patrick turner tube preamp
Am1d1c
dynamics ,good freq. response and very open. sounds good on reg. power supply
also on clc supply. I was not a fan of the 12ax7 but in this preamp it sounds
great good job patrick regards doug
Patrick Turner
Am1d1c wrote:
Well, I am amazed that someone might have tried this design of mine.I am glad you
like the outcome.
In case you were not aware,
I have a regulated power supply in mine.
The B+ is passed through a transistor emitter follower buffer,
using an MJ1002, but any high voltage darlington pair will do.
There is then about 100 uF at the output of the regulator.
Then there are still seperate R plus 47uF caps for each stage to work from,
after the regulator, and its cap.
Motorboating and cross talk is thus avoided, using one good psu.
I found there was no need for any L in the psu becuase the regulator
with a series current limiting R is far more effective than an L.
It is easier to use discrete hi voltage transistors
than a floating regulator, such as LM317.
And of course, with proper protection diodes, it is more reliable.
The heater supply is DC, with an LM350 used, and the whole DC
supply is floating at 72 volts, to avoid exceeding the heater to cathode
voltages going over ratings of 90 volts.
I have a steel box for the PSU, and the power tranny is an ex navy
potted type, with a steel or mu metal can.
There is a seperate heater tranny that I bought locally, and it
resides in a steel can, with sand all round it.
The 12 AX7 is a much maligned tube, but I find it quite blameless,
and better than most others in the phono section, where
it is most suited, as signal levels are so low, and easily handled.
One has to sort through different brands, for the quietest,
and most musical sounding ones.
Patrick Turner.
jim
"Patrick Turner" <in...@turneraudio.com.au> wrote in message
news:3C4658C0...@turneraudio.com.au...
Hiya Patrick
Any comments on using 6SL7s in your preamp instead of ECC83/12AX7s
Appreciate interelectrode capacitances are higher, and 6SL7s are lower mu,
but I've got about a dozen new Brimar 6SL7s lying about me with no obvious
home to go to.
Kind regards
jim
Patrick Turner
jim wrote:
I have nearly always found the big clunky 6SL7a bit more noisy than the 12AX7.
If you are lucky, this won't be the case,
and the mu follower input stage on my preamp
should work quite well, and there is plenty of gain.
Certainly, if you include the unity gain tone control stage,
the 6SL7 would be fine, as nearly any twin triode could be used
for this, and the NFB will keep noise low,
and overcome all capacitance problems.
The 6SL7 falls between the 12AT7 and 12AX7
in terms of electronic character.
Probably another better tube may be the rarer
12AY7, for the front end. I have never used that tube,
and I doubt it is still being made, but I could be wrong.
My SEUL amps have ECC35, which is very similar to 6SL7
at the front end, with both halves paralleled.
It is a nice tube in this application, and looks right, on the chassis.
Patrick Turner.
Jon Yaeger
What's your opinion of the 6EU7 vs. the 12AX7?
Thanks.
Jon
R.Pit
I found a Riaa amp with only 6SL7 .
Take a look at :
http://dddac.de (it's in English)
http://dddac.de/files/DDPRE2.pdf
A very intresting site anyway , good story's and projects.
Ronald
"jim" <j...@macsys50.freeserve.co.uk> schreef in bericht
news:a26an8$aj9$1...@newsg1.svr.pol.co.uk...
Chris Merren
> Any comments on using 6SL7s in your preamp instead of ECC83/12AX7s
> Appreciate interelectrode capacitances are higher, and 6SL7s are lower mu,
> but I've got about a dozen new Brimar 6SL7s lying about me with no obvious
> home to go to.
> Kind regards
> jim
Hi Jim....
Check this schematic out for RIAA using 6SL7....
http://groups.yahoo.com/group/tubetechnical/files/Schematics/PHONO-PRE.gif
Regards
Chris
ee
Where is the schematic available?
How do you like the 6922 or better yet the 6H30 for a preamp Patrick?
EE
Henry Pasternack
news:ZHC18.35049$UW.43...@zwoll1.home.nl...
> Hi,
>
> I found a Riaa amp with only 6SL7 .
