will.
That's what they all say.
The caps get the blame for all the ills in the circuit that lay
unnoticed by the apprentices.
>
> > Most of my ideas about how to go about this grate with all the low
> > priests I meet in your religious cult. I quite like NFB, using large
> > electrolytic caps, and bypaseed Rks, and IDH tubes 6550 which IMHO are
> > just as linear or more linear than a 300B in triode.
>
> NFB = evil
> Electros = evil
> Bypassed Rks = evil (unless quality film caps used)
> IDH tubes = evil
> 6550 more linear than a 300B = utter arse-dribble
I don't expect you to change your absurd one eyed view on audio amps.
But I do want the group to know that you are one eyed.
And inconsistent. For example, when you leave an Rk un-bypassed, there
is series current FB. Its not always good for the sound because such a
low amount of NFB does not reduce distortion much, but it sure does
increase effective Ra and thus reduce the bandwidth of the amp.
The current FB isn't bad in most cases because the open loop THD is
negligible anyway, and the application of a low amount of NFB is not
then going to make the mix of harmonic spectra much more complex or
audible. The current NFB thus generally reduces the major 2H a little
and thus the major effect dominates.
But you are grossly stupid and dull minded when you can come so such an
overly simplistic list of 5 generalities.
I could write a book to refute your utter BS but you ain't worth my time
Doug.
>
> > I just DO NOT know anyone who doesn't like the sound I get with 12AU7 in
> > preamps or anywhere else. It could have been another 6SN7, or a trioded
> > pentode, or just about anything else with LOW GAIN and low Ra, but the
> > 12AU7 is and ideal choice with 6SN7 so that the amp isn't too sensitive.
> > There is some GNFB in those amps, but only to get the Rout low enough
>
> I don't dispute that it sounds good, but the AU7 is a shitty, high
> distortion tube. The SN7 is vastly better, just as an SL7 is vastly better
> than a AX7.
Again, you are being absurd, and showing what a fucking idiot you are.
The differences in THD at a typical 2Vrms level between the tubes you
nominate and when set up optimally is nothing like vast as you suggest.
None of the tubes you nominate are high THD when used in input driver
stages properly.
All can sound marvellous.
All can be made to make less than 4% THD at 50Vrms of output with all R
loading and R used for dc supply.
When the dc supply is via a CCS of choke with an R to isolate the shunt
L and C or when the tubes are used in a �-follower combo the THD can
become **much** lower. A typical THD result in a preamp is then 0.2% for
10Vrms, and this does not qualify for high THD especially if you have
the gain pot before the gain stage and the tube output is only a typical
0.1Vrms to get average levels in a following power amp.
Then THD becomes < 0.005%, so why do you so idiotically say that the
12AU7 "is a shitty, high distortion tube." ????
The sound of what is used depends on how it is used Doug. You simply
dunno what your're talking about.
And a whole pile of people have come to different conclusions than you
have.
They are all more open minded and apparently equipped with better
hearing.
There are *so many* who'd say the opposite is true to what you say.
>
> > I have been able to compare the re-engineered samples of amps with DH
> > tubes I've worked on with those I've made with IDH triodes or
> > multigrids. The amps with 300B were never any better than amps of the
> > same power using half the number of multigrids.
>
> Maybe you need better speakers.....
I don't need better speakers. Nor do I want any.
I have never heard any better or measured any better than what I have
made myself.
There are a few others I know who have tried valiantly to make their own
and they have not convinced me I have to follow them.
And the ones who have laboured so hard and long to make horn speakers
merely illustrated to me how easy it is to build expensive noisy
firewood.
There could well be better speakers out there somewhere but after
hearing so many other people's speakers and repairing and re-engineering
so many clients speakers I can safely predict that its unlikely I'll
ever hear better speakers.
There are speakers which might be better, such as Wilson-Watt puppies
and so on thet people rave about but I don't know anyone with such
things.
Meanwhile I've heard many speakers which did seem good enough and were
"drinkable wine"
I have also enjoyed listening to systems which were marvels and which
used solid state amps and Martin Logan ESLs.
Don't ever be so stupid as to believe that your ideas sum up to being
the only valid way to a good listening experience.
