ultralinear: too many compromises.
Anne Merete Driveklepp
>Randy Nachtrieb:
>You can't have your cake & eat it too. The type of output signal
>that comes out of a McIntosh is about as different as that from a SE
>triode as one could get. The asymetrical nonlinearities of a SET produce
>predominantly even harmonic products of low order, the low NFB producing
>a low damping factor with high distortion.
NO! This is just a narrow wiew to it. A properly constructed SET amp has
low
distorsion, not high...Remove the nescessary feedback loop from one typical
pent. p-p amp... THEN you might experience high distortion..............
The damping factor is also wery good, considered the normal practice of no
feedbaack.Try to disconnect the feedback loop from one typical pent. p-p
amp....THEN you might experience some real absent damping factor.......
> The highly balanced
>Unity-Coupled output stage of the Mac however, produces predominantly
>odd order harmonics; furthermore, in contrast to low NFB designs (& most
>UL amps), the overload performance of a Mac is rather ill behaved.
>As to UL amps having "oscillations", poor stability & "lousy"
>damping factor, I suggest you take the time to learn how to bench test
>these parameters. If you do, you will find that your esteemed Mac
>performs every bit as "poorly".
This is partly incorrect: The design, as any p-p amp cansel the equal (2, 4
etc.... )harmonics.It does not 'produse more' of the odd order types,- Au
contraire,- the stage has wery low distorsion alltogether....It has also
better damping factor vs. the ultralin. output stage I believe, open loop
vice,
because of the heavy l o c a l feedback applied..The design IS in fact
more
stable in itselves, but the fact remains that all Mac-amps was designed for
POWER!! (And by measurements I believe....)
It has wery low idling current, and quite a sloppy power-supply for being a
(near) klass 'B' amp....That should explain inapropriate behaviour near
clipping...........
> In the days of Hafler & Keroes, such things as the relative balance
>of generated harmonics were not taken as major design factors(perhaps
>they shouldn't be). The UL configuration was an invention & was
>patented. Furthermore, its design goals were genuine & not an attempt to
>fleece the public, the intentions of the designers being far less open
>to speculation than those who peddle 5 watt SET amps.
>If you decide to gain more technical knowledge, you may want to
>acquire considerably more(& varied) listening experience to go along
>with it. As most of the "oldtimers" on this newsgroup would likely
>attest,you can find amps which sound quite good, along with the bad,
>among most configurations.
Regards,
Randy Nachtrieb
Enlightened Audio
New Art Loudspeakers
Yepp,I agree completely.
----------------------------------------------------------------------------
-
>I think, this about sums up your post nicely.
>Read the replies to your original post, especially
>those from John Byrns, Randy Nachtrieb and myself.
>AFAIK, tube courses aren't given anywhere, but there
>are a lot of nice books out there, previously listed in this
>NG.
_
>Sander deWaal
>postm...@pegasus.demon.nl
Hi Sander,
I have a feeling in my bones that you are riding your high
horse here...Of course I might be wrong.
I'm not subscribing to all the wievs of mr. Macfanatic, but he has,
and should be allowed to have some v a l i d points..Now, the
disadvantages of the UL-stage is I guess well known, and have been
described in some detail by a constructor of 'some' knowledge,-
Mr Jean Hiraga......Somewere I have a few articles on a construction
of an EL34 amp.of his.The descision turned out to favor the plain
pentode amplifier.....
As I have tried to hint above, a separation of 'principle' and 'execution
of principle' is in demand while discussing these topics.
Torbjoern Lien.
.
Sander deWaal
On 17 Jan 1998 15:47:53 GMT, "Anne Merete Driveklepp"
<mdri...@online.no> wrote:
Randy Nachtrieb wrote:
>>You can't have your cake & eat it too. The type of output signal
>>that comes out of a McIntosh is about as different as that from a SE
>>triode as one could get. The asymetrical nonlinearities of a SET produce
>>predominantly even harmonic products of low order, the low NFB producing
>>a low damping factor with high distortion.
And T Lien wrote:
>NO! This is just a narrow wiew to it. A properly constructed SET amp has
>low distorsion, not high...Remove the nescessary feedback loop from one typical
>pent. p-p amp... THEN you might experience high distortion..............
>The damping factor is also wery good, considered the normal practice of no
>feedback.Try to disconnect the feedback loop from one typical pent. p-p
>amp....THEN you might experience some real absent damping factor.......
All Randy's comments are relative to minimal standards of hifi.
Even a properly designed SET has high distortion and low damping
factor, as compared to modern standards.
But, this aside, he wanted to point out that there are *huge*
differences between typical SE- and UCC or even UL sound.
This doesn't mean, that one could not prefer the sound of
SET's, just the technical facts.
AN:
>> The highly balanced
>>Unity-Coupled output stage of the Mac however, produces predominantly
>>odd order harmonics; furthermore, in contrast to low NFB designs (& most
>>UL amps), the overload performance of a Mac is rather ill behaved.
>>As to UL amps having "oscillations", poor stability & "lousy"
>>damping factor, I suggest you take the time to learn how to bench test
>>these parameters. If you do, you will find that your esteemed Mac
>>performs every bit as "poorly".
TL:
>This is partly incorrect: The design, as any p-p amp cansel the equal (2, 4
>etc.... )harmonics.It does not 'produse more' of the odd order types,- Au
>contraire,- the stage has wery low distorsion alltogether...
Randy says: "predominantly odd order harmonics", not "more".
A question of magnitude.
This nature of distortion, as opposed to the type of distortion,
a SET produces.
TL:
>.It has also
>better damping factor vs. the ultralin. output stage I believe, open loop
>vice,
>because of the heavy l o c a l feedback applied..The design IS in fact
>more
>stable in itselves, but the fact remains that all Mac-amps was designed for
>
>POWER!! (And by measurements I believe....)
Indeed the UCC has a higher damping factor than the UL.
UL is also a form of local feedback.
But why should it be more stable?
TL:
>It has wery low idling current, and quite a sloppy power-supply for being a
>(near) klass 'B' amp....That should explain inapropriate behaviour near
>clipping...........
It was designed, to overcome the crossover distortion of
(near) class B amps.
It introduces more distortion of another kind, up to severe
high frequency oscillation into reactive loads.
So, why should it be more stable?
Sander deWaal wrote:
>>I think, this about sums up your post nicely.
>>Read the replies to your original post, especially
>>those from John Byrns, Randy Nachtrieb and myself.
>>AFAIK, tube courses aren't given anywhere, but there
>>are a lot of nice books out there, previously listed in this
>>NG.
T'Lien wrote:
>Hi Sander,
>I have a feeling in my bones that you are riding your high
>horse here...Of course I might be wrong.
Nope, I don't have special thoughts about either one
of the mentioned output circuits, it's just that mr. MacFanatic
was just plain out WRONG with his conclusions.
TL:
>I'm not subscribing to all the wievs of mr. Macfanatic, but he has,
>and should be allowed to have some v a l i d points..Now, the
>disadvantages of the UL-stage is I guess well known, and have been
>described in some detail by a constructor of 'some' knowledge,-
>Mr Jean Hiraga......Somewere I have a few articles on a construction
>of an EL34 amp.of his.The descision turned out to favor the plain
>pentode amplifier.....
As you probably have noticed, I never said that UL-amps
were without problems or sonic disadvantages.
Except from the fact, that I use an amp with negative grid bias,
as opposed to cathodebias or UCC.
I wanted to address the claim, that short tube life and/or
red glowing plates, were due, solely to the UL configuration
of the output stage.