> Take a look at :
> http://dddac.de (it's in English)
> http://dddac.de/files/DDPRE2.pdf
> A very intresting site anyway , good story's and projects.
>
> Ronald
The focus of the site seems to be the DAC. It's kind of sad, really,
because although the project is interesting, he's really broken all
the rules for constructing high-speed circuits with his perf-board
wiring and strap on "ground plane". Ouch! If his DAC sounds good
now, imagine how much better it would be if it were laid out properly
on a quiet multi-layered PC board with surface mount bypass caps and
an intelligently sectioned ground plane.
Anyone interested in state-of-the-art DAC construction owes it to
himself to do a little research. The difference, I think, between
an adequate DAC and a superb one lies in the construction. It's a
lot more complicated than it seems. For some informative but often
confusing reading, check out the late-nineties postings by Pete
Goudreau on rec.audio.high-end concerning his mods of the Parts
Connection DAC-1.
-Henry
Patrick Turner
Jon Yaeger wrote:
> Patrick,
>
> What's your opinion of the 6EU7 vs. the 12AX7?
I have never used the 6EU7, so I guess I can't have an opinion.If I had met some
along the way, I may have tried to use them.
But I am a little commercially minded, and so I am a bit reluctant
to make amps for which there are no tubes being made.
Its like the 12AY7, they say it is a regal triode, like the 6SN7,
but is it still made?
There were so many tubes made, and life is short,
and to try them all takes time....
Patrick Turner.
Patrick Turner
ee wrote:
> Good job.
> Where is the schematic available?
At my website, http://www.turneraudio.com.au ,
in the 'circuits-philosophy-images' page,
see 'april 2000 preamp'
Go down the page for the notes about it.
A pic is in the 'preamp' page, listed as 'prototype preamp'
There are quite a few different types of topologies used in this
preamp. Horses for courses.
Seldom do I use the tone control, and line gain stage,
but when I do, I can't hear any change when I switch them
in or out.
With LP replay, I have just the phono section, and balance /
volume controls, and a cathode follower in operation,
and usually there is enough steam to power most poweramps,
if I want it loud with an insensitive amp, I switch in the line stage.
With CDs, I seldom use the line gain amp, so the signal goes straight
on into the poweramp, with only a cathode follower buffer,
to cope with a long cable.
A friend is building a SET preamp with one EL84 in triode,
and with a choke load, and 10 kOhm pot, to power
his Nelson Pass SEM amp. We expect music.
One could use a low Ra triode, such as the EL86
wired as a triode for a line preamp, it has a gain of 7,
and Ra of about 1.3 kOhms, at 15 mA of plate current.
This would work nearly as well as a cathode follower,
but may in fact sound better, as there isn't the local NFB
involved when the cathode follower is used.
It isn't a big issue, as signal levels are very low anyhow.
These small output tubes make very nice driver tubes in
many power amps, and they look right, beside
6550 or KT88.
> How do you like the 6922 or better yet the 6H30 for a preamp Patrick?
Well, nearest thing to 6922 that I have tried is a 6DJ8, old german madestock.
I've never tried 6H30.
Like a lot of things, how it sounds is a bit dependant on the way you
load the tube, as well as its natural harmonic structure.
But somehow the "sound of a certain tube"
is a bit like taking different bottles of fine wine from the stack.
bottles of the same wine taken from the same cask at different times affects
the taste, bouquet, texture, finish, aftertaste, and so on.
The time you leave the wine bottle opened before drinking
affects the experience.
Its a bit the same with tubes. leave the amps on for 3 hours.
I bet they sound better than after ten minutes.
I have heard people say that this or that brand of 12AX7 are very
different, and that one is more better than another.
I am just left to wonder.
I have noticed differences, but its all a bit flukey.
Most Mullards are a bit flompy and warm.
That's probably all I should say.
When you find a good one, then stick to it.
People used to use discrete transistors much more 30 years ago.
Now its all opamps.
Seriously, folks used to argue about which sort of
transistor sounded the best.
I have a book with 32,000 transistors listed within.
There are parameter variations within the same batch.
Lords knows which 6 transistors are the best sounding!!!
My experiences of 6DJ8 is that I found it worked
OK as a paralleled input tube for a phono amp.