When you ram your narrow ideas down everyone's neck here you only seem
like a fool with a broom handle.
Patrick Turner.
Patrick Turner
Both a and b are true.
Patrick Turner
Doug Flynn
>> Doug Flynn wrote:
>>
>> NFB = evil
>> Electros = evil
>> Bypassed Rks = evil (unless quality film caps used)
>> IDH tubes = evil
>> 6550 more linear than a 300B = utter arse-dribble
>
> I don't expect you to change your absurd one eyed view on audio amps.
Guilty as charged, yer honour. My one-eyedness comes from years of
experience. I am only one-eyed against things that suck, and which I have
proven through experimentation to suck.
> And inconsistent. For example, when you leave an Rk un-bypassed, there
> is series current FB. Its not always good for the sound because such a
> low amount of NFB does not reduce distortion much, but it sure does
> increase effective Ra and thus reduce the bandwidth of the amp.
> The current FB isn't bad in most cases because the open loop THD is
> negligible anyway, and the application of a low amount of NFB is not
> then going to make the mix of harmonic spectra much more complex or
> audible. The current NFB thus generally reduces the major 2H a little
> and thus the major effect dominates.
I realise that an unbypassed Rk suffers FB. I always bypass (with a quality
film cap, of course), unless I am using fixed bias or battery bias. Using an
electro as an Rk bypass *will* destroy the sound as sure as the sun comes up
in the morning. If you use these evil devices ANYWHERE in the signal path
then it's hardly surprising you can't tell the difference between a 300B and
something you found in the back of an old TV.
> But you are grossly stupid and dull minded when you can come so such an
> overly simplistic list of 5 generalities.
Or perhaps I have just reached a Zen-like state of consciousness :-) Are you
*seriously* telling me a 6550 is MORE linear than a 300B???? I wonder if the
curves of a 300B versus a trioded 6550 bear that out.
>> I don't dispute that it sounds good, but the AU7 is a shitty, high
>> distortion tube. The SN7 is vastly better, just as an SL7 is vastly
>> better
>> than a AX7.
>
> The differences in THD at a typical 2Vrms level between the tubes you
> nominate and when set up optimally is nothing like vast as you suggest.
> None of the tubes you nominate are high THD when used in input driver
> stages properly.
> All can sound marvellous.
> All can be made to make less than 4% THD at 50Vrms of output with all R
> loading and R used for dc supply.
> When the dc supply is via a CCS of choke with an R to isolate the shunt
> L and C or when the tubes are used in a ต-follower combo the THD can
> become **much** lower. A typical THD result in a preamp is then 0.2% for
> 10Vrms, and this does not qualify for high THD especially if you have
> the gain pot before the gain stage and the tube output is only a typical
> 0.1Vrms to get average levels in a following power amp.
> Then THD becomes < 0.005%, so why do you so idiotically say that the
> 12AU7 "is a shitty, high distortion tube." ????
All 9 pin tubes suffer heaps more distortion than octals. The tube makers
never tried to hide this fact.
roughplanet
Doug Flynn wrote:
> L and C or when the tubes are used in a µ-follower combo the THD can
Now who's being one-eyed Patrick. There is a guy down here in Mexico who's
name is Wally Roche and who builds the most amazingly simple (but
beautifully
designed) horn speakers all of which sound truly superb.
He also builds ribbon speakers to die for, but only at the rate of a pair
every 3 years
or so because of the time taken (his attention to detail & tuning is
exemplarary).
The top of the line hoen speakers - the 'Full Monties' sound so good I have
often
been temted to swap my beloved L110's for a set. I'd love your opinion on
them.
I'm sure you'd have to eat a generous serving of humble pie :-).
> There could well be better speakers out there somewhere but after
> hearing so many other people's speakers and repairing and re-engineering
> so many clients speakers I can safely predict that its unlikely I'll
> ever hear better speakers.
> There are speakers which might be better, such as Wilson-Watt puppies
> and so on thet people rave about but I don't know anyone with such
> things.
I use to own a pair of Wilson Watt Puppies & they were indeed great
speakers.
But when I had to return the Wolcott 220's, I didn't have another amp with
the
necessary grunt to drive them so off they went too.