A statement like that is pure nonsense, and could confuse
others.
The only valid point he has, IMO, is that the UCC *sounds*
better to him. No one can argue about preferences, and this
is his prerogative, of course.
>As I have tried to hint above, a separation of 'principle' and 'execution
>of principle' is in demand while discussing these topics.
I'll try to remember that.
But also, the mispresentation of facts to justify one's choice of
circuit, can be very confusing to newbies, and annoying to "veterans".
Mr MacFanatic as a record here, as long-time lurkers will know.
He revealed his name to me in private email, and I won't stoop that
low as to mention it here, but I think the RAT-regulars all know who
he is.
_
Sander deWaal
postm...@pegasus.demon.nl
______________________________________________________
Many real strides in technology and new inventions
came from backyard scientists and garage engineers,
hobbyists and tinkerers, et al. (Gabe Velez)
______________________________________________________
Randy Nachtrieb
Hi Sander,
Thank you for your most accurate clarifications of my previous
post.
Best,
Randy
Anne Merete Driveklepp
>Randy Nachtrieb wrote:
>>>You can't have your cake & eat it too. The type of output signal
>>>that comes out of a McIntosh is about as different as that from a SE
>>>triode as one could get. The asymetrical nonlinearities of a SET produce
>>>predominantly even harmonic products of low order, the low NFB producing
>>>a low damping factor with high distortion.
>>And T Lien wrote:
>>NO! This is just a narrow wiew to it. A properly constructed SET amp has
>>low distorsion, not high...Remove the nescessary feedback loop from one
typical
>>pent. p-p amp... THEN you might experience high distortion..............
>>The damping factor is also wery good, considered the normal practice of
no
>>feedback.Try to disconnect the feedback loop from one typical pent. p-p
>>amp....THEN you might experience some real absent damping factor.......
>All Randy's comments are relative to minimal standards of hifi.
>Even a properly designed SET has high distortion and low damping
>factor, as compared to modern standards.
Yes,-compared to modern standards which always including loads of
corrective feeback.
No, these (well-designed) amps have low distortion and good damping
factors,
considered the absence of feedback! Spec'-vise one could make a SET-amp
superior to most p-p pentode amps, with an addition that the former (p-p)
type
already has: Just put another gain-stage in, and remove the gain,ex. say
20dB,
with global feedback.....
>But, this aside, he wanted to point out that there are *huge*
>differences between typical SE- and UCC or even UL sound.
>This doesn't mean, that one could not prefer the sound of
>SET's, just the technical facts.
That's ok.
>AN:
>>> The highly balanced
>>>Unity-Coupled output stage of the Mac however, produces predominantly
>>>odd order harmonics; furthermore, in contrast to low NFB designs (& most
>>>UL amps), the overload performance of a Mac is rather ill behaved.
>>>As to UL amps having "oscillations", poor stability & "lousy"
>>>damping factor, I suggest you take the time to learn how to bench test
>>>these parameters. If you do, you will find that your esteemed Mac
>>>performs every bit as "poorly".
>TL:
>>This is partly incorrect: The design, as any p-p amp cansel the equal (2,
4
>>etc.... )harmonics.It does not 'produse more' of the odd order types,- Au
>>contraire,- the stage has wery low distorsion alltogether...
>Randy says: "predominantly odd order harmonics", not "more".
>A question of magnitude.
>This nature of distortion, as opposed to the type of distortion,
>a SET produces.
Ok, guess I hang up on:"produces ......
>TL:
>>.It has also
>>better damping factor vs. the ultralin. output stage I believe, open loop
>>vice,
>>because of the heavy l o c a l feedback applied..The design IS in fact
>>more
>>stable in itselves,but.....
>Indeed the UCC has a higher damping factor than the UL.
>UL is also a form of local feedback.
>But why should it be more stable?
OK, I'll try, although I have to put some lenght to it: I'll quote
myselves:
"The design IS infact more stable in itselves, but........."
Here I simply mean that the design,-as a constructing prinsiple is more
stable than the UL-design,-also as a constructing prinsiple....
Why is that? The purpose of both designs is to produse power into a
loadspeaker, which always has a complecs impedance..........
Now although the UL could be said to work with error reduction by means
of feedback to the #2-grids, this reduction is by no means l i n e a r ,
(because it is applied to G2), and we need some more help before the design
can be put to work. We need feedback from at least one preceeding gain
stage
to reduse distortion and output resistance, and by doing that we must deal
seriously with phase shift, poles, and so on....A global feedback system
is simply more phrone to bad behaviour because of this, and the fact that
you cannot foresee the complexity of the load..........
Now, the UCC-design is relying on cathode feedback, and infact COULD be
relying on that alone. The output resistanse, and distortion is redused in
a
linear manner, and with the proper selection of tubes (triodes?) distortion
and output impedance would already (before global feedback) be superior.
As this happens with local feedback only, all we have been doing is to
'modulate' or 'rearrange' the original tube spec's, so we have an
inherently
stable output stage..
>>TL
>>It has wery low idling current, and quite a sloppy power-supply for being
a
>>(near) klass 'B' amp....That should explain inapropriate behaviour near
>>clipping...........
>It was designed, to overcome the crossover distortion of
>(near) class B amps.
>It introduces more distortion of another kind, up to severe
>high frequency oscillation into reactive loads.
>So, why should it be more stable?
Again I'll quote myselves: "....but the fact remains that all Mac-amps
was
designed for POWER!! (And by measurements I believe....)"
I think I already have responded to that in the description of the
"execution of
the principle" (low, near B-biasing, a sloppy supply.) I might add: The
common
global feedback, and added leakage inductance to the output transformer.
(Shame on you Frank!) I guess that if the amp had been redesigned to music
use, it just might................(Well,- guess that depends on taste to,-)
>>
>Sander deWaal wrote:
>>>I think, this about sums up your post nicely.
>>>Read the replies to your original post, especially
>>>those from John Byrns, Randy Nachtrieb and myself.
>>>AFAIK, tube courses aren't given anywhere, but there
>>>are a lot of nice books out there, previously listed in this
>>>NG.
>T'Lien wrote:
>>Hi Sander,
>>I have a feeling in my bones that you are riding your high
>>horse here...Of course I might be wrong.
>Nope, I don't have special thoughts about either one
>of the mentioned output circuits, it's just that mr. MacFanatic
>was just plain out WRONG with his conclusions.
(Sorry, I did not follow his input that close....)
>TL:
>>I'm not subscribing to all the wievs of mr. Macfanatic, but he has,
>>and should be allowed to have some v a l i d points..Now, the
>>disadvantages of the UL-stage is I guess well known, and have been
>>described in some detail by a constructor of 'some' knowledge,-
>>Mr Jean Hiraga......Somewere I have a few articles on a construction
>>of an EL34 amp.of his.The descision turned out to favor the plain
>>pentode amplifier.....
>As you probably have noticed, I never said that UL-amps
>were without problems or sonic disadvantages.
>Except from the fact, that I use an amp with negative grid bias,
Yess! Cathodebias is argueably plain silly! (Oops! shouldn't have said
that,-guess I'm in for some flaming.......Heck, the NG probably don't
notice anyway............ :-).
>as opposed to cathodebias or UCC.
>I wanted to address the claim, that short tube life and/or
>red glowing plates, were due, solely to the UL configuration
>of the output stage.
>A statement like that is pure nonsense, and could confuse
>others.
Yepp, for sure is!!
>The only valid point he has, IMO, is that the UCC *sounds*
>better to him. No one can argue about preferences, and this
>is his prerogative, of course.