But it wasn't stunningly better than a 12AX7.
I used the DJ8 to try to take advantage of the higher tube transconductance
to get a lower noise figure. But despite all efforts,
the tube I selected out of stocks was 3 dB noisier, instead of
6 dB quieter, which the theory books predict.
Perhaps the handpicked AX7 was somehow defiant of theory
and succeeded in being 9 dB quieter than predictions.
The sound one hears is the product of the ALL the sonic influences
of the venue and microphone used in the recording
right through studio gear, and then via the silver or black disk,
perhaps a tape, or two, and thence to our pinkies,
which differ from person to person, in frequency response,
distortion, dynamic ability.
Some folks can hum along to opera, and understand
all the words, many of us fail at both.
Some hear differences in tubes that others cannot hear no
matter how hard they try.
What the brain does with what is heard varies enormously.
But you know, despite all these hearing variations, and variations
in audio gear quality, if you feel no emotion when the best of
Bethoven is played, even though you may be quite musically
uneducated, then why didn't you go fishing?
The amp with the 6DJ8 is in a pic at the 'repairs' page
at my web site at http://www.turneraudio.com.au
This preamp is obviously an old Quad 22, extensively
altered, and it has 7 twin triodes within.
It sure does the job ok.
In another preamp, I tried using a DJ8 with some shunt feedback
to reduce gain from 24 down to 6 for a linestage amp.
I may as well have built a "Granite Filter", as the sound
was hard and dry, and not quite right.
I replaced this tube with a 12AU7, still with shunt feedback
to reduce gain from 12 to 6, and wow, very nice indeed.
I tried several AU7 and they all sounded different.
An old Mullard was too mushy and warm.
Another was noisy, another was microphonic,
despite being milspec.
I found a suitable one that was not mushy,
not hard, not noisy, but was engaging, was musical,
and had me saying, "yessir, that'll do!"
My client was pleased.
Patrick Turner.
Kavlex
project. It is almost exactly the same.
Regards.
Billy...
"ee" <e...@hotmail.com> wrote in message
news:3c47229d...@news.telus.net...
ee
transformer load the MOSFETs source or drain?
I made a SE MOSFET amp and its really brings out the delicate detail
and ambiance in CDs.
I don't like to copy other people's schematic value for value or part
for part, but it can lead to new topographies like in this case.
EE
Patrick Turner
ee wrote:
> Patrick, regarding your SS power amp, does the primary of your output
> transformer load the MOSFETs source or drain?
There was a recent posting of mine,which may answer any questions more
fully...
Here is a copy of that.....
See more comments following the copy of my previous post.
Subject:
Re: Jensen Paper In Oil? and Pat's mosfet amps
Date:
Fri, 18 Jan 2002 19:00:45 +1100
ee wrote:
> I sure like to see the schematic of Patrick's MOSFET poweramps.
> Why pay for 300B when Mosfets can do the job.
Eventually I will publish the Mosfet class A PPschematic on the
website of mine.
Keen experimenters will design their own
like I have, rather than use another man's ideas.
The large compact 300 watt per channel
class AB amp is not too bad an amp, but a bit mundane,
and plain within, as it is just another direct coupled
SS amp with LTP front end, two transistor voltage amp, and source
follower output stage. It goes well, and makes 0.003 % of THD
at 300 watts. I couldn't measure the crossover distortion.
200,000 uF of caps are used, and the drive amp rails are
actively regulated.
These results are elementary for folks serious about SS.
The other smaller amp is a 50 watt class A with OPT.
It was harder to build than a 300B amp.
I made my own heatsinks.
I had all this scrap aluminium I had accumulated.
They needed to have a large total surface area.
But during our recent hot
weather where we had 34 degrees C in the lounge,
the amps remained stable, and the mosfet cases went up to about
60 degrees C.
During winter, all parts can be handled without fear of a scorch.
The output transformer for these amps are just as complex as
a tube type, but the wire is thicker, as less turns are used,
for the lower impedances, 35 ohms drain to drain,
to 5 ohms at the sec.
OPT bandwith was 10 Hz, at saturation, to 300 kHz.