> Meanwhile I've heard many speakers which did seem good enough and were
> "drinkable wine"
> I have also enjoyed listening to systems which were marvels and which
> used solid state amps and Martin Logan ESLs.
>
> Don't ever be so stupid as to believe that your ideas sum up to being
> the only valid way to a good listening experience.
>
> When you ram your narrow ideas down everyone's neck here you only seem
> like a fool with a broom handle.
Hear hear.
ruff
Patrick Turner
Doug Flynn wrote:
>
> "Patrick Turner" <in...@turneraudio.com.au> wrote
>
> >> Doug Flynn wrote:
> >>
> >> NFB = evil
> >> Electros = evil
> >> Bypassed Rks = evil (unless quality film caps used)
> >> IDH tubes = evil
> >> 6550 more linear than a 300B = utter arse-dribble
> >
> > I don't expect you to change your absurd one eyed view on audio amps.
>
> Guilty as charged, yer honour. My one-eyedness comes from years of
> experience. I am only one-eyed against things that suck, and which I have
> proven through experimentation to suck.
You now have the credibility of a gibbering idiot.
>
> > And inconsistent. For example, when you leave an Rk un-bypassed, there
> > is series current FB. Its not always good for the sound because such a
> > low amount of NFB does not reduce distortion much, but it sure does
> > increase effective Ra and thus reduce the bandwidth of the amp.
> > The current FB isn't bad in most cases because the open loop THD is
> > negligible anyway, and the application of a low amount of NFB is not
> > then going to make the mix of harmonic spectra much more complex or
> > audible. The current NFB thus generally reduces the major 2H a little
> > and thus the major effect dominates.
>
> I realise that an unbypassed Rk suffers FB. I always bypass (with a quality
> film cap, of course), unless I am using fixed bias or battery bias. Using an
> electro as an Rk bypass *will* destroy the sound as sure as the sun comes up
> in the morning. If you use these evil devices ANYWHERE in the signal path
> then it's hardly surprising you can't tell the difference between a 300B and
> something you found in the back of an old TV.
>
> > But you are grossly stupid and dull minded when you can come so such an
> > overly simplistic list of 5 generalities.
>
> Or perhaps I have just reached a Zen-like state of consciousness :-) Are you
> *seriously* telling me a 6550 is MORE linear than a 300B???? I wonder if the
> curves of a 300B versus a trioded 6550 bear that out.
I suggest you perform some disciplined testing of 6550 and 300B tubes to
find out the truth about linearity.
I did.
See some mentions about test results and curves I plotted at
http://turneraudio.com.au/loadmatch1-se-triodes.htm
I told you that unless you do some real work to find out the technical
attributes of tubes, then you will remain a stupid bloody idiot shouting
BS to everyone who is forced to read all your painfully ill-informed BS.
What you may find is that a 6550 may only be slightly marginally
different to a 300B with regard to distortion products.
When loaded with the load giving maximal power for a given Ea/Ia idle
condition, most triodes give between 4% and 6% of THD at the clipping
point and most of that is 2H.
So if you have a 6550 making 7W into 8 ohms, you have 7.5Vrms of signal
voltage and between 4% and 6% will be THD, and mainly 2H.
Ditto 300B.
Now to ensure fidelity with such an amp the average signal voltage
should never rise above 0.75Vrms, which allows for peaks up to 7.5Vrms.
Triode THD varies nearly in proportion to output load voltage, so let us
say you have 5% THD at 7.5V, then expect 0.5% at 0.75Vrms load voltage.
This is a considerable amount of THD, and 12dB of global NFB will reduce
this to about 0.125%, which is below the threshold of audibility as
specified in the 1955 Radiotron Designer's Handbook.
The 0.75Vrms average voltage mentioned gives 0.07W average into 8 ohms
and if the sensitivity is 89dB/W/M then the av SPL = 78dB approximately,
and this is a level which is a bit quiet for many people.
Contrary to what you and many flat earth believe about NFB, the applied
global NFB can make the sound more dynamic, precise, detailed and with
less bass blur and with more preserved warmth.