Hmm, I do not agree......as before stated, I find the UCC to be a design
with greater possybilities, both construction and sound-vice.(A good
and e a s y compromice I guess would be to lessen the work of the
voltage drivers, by redusing the amount of local feedback.( I think that
the quad 2 is a remarkable design in it's class.))
>>As I have tried to hint above, a separation of 'principle' and 'execution
>>of principle' is in demand while discussing these topics.
>I'll try to remember that.
>But also, the mispresentation of facts to justify one's choice of
>circuit, can be very confusing to newbies, and annoying to "veterans".
>Mr.MacFanatic has a record here, as long-time lurkers will know.
>He revealed his name to me in private email, and I won't stoop that
>low as to mention it here, but I think the RAT-regulars all know who
>he is.
Guess I'm a newbie then, I didn't know...............
>Sander deWaal
Torbjoern Lien
>
>
>
>
macfa...@hotmail.com
Dear Randy,
Thanks for taking your valuable time to respond. There is certainly no
intention to compare the Mac to a SET amp from a technical standpoint.
Both amps handle the signal in completely different ways, yet the
subjective musical result is pretty much the same. The music comes
through and that’s the important thing.
Concerning the overload characteristics of the McIntosh Circuit, well I
have to disagree with you completely. Before I bought my MC275 I was
always plagued by annoying clipping distortion no matter how powerful the
amp was. The Solid Staters were simply trash. I then bought my pair of
MKIVs hoping I could have a taste of real musical performance. I
unpacked and connected my Dynas, furnished by Audio Classics, and put
them through their paces with near mint UHQRs, Nautilus and other nice
vinyl recordings, as well as some of my reference CDs (just so see if
they sweetened nasty digital artifacts). Much to my chagrin the Dynas not
only failed to live up to my expectations but also gave me a number of
headaches (listening fatigue AND tinkering fatigue!!!). Thankfully, I
have the MC275 now. The thing has really worked wonders for me. At last I
got rid of nasty overloads and clipping distortion!!!. Music is
reproduced with amazing believability and astonishing “u-r-there”
perspective. At first, I though the McIntosh was all about looks and not
much more. Listening tests have confirmed otherwise. I can push the Mac
hard with no signs of break up. By comparison, the 40 watt Mark IVs DO
have poor overloading characteristics no matter what output tube I use.
It is the same with cheapie Teslas than with NOS Mullards.
Remember the inflated claims made by manufacturers in the 60s and 70s
which caused the FTC to issue specific rules to protect the consumer?.
Well, McIntosh amps were the only ones who stood up to their
specifications and certainly paved the way for modern scientific
amplifier measurements. The “competing” designs by contrast, exhibited
gobs of distortion and confused the poor consumer with peak power, music
power and other stupidities.
While I haven’t tested UL and Unity Coupled for stability and damping, I
think that pumping speaker cones are sufficient proof of amplifier
misbehavior. Further, strange audible high frequency artifacts are
undeniable proof of oscillation. In truth this phenomena can be
experienced with any tube amp. The difference lies in the fact that such
performance aberrations ARE part of normal UL operation as opposed to the
Mac circuit in which they are usually caused by a serious malfunction.
My MC30 does exhibit the pumping cone phenomena, but this is caused by
out-of-spec 40 year old parts and NOT by result of poor circuit design.
Of course, the UL was patented. But remember the UL idea was taken from
Blumein and therefore wasn’t anything revolutionary for that matter. Poor
Ed Villchur had his acoustic suspension patent revoked because he
registered a design developed years ago by Harry Olson at RCA Labs. Any
similarities with Hafler, Keroes and Blumein is pure coincidence.
Finally, I bet your “New Art” loudspeakers don’t mate well with McIntosh
amplifiers. That’s really a shame. But hey cheer up! ; perhaps you could
peddle your speakers as the ultimate transducers to turn a humble Dyna
into a decent musical performer. Send me a pair of your speakers to
audition with my Dynas. I´ll spin some Robert Fine engineered Mercury’s,
Shaded Dogs and perhaps some cool Brubeck, Miles, Coleman and Coltrane
LPs and CDs (yes CDs). If the Dynas, via New Art speakers, give me
Goosebumps, then man I’ll declare you a genius (provided you have
designed them, of course!).
Regards,
N J Salguero
The Macfanatic
-------------------==== Posted via Deja News ====-----------------------
http://www.dejanews.com/ Search, Read, Post to Usenet
John Byrns
> While I haven’t tested UL and Unity Coupled for stability and damping, I
> think that pumping speaker cones are sufficient proof of amplifier
> misbehavior. Further, strange audible high frequency artifacts are
> undeniable proof of oscillation. In truth this phenomena can be
> experienced with any tube amp. The difference lies in the fact that such
> performance aberrations ARE part of normal UL operation as opposed to the
> Mac circuit in which they are usually caused by a serious malfunction.
> My MC30 does exhibit the pumping cone phenomena, but this is caused by
> out-of-spec 40 year old parts and NOT by result of poor circuit design.
Dear Fanatic,
I think you are way off base on the speaker pumping issue. In my
experience, speaker pumping is almost always the result of playing vinyl
recordings through a system that doesn't include a proper infrasonic
filter, and has little to do with stability problems in the power amp. An
infrasonic filter is absolutely essential in a good vinyl playback system,
and ideally one should be included in the RIAA preamp circuit. I would
guess that you preamp doesn't include this essential circuit feature,
hence the cone pumping with power amps that don't include infrasonic
filtering. I suspect that the McIntosh amps include an infrasonic filter
in their input circuit, and that is why they work for you. I know the
Dyna mark IV's don't include an infrasonic filter, so it makes sense that
you would have cone pumping when using them with a substandard RIAA
preamp. I will have to check the MC275 circuit, and literature to see if
they include an infrasonic filter, is yours the original or re issue
MC275? Perhaps the cone pumping you experienced with the MC30, is not the
result of "out-of-spec" parts, but is the result of the lack of an
infrasonic filter in the MC30, I don't know, it's just a thought.
Regards,
John Byrns
Sander deWaal
On 17 Jan 1998 20:15:23 GMT, "Anne Merete Driveklepp"
<mdri...@online.no> wrote:
>>All Randy's comments are relative to minimal standards of hifi.
>>Even a properly designed SET has high distortion and low damping
>>factor, as compared to modern standards.
>
>Yes,-compared to modern standards which always including loads of
>corrective feeback.
My very own tube amps produce a whopping .6% THD, at
levels up to 15 Watt.
Above that, they go up slowly up to 3 % at 18 watts.
At average listening levels, it's not more than .3%.
They have 10 dB of global feedback (shoot me now!!!)
Compared to the average THD of a SET, which may be
5 % at decent listening levels, rather low.
Now, I believe that the *nature* of the distortion is far more
important than the figure alone.
THD in itself is a rather meaningless thing , as long
as it is under a 5 % or so.
As I use PP, most residual distortion is odd order, so it
HAS to have a lower THD than SET amps.
All of this is no judgment of the sound quality, let alone
the preference that one may have for one or the other sound,
but just a clinical fact.
>No, these (well-designed) amps have low distortion and good damping
>factors,
>considered the absence of feedback! Spec'-vise one could make a SET-amp
>superior to most p-p pentode amps, with an addition that the former (p-p)
>type
>already has: Just put another gain-stage in, and remove the gain,ex. say
>20dB,
>with global feedback.....
They have indeed, *considering the absence of feedback*.
Doesn't contradict what Randy or I wrote earleir, IMO.
And spec's don't mean much, if one doesn't like the sound
of an amp.