C-cores with a max Mu = 5,000 were used, but then
reduced by a small air gap to 2,200, for better
saturation behaviour at LF.
To guarentee stability, I used some gain reduction
internally to limit the output bandwidth to about 65 kHz.
So, with all things considered, the 300B amp
would be dead easy, compared to all this.
But these amps make 50 watts to 5 ohms,
and not just 8 watts, like the 300B.
I think 4 x 300B would be needed in PP for 50 watts, class AB.
But who cares, power is not everything.
Each side of the PP circuit uses two Hitachi TO3 pack mosfets.
Exicon TOP3 would have been better, but mine were donated
to me "free", in a batch for some design favour I did for
a Sydney amp maker.
I tested a couple out seperately to see what their transfer curves
were.
They are almost identical to a good pentode,
with not much 2H, and more 3H when the pace hots up.
So, like pentode amps, not much is to be gained by
using mosfets in SEM mode, IMHO.
Nelson Pass uses them but runs them dangerously hot, with
50 watts of quiescent dissipation, for one lonesome mosfet.
This is at 17 volts at 3 amps, where their transfer curve
is supposed to be more like a triode's, with more 2H.
I tried mine out at 32 volts at 0.8 amps, ie about 25 watts
quiescent,
and the first few watts quite blame free, and not too bad at clip.
( The question comes to mind, would triodes sound better without
their natural distortion? )
The mosfet circuit uses a simple small signal, 2 transistor
longtail pair,
with a low noise CCS transistor tail.
local emitter resistors keep this stage very linear.
This stage is cap coupled to the mosfets, like tubes are.
Fixed bias is used, commom for all mosfets.
Only 0.8 volts rms is needed to the mosfet gate to
drive the output stage to full power
Since they have tendency to reduce their own DC current flow
as the temperature rises, there is no need for any form of VBE
transistor to alter bias when things get hot.
The 4 N channel mosfets are set up in common source mode, like
a quad of pentodes.
So, the total number of stages is less than in many tube amps,
and a shirtload less than in nearly all other SS amps where
the transistor count can get up to 25 quite easily.
There are 3 loops of NFB.
( so its a bit like a Leak amp )
There is shunt feedback from the drain circuits of the mosfets to
the collectors of the LTP.
This amounts to 10 dB and reduces Zout from 72 ohms
down to 1.5 ohms.
It was still a little hissy, but sounded very like a triode PP
amp with no NFB.
I put a bit more NFB in by using the third path from the
OPT secondary to the second input to the LTP.
Voila, no noise, and 0.2 % THD at 50 watts, with
only a total of 22 Db of NFB, instead of the
usual total of 82 dB used in most SS amps,
which doesn't make them sound any better,
and I am not sure why.
Because the amps are real class A amps,
the distortion at a watt, even with ZNFB,
is quite acceptable, but the Zout is too high,
and I prefer to get it below 0.3 ohms,
so NFB has to be used.
Placing a 10 uf cap across the output
could not be heard during a music session.
The longer term afterglow from these amps is pleasant.
Its like some amps leave you with a good audio memory,
others you are not so sure about.
Compared in A-B tests, I couldn't hear any differences with
any large class A tube amps.
But A-B testing, although this is very revealing,
or tending to reveal nothing, where differences are alleged,
the results can be deceptive in that our subconscious works away
about what has been listened to, and may reach
some other conclusion over time.
The power supply has B+ = 35v, at 3 amps, to the CT,
and uses a high current bridge to charge 9,400 uF,
then a 150 mH inductor,
and 45,000 uF at the CT.
The CT ripple voltage is in millivolts.
Driver amp rails are shunt regulated by zeners, and huge caps.
Power tranny is a 300 VA toroidal, which is hard to see in the pic
at my site.
People think the OPT is the power tranny, but its not.
One cannot blow up speakers with this SS amp,
thanks to the OPT, and common mode ripple rejection
is excellent, like in PP tube amps.
Patrick Turner. http://www.turneraudio.com.au
> EE
>
>
> I made a SE MOSFET amp and its really brings out the delicate detail
> and ambiance in CDs.
>
> I don't like to copy other people's schematic value for value or part
> for part, but it can lead to new topographies like in this case.