There are ways to reduce Triode THD to blameless levels in the above
sample without NFB.
(1) Increase the sensitivity of speakers from say 89dB to 95 dB and then
you only need 1/4 of the power for the same SPL, or half the applied
voltage but THD will still be 0.25% average. Raising speaker sensitivity
to 99dB by using horn loaded speakers means for 78dB av level you only
need 0.007W and 0.75V / 3.16 = 0.23V applied so THD = approx 0.23V x 5%
/ 7.5V = 0.153%.
Even with horn speakers, THD isn't negligible. But most people won't
hear the THD and the larger amount of resultant un-harmonic IMD which is
the real crud that ruins the music.
(2)Use more paralleled tubes or bigger single tubes to raise the maximum
power. Say we have 4 x 300B for 28W max.
This gives us 15Vrms into 8 ohms at 5%, so with 0.75V with ordinary
speakers the THD = 0.25%, or half what we get with one 300B.
Gee, we have to spend real bucks to get the THD down this way.
(3)Use two parallel tubes but change the loading so the pair produce
only 7W. This means tou need an OPT with a turn ratio 1.41 times greater
and it results in halving the output resistance, and perhaps halving the
THD/IMD at the same powers raised by just using one tube with the
nominal load equal to that giving maximum power.
(4)Use 4 output tubes anyway, but to save money for excessively and
pretentiously priced 300B, use 6550, KT88, EL34, 6L6, 5881, KT90 or
whatever other tube takes our fancy which can be connected as a triode.
The arrange the OPT to have several taps where 5 ohms is the load for
say 28W, and then we have taps for 28W at 3.5 ohms and 2.5 ohms.
When you connect the 8 ohms to the 2.5 ohm tap the power max will not be
huge but still more than with 1 tube and and the THD will thus become
lower and the amp will definately measure as well as the single tube
with NFB.
This will in fact suit most real hi-fi listeners who unlike their
teenage kids won't ever need more than an average SPL of 78dB using
speakers they just bought at Miranda Hi-Fi or Duratone.
The guy with the one lone 300B, ( or 6550 etc ) may struggle with
generic shop sourced speakers of average sensitivity. He should hear a
betterment with NFB connected but many don't like that and won't use the
NFB. I have many clients who prefere NFB connected. No apologies Doug,
but your ideas about NFB are nowhere near being likely to ever be
universally accepted.
(5). Use PP connection and class A. This means 2 x 300B, and instead of
7W we could have 14W in pure class A and if the load is adjusted so each
tube sees the same class A load as in the SE case then the THD will fall
dramatically from 5% to about 1% at clipping but it will be nearly all
3H instead of 2H. The IMD products produced by the 3H, ie, the slight
flattening of wave peaks, ie, compression, are less harmonious than
those produced by the large amount of IMD as a result of 2H being the
main THD.
So despite reducing THD at all levels by -12dB just by using PP, the
betterment is only about 1/2.
Still, because we have two output tubes instead of one the betterment
goes back up to being very reasonable if all we want is 0.75vrms into 8
ohms av where the THD will measure 0.07% without any GNFB.
Less than 0.1% is considered a good result, and nobody I know complains
about it.
If you load the PP amp with a load higher, or have a higher OPT turn
ratio you will further reduce the THD but not by much because even
without any loading the voltage output signal will still have some THD
with either SE or PP because the constant current loading of the tubes
displays such distortions when you measure it. PP triodes with CCS
loads, ie, no load resistance are remarkably linear, and the THD one
does measure could then be mainly a product of the iron caused
distortions.
We have to assume we are driving all these examples with distortionless
input and driver stages.
It is common that DIYers without much test gear or experience or
knowledge will set up their driver amps appallingly so the driver stages
will produce as much THD as the output stages.
In the case of an SET, the driver tube can also make 5% THD 2H when the
output tube is clipping but its not all bad because this THD will cancel
the THD made in the output stage. The THD of two consecutive triode
stages are self canceling, but not the resultant IMD which is what
spoils the sound. THD is never a problem on its own; its always the IMD.
So if you linearize the driver stage, then you get less cancellations
but less IMD; the better solution is to minimise THD everywhere, period,
and not depend much on THD cancellings.