Why trying to convert a SET amp so, that it meets the
specs of a PP penthode amp? (Not that that would be easy,
far from that!)
>OK, I'll try, although I have to put some lenght to it: I'll quote
>myselves:
>"The design IS infact more stable in itselves, but........."
>Here I simply mean that the design,-as a constructing prinsiple is more
>stable than the UL-design,-also as a constructing prinsiple....
>Why is that? The purpose of both designs is to produse power into a
>loadspeaker, which always has a complecs impedance..........
>Now although the UL could be said to work with error reduction by means
>of feedback to the #2-grids, this reduction is by no means l i n e a r ,
>(because it is applied to G2), and we need some more help before the design
And here it is where we digress.
If one takes a look at what really is happening with G2 feedback,
one would see that the Ia-Ua characteristics of a beamtetrode or
penthode are highly linearised.
Meaning, that for a given point of operation on the characteristics,
less distortion will be generated.
In fact, less than in a triode, and far less than pure penthode.
Now, the only drawback is, that the 43% tap on the primary
winding is highly frequency-dependent, so it's not possible
to ensure these characteristics for the entire frequency spectrum.
Not to metion stray inductance and -capacitance.
Now take the Quad/Mac approach: In order to make the output
stage to work in class B, they added an extra winding on the
output transformer, to overcome the crossover distortion that
would occur without it.
As long as we don't look at the frequency extremes, and drive
the stage into overload, all will function relatively well.
At extremes, and overload, all kinds of phaseshifts will occur,
especially when global feedback is applied also, as do
the Mac and the Quad.
(And quite heavily, I might add).
This may well lead to oscillations, in case of a highly reactive load.
So, both circuits do have their drawbacks, with the advantage of UCC
that it behaves more civil in average signal handling, in not too
reactive loads.
>can be put to work. We need feedback from at least one preceeding gain
>stage
>to reduse distortion and output resistance, and by doing that we must deal
>seriously with phase shift, poles, and so on....A global feedback system
>is simply more phrone to bad behaviour because of this, and the fact that
>you cannot foresee the complexity of the load..........
Yup, one reason why pure penthode or triode stages are more stable
than either UL, UCC or PCCC.
>Now, the UCC-design is relying on cathode feedback, and infact COULD be
>relying on that alone. The output resistanse, and distortion is redused in
>a linear manner, and with the proper selection of tubes (triodes?) distortion
>and output impedance would already (before global feedback) be superior.
>As this happens with local feedback only, all we have been doing is to
>'modulate' or 'rearrange' the original tube spec's, so we have an
>inherently stable output stage..
With triodes, I would agree to a certain point.
It takes a healthy design of the output transformer, a qualification
that the Quad trannies would meet earleir than the Mac's, IMO.
>I think I already have responded to that in the description of the
>"execution of
>the principle" (low, near B-biasing, a sloppy supply.) I might add: The
>common
>global feedback, and added leakage inductance to the output transformer.
>(Shame on you Frank!) I guess that if the amp had been redesigned to music
>use, it just might................(Well,- guess that depends on taste to,-)
I still think that the UCC could provide as much pleasure
as any other output stage, providing good trannies and
a friendly load are used.
I could have been more clear about this in my original reply,
though.
Guess I was blinded by mr. MacFanatic's rhetorics..... :-)
>>As you probably have noticed, I never said that UL-amps
>>were without problems or sonic disadvantages.
>>Except from the fact, that I use an amp with negative grid bias,
>
>Yess! Cathodebias is argueably plain silly! (Oops! shouldn't have said
>that,-guess I'm in for some flaming.......Heck, the NG probably don't
>notice anyway............ :-).
I do! :-)
And cathode bias isn't silly, I never said that.
mr. MacFanatic used UL as a tool to show that
tube life would be shorter than in a UCC circuit.
I wanted to point out to him that the way, bias is obtained,
has nothing to do at all with how the output tubes are wired.
So, the crux of it all is: tube life depends on other factors
than the way, the output is loaded.
Nothing more, nothing less.
I don't go around bashing cathode bias, nor any other possible
way to use tubes in a circuit, just pointing out that UL has *nothing*
to do whatsoever with biasing or tube life.
If I wasn't clear enough, I apologize for that.
>>The only valid point he has, IMO, is that the UCC *sounds*
>>better to him. No one can argue about preferences, and this
>>is his prerogative, of course.
>
>Hmm, I do not agree......as before stated, I find the UCC to be a design
>with greater possybilities, both construction and sound-vice.(A good
>and e a s y compromice I guess would be to lessen the work of the
>voltage drivers, by redusing the amount of local feedback.( I think that
>the quad 2 is a remarkable design in it's class.))
I may have been overreacting on his part, and I would welcome a
discussion of the merits of all mentioned circuitry.
Best regards,
Randy Nachtrieb
Nestor,
It is probably silly of me to respond to one who is both ignorant &
obviously wishes to remain so, but here goes.
First off, I am curious as to what sort of speakers(or source) you
have which are so demanding that they make virtually all amplifiers
clip, regardless of their power.
Secondly, as to lack of clipping, I believe you are confusing your
MC-275 with the modern Mac amps containing "Power Guard", which
alleviates clipping but squashes dynamics. The "dynamic compression"
problem is probably the reason why McIntosh contacted me several years
ago about the possibility of replacing "Power Guard" with a circuit of
mine known as "Dynamic Waveform Preservation".
Anyway, over the years I have owned atleat 5 or 6 MC-275's, as well
as modifying them for customers as far away from Los Angeles as Turkey.
While IMHO they are one of the better Macs, mainly because of their SS
full-wave bridge power supply, in stock configuration they are not
particularly good dynamic performers either audibly or on the bench. By
the same token however, neither is a stock Dyna.
Your assertion that UL amps create loudspeaker pumping is absurd,
the UL circuit contains no physical mechanism for this. Secondly, true
"pumping" (as opposed to motorboating, rumble or record warp) is of
sufficiently low frequency, that it is highly doubtful that one would
encounter it during the operation of a xfmr coupled tube amp.
Yes, my speakers (whose design is original) would most likely make
your Dynas & even your Macs sound quite good, it is however my
experience, that people with your mind set do not listen with their
ears. Rather their opinion is based upon some arbitrary agenda or
fetish. Perhaps in your case the obsession with the MC-275 was brought
about by the sub-conscious knowledge that your parents named you after
its designer, Mr. Nestorovic.
Kirk R. Patton
Man, this thread has made for some interesting reading!!! Most of the
incredulous postulating has been pretty well picked apart and clarified,
and especially enjoyed Russ Sadd's excellent post. An excellent
discussion of the pros and cons of distributed loading can be found in
the Mullard book "Tube Circuits for Audio Amplifiers" sold through a
number of catalogs -- it has some brief, concise theory, with graphs and
analysis of real circuits, and projects using UL, and discussion of
their performance. Fun book, easy to read.
I felt a need to throw in my own couple of cents regarding a couple of
points. One key factor in the implementation of the Unity Coupled
design, as outlined in Frank McIntosh's patent, is the quality factor of
the output transformer. In fact, his discussion of bifilar winding
techniques seems to be as important to him as the circuit itself. I
think most of us would agree that Mac's output trannies were, at the
very least, really nice. And the fact that they didn't liscense the
patent to anyone means that there aren't lots of examples of
poorly-implemented Unity Coupled designs kicking around for us to
criticize. The same is not true for UL. Poor coupling between the
windings of a Unity Coupled transformer would lead to a very poor phase
margin before overall NFB was even applied -- add an equally poor driver
stage and it's unlikely that the thing would work at all (except as a SW
transmitter). My point is that the reputation of the Unity Coupled
design rests solely on McIntosh -- who probably would have built a
pretty good amp of any design.