>
> EE
In case it wasn't clear, when people say a tube is set up in common
cathode mode, they mean the cathode is grounded,
and the load is in the plate circuit, with the grid voltage able to move
to
change the load current, and alter plate voltage.
Nearly all tube amps are set up like this.
In the case of NPN mosfets, the source is just like the cathode,
drain is like the anode, and gate is just like the grid.
When a mosfet amp is in common source mode,
the source is grounded, and the drain voltage is allowed
to move as the load is between the drain and power supply.
This is how I have my mosfets set up, and it very
similar to most PP tube output stages, such as an ST70's.
Most solid state amps, mosfet types included, have their
output mosfets set up in source follower mode,
just like a cathode follower. This is a way of
applying local voltage negative feedback.
Cathode follower tube output stages are not much employed,
because the grid drive voltage would have to be greater
than the output voltage which normally appears only
at the anode.
But with follower action in SS devices, the drive
voltage required to turn the device on and off
between the gate and source is very low,
only a volt or two, and the output voltage which goes to a low
impedance load is not very high either, 10 to 20 volts perhaps.
So in follower mode, the input gate voltage must be
just marginally more than the output voltage, which is easy to
to do.
Say a mosfet has a voltage gain of 10.
In a follower, it might be only 0.9, and gain reduction
is from 10 to 0.9, about 21 times, and the amount of NFB is about
27 dB, which is a lot.
In my mosfet amps, there is no such follower action,
but I use shunt feedback from the drains to the drive amp
collector cicuits.
So I have a form of an anode follower.
The equivalent principles in tube circuitry
are explained in RDH4 on pages 332 and following pages.
An application of the principle is shown in my amp circuit
at my site
http://www.turneraudio.com.au
when you scroll 2/3 down the page titled "Philosospy, Circuits and DIY
images"
and click on 'Local Shunt Feedback.gif'
Because I only have to produce 1 vrms to drive the mosfet gates,
instead of 21 vrms, I can have a very simple and linear drive amp.
In SS circuitry, shunt feedback form the output to the drive
amp collector circuits is very effective as the drive amp
is a current source, unlike a triode, which is a voltage source.
Therefore any small variation in output voltage
and subsequent current fed back will
change the drive amp GAIN, due to the change of load value the
drive amp sees, and thus a correction for output
voltage change is possible.
It is especially effective in mosfet amps, because they
have a very high input gate Z, like a tube, and their gain
can, like I have used them, be quite high, like a pentode's,
where voltage gain is about the
product of transconductance x load value.
I have posted before on this subject of local shunt feedback, but
nobody had the slightest idea of what the heck I was
trying to tell them.
Patrick Turner.
Patrick Turner
I am just sending this again, after resetting my PC clock,
which somehow was causing the post to appear well up in the list,
and may have been missed by some readers. P.T.
ee wrote:
> Patrick, regarding your SS power amp, does the primary of your output
> transformer load the MOSFETs source or drain?
There was a recent posting of mine,which may answer any questions more
ee wrote:
Patrick Turner. http://www.turneraudio.com.au
> EE
>
>
> I made a SE MOSFET amp and its really brings out the delicate detail
> and ambiance in CDs.
>
> I don't like to copy other people's schematic value for value or part
> for part, but it can lead to new topographies like in this case.
>
> EE
In case it wasn't clear, when people say a tube is set up in common
Ned Carlson
wrote:
>Patrick,
>
>What's your opinion of the 6EU7 vs. the 12AX7?
>
>Thanks.
>
>Jon
I know some folks claim 6EU7 sounds better than 12AX7,
but given that they're from the same manufacturer and vintage,
they're electrically and mechanically identical except for the
pinout.The idea behind 6EU7 is that by rearranging the pins, there'd
be less filament hum pickup.
Sovtek makes a 6EU7 currently.
Ned Carlson Triode Electronics "where da tubes are!"
2225 W Roscoe Chicago, IL, 60618 USA
ph 773-871-7459 fax 773-871-7938
12:30 to 8 PM CT (1830-0200 UTC) M-F
http://www.triodeelectronics.com
"Beer is proof that God loves us and wants
us to be happy."
Benjamin Franklin