In the case of PP amps the SE input stage generally makes negligible
amounts of THD and its the driver stage that performs worst. it is
usually a LTP which can make 1% THD when the OP stage clips, and this is
all 3H and with a phase that is nearly always additive so a typical
class A triode amp makes 2% THD at clipping instead of only 1% without
GNFB.
There are a range of circuits published at my website which readers are
welcome to try out and all give excellent sound.
A few have triode output stages, but most have multigrid tubes connected
with less simplicity.
There are a number of ways to overcome the serious shortcomings of low
power class A single ended triode amps if you want more level than a
mere 78dB av level. Just going to 81dB requires twice the power, ie, 2 x
300B, and to 84dB av when women will begin to complain means you need 4
x 300B.
With average modern dynamics or with Quad ESL57, 4 x 300B in parallel
for 28W max into 8 ohms woulod be a minimum amp IMHO.
I have heard some nice horn speakers powered by SE amps as well as a
pair of Tannoy dual concentrics in 7cu.ft boxes.
The Tannoys boasted 96dB efficiency, and a lone 300b sounded just fine
to me.
The owner had a bother with hum though which took some solving.
My own tubed kitchen radio I designed and built in 1999 has a lone EL34
in triode with 12dB GNFB to a paralleled 12AX7 input triode. I have h a
carefully re-enginnered speaker cabinet with full reflexing with old
Rola Deluxe 12" from 1953 and a dome tweeter from 1974. Its magical
sound from such simple circuitry and because I listen with the speaker
beside me I only need low levels. The radio has 20Hz to 10kHz of audio
bandwidth so music from Radio National sounds excellent.
I just cannot see that it would be better to use a 300B. A waste of
expense IMHO.
I recently had a 1993 Metz TV set given to me for services rendered. Its
all SS digital but has a CR screen, very nice. Its sound is lamentable
and I must get around to using external speakers. I'd say there probably
are generic SS chips used for the audio, and frankly, I cannot see why
switching to DHT SE amps are going to make any difference once I achieve
the 90% betterment with hi-fi speakers.
>
> >> I don't dispute that it sounds good, but the AU7 is a shitty, high
> >> distortion tube. The SN7 is vastly better, just as an SL7 is vastly
> >> better
> >> than a AX7.
> >
> > The differences in THD at a typical 2Vrms level between the tubes you
> > nominate and when set up optimally is nothing like vast as you suggest.
> > None of the tubes you nominate are high THD when used in input driver
> > stages properly.
> > All can sound marvellous.
> > All can be made to make less than 4% THD at 50Vrms of output with all R
> > loading and R used for dc supply.
> > When the dc supply is via a CCS of choke with an R to isolate the shunt
> > L and C or when the tubes are used in a ľ-follower combo the THD can
> > become **much** lower. A typical THD result in a preamp is then 0.2% for
> > 10Vrms, and this does not qualify for high THD especially if you have
> > the gain pot before the gain stage and the tube output is only a typical
> > 0.1Vrms to get average levels in a following power amp.
> > Then THD becomes < 0.005%, so why do you so idiotically say that the
> > 12AU7 "is a shitty, high distortion tube." ????
>
> All 9 pin tubes suffer heaps more distortion than octals. The tube makers
> never tried to hide this fact.
Wrong again. Many samples of 6CG7 had **exactly the same electrodes** as
used in 6SN7. The makers just crammed them in closer and sometimes
inserted a screen between the two anodes then lowered the Pda rating
slightly. This was done to reduce the tube size and hence the chassis
size and cost of much gear made. The 6CG7 was commonly used in TV sets
which had a large number of tubes and thus were prone to cost cutting.
6CG7 and its cousin, the 6FQ7 were made in vast numbers and using
different shapes and sizes of electrodes. All had ľ = 20 approximately,
Ra = 8k at Ia of 10mA approx, so gm = 2.5mA/V. People will say the many
brands of 6SN7 and 6CG7 may vary in sound quality.
All of them will argue which is best.
So wash your mouth out with capacitor oil please.
I know folks who say the German Siemans 6CG7 NOS were the best sounding
signal triodes ever made. But I like NOS Miniwatts made in Oz.