The biggest drawback of Unity Coupled is the very, very low gain of the
output stage as a result of the heavy application of local feedback.
This creates other engineering dilemmas, which McIntosh did a pretty
good job of dealing with, bit IMO some improvements could be made here.
In regards to the question of who manufactures transformer for Unity
Coupled designs other than McIntosh -- I know Plitron makes some
torroidial units that have excellent specs. I'd love to try them out,
but alas, so little time . . . so many amps . . .
Wanted to clear up some of the weird claims made regarding the Marantz
Model 2. I don't think that the inclusion of the triode switch shows
that Saul was insecure about the design of the amp. To quote the
instrucions included:
"This unit has been designed primarily as a 40 watt Ultralinear
Amplifier. Speaker systems of a lower power rating could easily be
damaged by unwanted pulses coming through at full power. To protect
these systems we have incorporated a switch which changes the output
connections to Triode operation limiting the power output to 22 or 23
Watts."
Also, I couldn't understand references made to the damping control being
included to make up for some of the "problems" with the sound of UL. I
used to own a pair of these, and as it happens, have another pair in my
shop for service. So this morning I hooked them up and listened to them
again, and memory does serve me correctly, the Triode/Ultralinear switch
doesn't really change the sound! Mainly just the difference in
headroom. But the tonal balance remains virtually the same. A little
more gain in UL mode. Then I played with the Variable Damping control.
This definately does change the sound, mainly the bass. But I can't
(nor could I when I owned them) see why anybody would want to use this
in a modern speaker system -- the change was definately NOT for the
better. Incidently, the rated damping factor of the Marantz model 2 is
20 in fixed-damping, and .5 to 5 in variable-damping mode. 20, by the
way, is excellent for a tube amp, especially one of this vintage.
Regarding reliability of UL amps -- if you're having problems with tube
failures, and the amp is in good shape, then it is the tubes, plain and
simple. There are A LOT of crappy EL34's out there, some new, some NOS.
There are others who post here who can cite reliability records of
various types, but getting a good quad of EL34's -- my personal
experience here -- has been very hit-and-miss. And I use several
suppliers and brands. When I do get a good set, they last for years in
my customer's Dynas and Marantzes -- and these are people who don't have
the slightest idea of how to diddle with the bias. If someone can give
me some input on this problem, I'd be more than happy to hear it.
McIntosh gear does bias the tubes rather conservatively, but believe me,
the stuff does break. And Mac owners are usually willing to spend
decent money to fix them. Failures in old Macs help put food in my
mouth . . .
Does the macfanatic have experience with UL gear other than Dyna? Ever
lived with a pair of Marantz 2's? I just got rid of a MC240 that had a
cost-no-object restoration, and yes, it did sound really, really, good.
But not like the Marantzes . . . . if you get a chance to try out a pair
in good condition, do it . . . I think you'll end up with a better
perspective on Ultralinear in general.
Best regards to all who posted, enjoyed it all tremendously . . .
Kirk Patton
John Byrns
In article <69tfav$3sb$1...@newsd-153.iap.bryant.webtv.net>,
gat...@webtv.net (Randy Nachtrieb) wrote:
> Secondly, as to lack of clipping, I believe you are confusing your
> MC-275 with the modern Mac amps containing "Power Guard", which
> alleviates clipping but squashes dynamics. The "dynamic compression"
> problem is probably the reason why McIntosh contacted me several years
> ago about the possibility of replacing "Power Guard" with a circuit of
> mine known as "Dynamic Waveform Preservation".
Randy,
If you don't squash the dynamics, as the "Power Guard" in the newer
McIntosh amps can do, what else is there left to do besides something like
clipping, or time dispersion, which will preserve most of the dynamics,
but not the wave form? How do you accomplish "Dynamic Waveform
Preservation"? It seems to me that if you want to reduce the peaks, you
only have two choices, scale the whole signal wave form, or distort the
wave form in some fashion. How does "Dynamic Waveform Preservation"
resolve this dilemma?
Regards,
John Byrns
Russ Sadd
Kirk R. Patton wrote in message <34C254...@sprintmail.com>...
>Man, this thread has made for some interesting reading!!! Most of the
>incredulous postulating has been pretty well picked apart and clarified,
>and especially enjoyed Russ Sadd's excellent post. An excellent
>discussion of the pros and cons of distributed loading can be found in
>the Mullard book "Tube Circuits for Audio Amplifiers" sold through a
>number of catalogs -- it has some brief, concise theory, with graphs and
>analysis of real circuits, and projects using UL, and discussion of
>their performance. Fun book, easy to read.
Yep, this is the biggest war of words since we were arguing over Golden
Dragon and their reliability! Thanks for the compliment too! Can I also
suggest GECs little booklet on amplifier design? It's been brought back into
print like the Williamson booklet and the Mullard one you mention.
Ultralinear considerations feature highly here.
>I felt a need to throw in my own couple of cents regarding a couple of
>points. One key factor in the implementation of the Unity Coupled
>design, as outlined in Frank McIntosh's patent, is the quality factor of
>the output transformer. In fact, his discussion of bifilar winding
>techniques seems to be as important to him as the circuit itself. I
>think most of us would agree that Mac's output trannies were, at the
>very least, really nice. And the fact that they didn't liscense the
>patent to anyone means that there aren't lots of examples of
>poorly-implemented Unity Coupled designs kicking around for us to
>criticize. The same is not true for UL. Poor coupling between the
>windings of a Unity Coupled transformer would lead to a very poor phase
>margin before overall NFB was even applied -- add an equally poor driver
>stage and it's unlikely that the thing would work at all (except as a SW
>transmitter). My point is that the reputation of the Unity Coupled
>design rests solely on McIntosh -- who probably would have built a
>pretty good amp of any design.
As you've seen from my post, there is a design rather simpler than the
Mac's. No bifilar winding present, honest! The author claims it gives a good
sound due to the tertiary feedback design, despite suffering from being
apallingly cheap! Quad's were standard silicon steel c-core I believe, and
the first generation were made of a much poorer grade steel. I get the
feeling that one of the strengths of this design was that it could work
miracles with a poorer output transformer.
>The biggest drawback of Unity Coupled is the very, very low gain of the
>output stage as a result of the heavy application of local feedback.
>This creates other engineering dilemmas, which McIntosh did a pretty
>good job of dealing with, bit IMO some improvements could be made here.
>In regards to the question of who manufactures transformer for Unity
>Coupled designs other than McIntosh -- I know Plitron makes some
>torroidial units that have excellent specs. I'd love to try them out,
>but alas, so little time . . . so many amps . . .
I prefer c-cores - this is just a personal thing, really. It might be a
British fixation - certainly seems to be if Partridge, Parmeko and others
tended towards them. Besides, the low gain can be offset quite easily by the
circuit - it isn't as bad as all that. I've got one that is driven straight
from a triode in Concertina, fed by a standard triode gain stage. I plan to
use a transformer for the phase-splitting in mine, which again makes it much
simpler to swing the load. In the next couple of months, I'll be able to
quote a model number and a company name for both the transformer used in the
article, and also a 6L6 version. Promise.