There are a range of triodes with either octal or nine pin bases or
other older pin outs which have similar characeristics and which all
have good linearity.
Some nine pin tubes made were more non linear but they often produced
more gain because the gm was higher.
If you look at the original data curves for old triodes ( and curves
were drawn with gear which often had THD greater than the tubes ), then
there wasn't much wrong with many samples of nine pinners, octal tubes
or the 4 pinners from really old times such as the 1920 101D used as
voltage regs in Melody amps now. The 101 like many old designs had boxy
anode shapes and multiple strands of wire for their filament which
affected the theoretical performance but basically, many old triodes
were very linear when you examine the spacings of the Ra lines across
any constant value of Ia. I've measured the THD when I have loaded
octals and nine pins using a CCS anode load and usually I have gained
less THD than the curves indicate, and that's the case even with 6550,
EL34, 807 etc.
In PP, pair of EL84 in triode used as a driver in my LTP design produces
less than 0.1% THD at 75Vrms from each anode.
The SET driver for my 845 amps uses EL84 in triode and in parallel and
is remarkably linear up to the required 106Vrms drive needed to drive
the 845. To try to use DH triodes would be a royal PITA because they
have far less gain and I'd need an extra stage in the amp and then I'd
have all the problems of microphony and filament noise. In the old times
when battery power was often used they woul never use resistances to get
dc to a triode anode; choke or transformer coupling was universal. But
the chokes and transformers used in 1932 were **PHARKING AWFUL** and
contributed serious THD and bandwidth restrictions. The tubes always
were the least blameworthy.
Once people kept RL > 4Ra across a wide BW then triode circuits became
very linear and the triodes were never the cause of poor sound.
With slightly more modern 9 pinners that Doug hates with a vengeance
such as 12AX7 and 12AU7, and I could add 12AT7, the use of RL > 20Ra
makes them sound glories because THD and the IMD has shrunk to
negligible levels. As RL ohms are raised, the gain of the tube
approaches ľ and the application of internal NFB triode NFB becomes
maximised. What would Doug have us do? chop open the triode and rip out
the NFB? It'd be like chopping open the hapless human sacrifice on an
ancient Peruvian alter and removing the heart to appease the Gods.
Lemme tell ya, the God Of Triodes intended triodes to have their inbuilt
natural NFB and I for one rely on what this GOT done to make amps sound
so blameless. However, a few blokes after about 1930 did manage to
remove the heart of some triodes without upsetting too many gods by
inserting a second and third control element within a triode and you had
a pentode, with control grid like a triode's and a screen and a
supressor grid.
This allowed tubes to amplifly RF signals which was difficult to do with
triodes because of their NFB and Miller C.
The screen interupted the triode's anode field effects on Ia. So to get
the same linearity with a pentode you simply connected the screen to the
anode, or you used external loop NFB. The supressor had a minor effect
to merely supress secondary anode emisions but that's another story.
Soon after pentodes cae beam tetrodes, the most famous being the KT66
and 6L6 and 807. All make fabulously linear triodes when triode
connected.
Tubes like 6EJ7 when triode connected *are* less linear than say
6CG7/6SN7, but they have a gain of maybe 50 instead of 18.
When used in an LTP though the slight increase in THD is negligible
compared to what is made by the OP stage.
Pentodes such as 6EJ7 can be set up as a pentode with direct coupled
cathode follower for an open loop gain of 200+ and a simple loop of
shunt FB used to reduce the gain to that of a 6CG7, and THD will be less
than the CG7, and Rout MUCH less.
Depending on how pentodes are used you can get extremely fine sound from
them.
Many of the greatest ancient vinyl recordings which Doug treasures so
much were recorded using gear riddled with nine pin pentodes like the
EF86, or seven pin pentodes like the 6AU6, or piles of 12AX7.
The subjective quality of the sound you hear with SE triodes isn't
always a result of the THD of the tubes used. Its the total of all the
circuit elements. And certainly the microphony has a large part to play.
When testing tube circuirs with a dummy load there is no sound to upset
the tubes, but in the real world the sound does cause minor levels of
accoustic FB and if measured carefully the artifacts based around
resonant structures within the tubes could easily be way above the
THD&IMD, and because microphony excites electrode resonanaces commonly
between 500Hz and 2kHz, then the sound may be coloured and depending on
the key of the music. An natural tube renonance which is a C note will
enhance the sound of music played in the key of C.
If for example NFB is used around any line stage amp to reduce its open
loop gain from say 50 to 5, then the effects of microphony are reduced
by 10 times, and you'd not think you had a micro tube in there. I can
only say the sensible thing is to ensure tubes are not mircophonic and
not depend on corrective measures to remove sonic defects.
So try to use non-microphonic tubes, especially in input stages. Non
micro tubes are essential in input stages in phono and microphone amps
especially where NFB is used for RIAA eq. So called NOS triodes which
one finds to be microphonic are often old tired used tubes which still
look OK after doing 10,000 hours in somebody's preamp. They are
carefully returned to the old original boxes and sold on E-bay. The same
people would sell their grandma if the price was right.
300B tube in output stages can be pinged with a finger to produce an
audible donging sound from the tube, and often audible sound at a
speaker. The cathode is a filament of fine metal wire strung between
tensioning springs, and it all dongs musically if the tube is touched.
The slightest vibration can cause tube noises higher than any measured
THD/IMD or spurious PS noise.
With 6550, and other IDH triodes the same microphonic problem simply
does not exist.
Its a strange thing, but one NEVER hears anyone say anything about
picture quality of old TVs. But the more linear the circuits became with
applied NFB where required, the better the picture and the sound.
Meanwhile digital amps with PWM have some maturing to do and many of
them have passed the quality of generic class AB SS amps.
It isn't too dufficult to get a triode amp to provide excellent clarity
and preserved warmth and musicality. These halmarks of good amp systems
should not be manufactured by the amps and should reside within the
recording signal from vinyl or CD.
As the future develops and the past old master tapes eventually thrown
away because they have perished then we are left with modernity and all
its ills of post recording processing which ruins music so badly that no
matter what Doug says there is no possibility of fidelity.
A really decent speaker and amp system should have the audiophile
swearing at the music makers and never at his triodes, even if they are
pentodes or beam tubes strapped in triode mode to maximize the internal
NFB within the tiode.
Patrick Turner.
Patrick Turner
> > L and C or when the tubes are used in a ľ-follower combo the THD can
I didn't say all horns suck.
Indeed there must be some which sound OK.
None of the local diyers have succeded here. On eguy has a collection
of old professional Altec and JBL horns with compression drivers and
they have been the best I've heard when properly set up to measure well
by adjusting the active crossover between preamp and the 3 power amps
for each channel.
But the guy doesn't stick to the settings for the system and adds things
and alters things and its basically a sound I don't like.
he like high treble and low bass.
> The top of the line hoen speakers - the 'Full Monties' sound so good I have
> often
> been temted to swap my beloved L110's for a set. I'd love your opinion on
> them.
> I'm sure you'd have to eat a generous serving of humble pie :-).
I never judge a speaker until I've heard it, so I think the pie will
remain uneaten.
>
> > There could well be better speakers out there somewhere but after
> > hearing so many other people's speakers and repairing and re-engineering
> > so many clients speakers I can safely predict that its unlikely I'll
> > ever hear better speakers.
> > There are speakers which might be better, such as Wilson-Watt puppies
> > and so on thet people rave about but I don't know anyone with such
> > things.
>
> I use to own a pair of Wilson Watt Puppies & they were indeed great
> speakers.
> But when I had to return the Wolcott 220's, I didn't have another amp with
> the
> necessary grunt to drive them so off they went too.
!
>
> > Meanwhile I've heard many speakers which did seem good enough and were
> > "drinkable wine"
> > I have also enjoyed listening to systems which were marvels and which
> > used solid state amps and Martin Logan ESLs.
> >
> > Don't ever be so stupid as to believe that your ideas sum up to being
> > the only valid way to a good listening experience.
> >
> > When you ram your narrow ideas down everyone's neck here you only seem
> > like a fool with a broom handle.
>
> Hear hear.
!
Patrick Turner.
>
> ruff