Russ Sadd,
ABS
Jim Kroger
> In article <885104787...@dejanews.com>, macfa...@hotmail.com wrote:
>
> > While I haven’t tested UL and Unity Coupled for stability and damping, I
> > think that pumping speaker cones are sufficient proof of amplifier
> > misbehavior. Further, strange audible high frequency artifacts are
> > undeniable proof of oscillation. In truth this phenomena can be
> > experienced with any tube amp. The difference lies in the fact that such
> > performance aberrations ARE part of normal UL operation as opposed to the
> > Mac circuit in which they are usually caused by a serious malfunction.
> > My MC30 does exhibit the pumping cone phenomena, but this is caused by
> > out-of-spec 40 year old parts and NOT by result of poor circuit design.
>
I'm curious what makes you think this is caused by old parts. My MC-30s
are completely original and I have no "cone pumping" if you mean visibly
seeing the cone move in and out. I do not have my phono set up so use CD
and FM.
Jim
---------------------------------------------
The author is singly responsible for the preceding content. It
expresses no opinion, position, or policy of the institutions
which have sponsored this here internet account.
---------------------------------------------
Anne Merete Driveklepp
Hello Sander deWaal (,and the rest of you...)
> >>All Randy's comments are relative to minimal standards of hifi.
> >>Even a properly designed SET has high distortion and low damping
> >>factor, as compared to modern standards.
(The SET might not be hifi?? What is it then?)
> >Yes,-compared to modern standards which always including loads of
> >corrective feeback.
>
> My very own tube amps produce a whopping .6% THD, at
> levels up to 15 Watt.
> Above that, they go up slowly up to 3 % at 18 watts.
> At average listening levels, it's not more than .3%.
> They have 10 dB of global feedback (shoot me now!!!)
> Compared to the average THD of a SET, which may be
> 5 % at decent listening levels, rather low.
Oops,-not nescessarily so! I believe that people who are serious about
SET-amps, as I am myselves, of course use appropriate speakers with
them! Speakers with a sensitivity that allows you to have the same (,or
even better) dynamic range, compared to the 'ordinary' 86-94 dB/W
stuff,- Now, this can be done, it's quite possible, but a little labourious
and,-eh,-space consuming & wery expensive....But you simply can
not evaluate the merits of the SET-amp, as a lone-standing apparatus!!
Every amp should be evaluated integrated with the proper use of it.
If not, the discussion becomes MEANINGLESS........ ;-)
Where do you have 5% from? This is a t a r g e t figure of design.
You might design a SET amp with a distortion of 5% at full output,
but as you dont listen to a 3, or 5W sine-wawe, but music, things
simply are different. Now, If you have 20 dB of headroom, the
average distortion is instead 0,5%, due to the simple circuit, that
behaves, distortion-vice in a wery simple manner. And we're talking
2. harmonic's here, with a decaying spectre....
> Now, I believe that the *nature* of the distortion is far more
> important than the figure alone.
> THD in itself is a rather meaningless thing , as long
> as it is under a 5 % or so.
I agree completely!!
> As I use PP, most residual distortion is odd order, so it
> HAS to have a lower THD than SET amps.
> All of this is no judgment of the sound quality, let alone
> the preference that one may have for one or the other sound,
> but just a clinical fact.
As I have shown, a rather meaningless (excuse me! :-) clinical fact.
All amps have to be evaluated in a context of use.If you have a Mac,
a Dyna, a Quad, or something else, the amp itselves dictate the context.
You should evaluate amps in an appropriate manner, but that should go for
any amp! Now, if you have a Powermac75, you probably have it to allow
yourselves to use anemic, kind'a modern speakers.You would evaluate it,
music-vice at power levels around 1-4W......To evaluate the 5W SET-amp
in a similar manner, you would simply have to be serious about speakers,
cause the power levels now is 80-300mW......
>
> >No, these (well-designed) amps have low distortion and good damping
> >factors,
> >considered the absence of feedback! Spec'-vise one could make a SET-amp
> >superior to most p-p pentode amps, with an addition that the former
(p-p)
> >type
> >already have: Just put another gain-stage in, and remove the gain,ex.
say
> >20dB,
> >with global feedback.....
>
> They have indeed, *considering the absence of feedback*.
> Doesn't contradict what Randy or I wrote earleir, IMO.
> And spec's don't mean much, if one doesn't like the sound
> of an amp.
> Why trying to convert a SET amp so, that it meets the
> specs of a PP penthode amp? (Not that that would be easy,
> far from that!)
As I have shown, it would be easy to exeed it. But my point here might be
a little
unclear, or misunderstood: I mentioned the extra gain stage/feedback thing,
just to
show the superiority of the SET design:It does not need it! It has amongst
other things
full bandwith, adequate damping factor, and low distortion.Now, 'any'
pentode p-p
amp just needs those 15-20dB of corrective feedback. And as the open-loop
power
gain seldom is of adequate power bandwith, this correction is frequency
dependent
too.....
> >OK, I'll try, although I have to put some lenght to it: I'll quote
> >myselves:
> >"The design IS infact more stable in itselves, but........."
> >Here I simply mean that the design,-as a constructing prinsiple is more
> >stable than the UL-design,-also as a constructing prinsiple....
> >Why is that? The purpose of both designs is to produse power into a
> >loadspeaker, which always has a complecs impedance..........
> >Now although the UL could be said to work with error reduction by means
> >of feedback to the #2-grids, this reduction is by no means l i n e a r ,
> >(because it is applied to G2), and we need some more help before the
design
>
> And here it is where we digress.
> If one takes a look at what really is happening with G2 feedback,
> one would see that the Ia-Ua characteristics of a beamtetrode or
> penthode are highly linearised.
> Meaning, that for a given point of operation on the characteristics,
> less distortion will be generated.
> In fact, less than in a triode, and far less than pure penthode.
I'm sorry my friend, but you are wrong here. Applying feedback to G2 is
not,
i repeat, not, linearizing the characteristics in a l i n e a r manner.If
you
consult any decent litterature, you could verify this for yourselves.Also,
the
valve characteristics are not h i g h l y linearized.What you end up with
is
curves, somewhat like a mix between pentode and triode, whith a slope on
all the 'G1 bias curves' that is c o n s t a n t l y varying, meaning a
changing
gm and Rp. They are never of the same value anywere on the load line, so to
speak. The net effect is distortion-vice better than a pentode, but the
pentode
has a more 'simple', level dependent distortion.And the triode is superior
to
both...(The best species has a nearly constant slope, meaning near constant
gm and Rp, and equal spacing, meaning constant mu. Also, check out the
GEC application papers for the KT66-77, and 88.and see for yourselves.)
I really think "Ultra Linear" ir a misnomer, because it is so far from
"Ultra", and
,to a lesser extenth, far from linear....It should be called something that
shows
the principle's 'tube type b l e n d i n g ' properties.
> Now, the only drawback is, that the 43% tap on the primary
> winding is highly frequency-dependent, so it's not possible
> to ensure these characteristics for the entire frequency spectrum.
> Not to metion stray inductance and -capacitance.
I partly agree.As I have tried to show, the picture is far more complex.And
speaking of the X-former, yes, both the UCC, and Circlotron-types are
'easier' to make well....But any ill-working UL-tranny should just be
ripped
out like a rotten teeth, and be replased with a desent one. It's possible.
> Now take the Quad/Mac approach: In order to make the output
> stage to work in class B, they added an extra winding on the
> output transformer, to overcome the crossover distortion that
> would occur without it.
I'm sorry, halve of this is incorrect: The Quad is a class 'A' design as
good as
any, NOT a 'B', or even 'AB'. You can NOT compare the two at all. The
cathode
feed vs. anode; ratio in the Mac, it is 1:1, in the Quad it is 1:9,...
something !!
The late Peter made this choise to cut down output impedance and to 'shape'
distortion, simultaneously. This is done in a far more elegant way than in
any
UL-amp, I believe.You should remember that the operation of the KT66 in
this
amp is quite intruiging,-it has been shown that this particular operation
of the
tubes reduses the 3.harmonics with 8 dB MORE than the figure one might
expect, by wieving feedback factor alone.....Contrary to the Mac, i belive
the
Quad really was designed for music! (Crossover distortion newer was an
issue
here.)
Now, the Mac is just a complete different story! Yes,- power was the issue.
But I strongly believe that with a different aproach....... :-))
> As long as we don't look at the frequency extremes, and drive
> the stage into overload, all will function relatively well.
> At extremes, and overload, all kinds of phaseshifts will occur,
> especially when global feedback is applied also, as do
> the Mac and the Quad.
> (And quite heavily, I might add).
NO no!! The Quad has not heavy feedback! It does not n e e d that much,
because it is a brilliant and wery well executed design!The Mac is also a
brilliant
design, but although good x-formers, not so brilliantly executed...
> This may well lead to oscillations, in case of a highly reactive load.
> So, both circuits do have their drawbacks, with the advantage of UCC
> that it behaves more civil in average signal handling, in not too
> reactive loads.
>
> >can be put to work. We need feedback from at least one preceeding gain
> >stage
> >to reduse distortion and output resistance, and by doing that we must
deal
> >seriously with phase shift, poles, and so on....A global feedback system
> >is simply more phrone to bad behaviour because of this, and the fact
that
> >you cannot foresee the complexity of the load..........
>
> Yup, one reason why pure penthode or triode stages are more stable
> than either UL, UCC or PCCC.
I n c o r r e c t , as earlier mentioned.The UCC is in i t s e l v e s
inherently
stable. As I said:
> >Now, the UCC-design is relying on cathode feedback, and infact COULD be
> >relying on that alone. The output resistanse, and distortion is redused
in
> >a linear manner, and with the proper selection of tubes (triodes?)
distortion
> >and output impedance would already (before global feedback) be superior.
> >As this happens with local feedback only, all we have been doing is to
> >'modulate' or 'rearrange' the original tube spec's, so we have an
> >inherently stable output stage..
>
> With triodes, I would agree to a certain point.
> It takes a healthy design of the output transformer, a qualification
> that the Quad trannies would meet earleir than the Mac's, IMO.
>
I'm sorry, I don't take your point? Both the Mac, and the Quad x-formers
are wery well executed.
>
> >I think I already have responded to that in the description of the
> >"execution of
> >the principle" (low, near B-biasing, a sloppy supply.) I might add: The
> >common
> >global feedback, and added leakage inductance to the output transformer.
> >(Shame on you Frank!) I guess that if the amp had been redesigned to
music
> >use, it just might................(Well,- guess that depends on taste
to,-)
>
> I still think that the UCC could provide as much pleasure
> as any other output stage, providing good trannies and
> a friendly load are used.
Yes, I fully agree with you!
> I could have been more clear about this in my original reply,
> though.
> Guess I was blinded by mr. MacFanatic's rhetorics..... :-)
>
> >>As you probably have noticed, I never said that UL-amps
> >>were without problems or sonic disadvantages.
> >>Except from the fact, that I use an amp with negative grid bias,
> >
> >Yess! Cathodebias is argueably plain silly! (Oops! shouldn't have said
> >that,-guess I'm in for some flaming.......Heck, the NG probably don't
> >notice anyway............ :-).
>
> I do! :-)
> And cathode bias isn't silly, I never said that.
No you didn't, but I do! It's s a f e but sonicvice ....huh,-!....
> mr. MacFanatic used UL as a tool to show that
> tube life would be shorter than in a UCC circuit.
> I wanted to point out to him that the way, bias is obtained,
> has nothing to do at all with how the output tubes are wired.
> So, the crux of it all is: tube life depends on other factors
> than the way, the output is loaded.
> Nothing more, nothing less.
> I don't go around bashing cathode bias, nor any other possible
> way to use tubes in a circuit, just pointing out that UL has *nothing*
> to do whatsoever with biasing or tube life.
> If I wasn't clear enough, I apologize for that.
>
> >>The only valid point he has, IMO, is that the UCC *sounds*
> >>better to him. No one can argue about preferences, and this
> >>is his prerogative, of course.
> >
> >Hmm, I do not agree......as before stated, I find the UCC to be a design
> >with greater possybilities, both construction and sound-vice.(A good
> >and e a s y compromice I guess would be to lessen the work of the
> >voltage drivers, by redusing the amount of local feedback.( I think that
> >the quad 2 is a remarkable design in it's class.))
>
> I may have been overreacting on his part, and I would welcome a
> discussion of the merits of all mentioned circuitry.
>
Well, this should be a start then. I did not notice that this initially was
some
debate over bias-curcuits and safety, linked with various amp-topics,-
Guess
I was a sloppy reader in the beginning, never to notice. I thought it was
about
some of the real interesting topics we've been dealing with here!
>
> Best regards,
> Sander deWaal
Best regards,
Torbjoern Lien
>
>
>
>
André Jute
First, I snipped a huge amount of very interesting stuff on why a Quad
II works like it does, because I am fascinated by what is left:
T Lien wrote:
> Hello Sander deWaal (,and the rest of you...)
>
> > >>All Randy's comments are relative to minimal standards of hifi.
> > Compared to the average THD of a SET, which may be
> > 5 % at decent listening levels, rather low.
> Oops,-not nescessarily so! I believe that people who are serious about
> SET-amps, as I am myselves, of course use appropriate speakers with
> them! Speakers with a sensitivity that allows you to have the same (,or
> even better) dynamic range, compared to the 'ordinary' 86-94 dB/W
> stuff,- Now, this can be done, it's quite possible, but a little labourious
> and,-eh,-space consuming & wery expensive....But you simply can
> not evaluate the merits of the SET-amp, as a lone-standing apparatus!!
>
> Every amp should be evaluated integrated with the proper use of it.
> If not, the discussion becomes MEANINGLESS........ ;-)
> Best regards,
> Torbjoern Lien
Whoever said SET amps are operated at 5% distortion either has an agenda
or has never heard one. The tubes and possibly a few amps may have been
specified at 5% distortion in days of yore, but modern designers know
the standards of perception are higher now, in fact have known since the
1940's. It is no big deal to design an SE amp with zero negative
feedback and 1% or 2% of distortion at 20W or 25W. But as others have
pointed out, that is not where you will play it. The point is in any
event not what a bunch of EEs can measure, but what a critical,
experienced listener can hear, and he hears less distortion in SET than
in anything else because SET sounds more like the real thing.
(Later this year I shall publish SE designs up to 80W, with which one
should be able to drive any reasonable speaker in any reasonable room.
Clearly, in everyday operation these amps will have vanishingly low
distortion.)
That said, I find these remarks a bit elitist:
>Speakers with a sensitivity that allows you to have the same (,or
> even better) dynamic range, compared to the 'ordinary' 86-94 dB/W
> stuff,- Now, this can be done, it's quite possible, but a little labourious
> and,-eh,-space consuming & wery expensive....
The politically correct code for "costs a lot money and is very
inconvenient", young master Lien, is "SET owners have to be truly
committed". (Where's Zip when we need him to straighten out some
foreigner so that he doesn't frighten off the punters with blunt
language?)
It is anyway bull that SET speakers have to be especially expensive and
space consuming. A little lateral thinking will soon convince anyone
that to drive cheap 85dB/W/m speakers, all you need do is build a bigger
SE amp or get a smaller room or stack the cheap speakers until you get
the sensitivity and SPL you want. Under 20W, say PSE 300B, does nicely
in 30 x 16 room on QUAD ESL-63, and putting two 300B or the heftier KR
VV per channel on the amp for PSE is cheaper than stacking stats or
buying a Tannoy Royal Westminster. I am not just cracking jokes: I
switch between Lowthers and electrostats at will with almost all my amps
except the smallest; for instance I often drive -63 (86dB) with Quad
II, and in a smaller room my standard setup used day in and day out was
12W QII/ESL63 until I started rolling my own.
There are a lot of unproven opinions in SET. (Including some of mine,
sure--but I back them with over forty years of concert going and writing
about music, a fair number of prototypes that never make it out of the
shop, and the advice of people who where there when SET was all there
was.)
One misconception, generally subsumed by the less thoughtful of the SE
trendies, is that a Lowther is 96dB to 100db sensitive in its box and
the room. It is utterly untrue. A Lowther, in almost any box but the
most space consuming and expensive, is an approximately 90db speaker
with what appears to be random peaks up to 100db. If you're going to
spend that much money and allocate that much space, you may as well go
for a decent horn speaker, a Tannoy or one of the American types.
Another is that twee fleapower amps are more "sensitive" to the soul of
the music than big brutes like the 300B and bigger. Well, even if you
make allowances for the shortcomings of Lowthers, even 2A3 (several
times the power of *real* fleapower amps) sound washed-out to me because
they don't have the power for decent dynamics and transient response
even when linked to 96-99dB sensitivity speakers. (Note that my
preferences are for unaccompanied voices and chamber music--for big
orchestral stuff the position gets worse.) I have recently started
wondering if there isn't some minimum level of dynamics and transient
response (for each kind of music, of course) that is commonly accepted,
or if it is a personal expectation, informed by experience. It is no
accident that the probably the most popular 2A3 amp ever is the PP Acro,
giving 7W...
(A corollary is: "everyone *knows* big SE amp *can't* be any good".
Crap. They have just never heard a really big SE amp, among other
reasons because output transformers for really big SE amps are very rare
and expensive.)
My feeling is that loudly promoted misconceptions like these do as much
harm to SET as those ludicrous claims of power requirements that speaker
manufacturers publish do to their credibility.
Time and again we find that some cheap makeshift brought in to save
money suddenly becomes an article of faith with people who have never
bothered to examine it objectively; that with a little time passing the
very shortcomings of the solution are touted as a virtue. (Which is
where we came in in the ultralinear thread from whose rib I carved this
new thread...)
Andre
Currently listening to Maria Denayova play Schubert Piano Sonatas
(Nimbus NI 1779, 3 discs) on 85db/W/m speakers driven to concert level
in a 4800 cu ft room by 18W of PSE 300B
Andre Jute
an...@indigo.ie
COMMUNICATION JUTE
--we support pages for music lovers, writers and audiophiles at
http://www.foundmark.com/ComJute/ComJuteF1.html
Sander deWaal
On 18 Jan 1998 23:48:34 GMT, "Anne Merete Driveklepp"
<mdri...@online.no> wrote:
>(The SET might not be hifi?? What is it then?)
AFAIK, HiFi means: High Fidelity, usually translated as
"accurate".
While I don't like the "accurate" sound of most of today's SS
gear, it doesn't mean that the meaning of HiFi has changed
over time.
If I had the speakers for it, I would perhaps enjoy SET's , too,
but as it is now, I can't drive my speakers to desired levels with
a 300B SET.
Maybe a 845 or 211.....a whole different ball game here! :-)
>Oops,-not nescessarily so! I believe that people who are serious about
>SET-amps, as I am myselves, of course use appropriate speakers with
>them! Speakers with a sensitivity that allows you to have the same (,or
>even better) dynamic range, compared to the 'ordinary' 86-94 dB/W
>stuff,- Now, this can be done, it's quite possible, but a little labourious
>and,-eh,-space consuming & wery expensive....But you simply can
>not evaluate the merits of the SET-amp, as a lone-standing apparatus!!
I agree, I like to see reproduced sound as a whole,
including source , amps and speakers.
Too bad I like my 86 dB/w/m speakers too much!
>Where do you have 5% from? This is a t a r g e t figure of design.
>You might design a SET amp with a distortion of 5% at full output,
>but as you dont listen to a 3, or 5W sine-wawe, but music, things
>simply are different. Now, If you have 20 dB of headroom, the
>average distortion is instead 0,5%, due to the simple circuit, that
>behaves, distortion-vice in a wery simple manner. And we're talking
>2. harmonic's here, with a decaying spectre....
I have those figures from my own experience (hey I used
to build SET's, too!), and reports from magazines.
If you're able to design a SET without global FB, with .5%
THD, you're certainly a better designer than those at AN,
Cary, AI, etc! :-)
2nd harmonics, can create a certain "aura" around instruments,
and can cause a huge soundstage that's not there in reality.
>As I have shown, a rather meaningless (excuse me! :-) clinical fact.
>All amps have to be evaluated in a context of use.If you have a Mac,
>a Dyna, a Quad, or something else, the amp itselves dictate the context.
>You should evaluate amps in an appropriate manner, but that should go for
>any amp! Now, if you have a Powermac75, you probably have it to allow
>yourselves to use anemic, kind'a modern speakers.You would evaluate it,
>music-vice at power levels around 1-4W......To evaluate the 5W SET-amp
>in a similar manner, you would simply have to be serious about speakers,
>cause the power levels now is 80-300mW......
Thus making it a hell of a design job............
But it can be done, I know.
>> They have indeed, *considering the absence of feedback*.
>> Doesn't contradict what Randy or I wrote earleir, IMO.
>> And spec's don't mean much, if one doesn't like the sound
>> of an amp.
>> Why trying to convert a SET amp so, that it meets the
>> specs of a PP penthode amp? (Not that that would be easy,
>> far from that!)
>
>As I have shown, it would be easy to exeed it. But my point here might be
>a little
>unclear, or misunderstood: I mentioned the extra gain stage/feedback thing,
>just to
>show the superiority of the SET design:It does not need it! It has amongst
>other things
>full bandwith, adequate damping factor, and low distortion.Now, 'any'
>pentode p-p
>amp just needs those 15-20dB of corrective feedback. And as the open-loop
>power
>gain seldom is of adequate power bandwith, this correction is frequency
>dependent
>too.....
Methinks that a PP triode stage offers more than a SET, in terms
of power and distortion.
It doesn't need 15...20 dB feedback, nor added gain, too.
(I stated that my amps use 10 dB feedback, I believe that more
than 12 dB does strange things to the sound. They also have
a bandwidth of over 100 KHz, with global feedback removed.....)
Due to the Miller effect, and the voltage source character, a triode
amp usually has a lower bandwidth than a comparable tetrode-
or pentode stage.
Again, it takes a *really* good design, to make any triode work
properly, and in fact, they're more critical about driving and
loading than tetrodes or penthodes are.
I guess my enthousiasm carried me away here.
The measurements I did, were with a 8 ohm resistor, not
a loudspeaker.
I believe you're right that a reactive load could ruin the
"linear" character of the stage.
>I'm sorry, halve of this is incorrect: The Quad is a class 'A' design as
>good as
>any, NOT a 'B', or even 'AB'. You can NOT compare the two at all. The
>cathode
>feed vs. anode; ratio in the Mac, it is 1:1, in the Quad it is 1:9,...
>something !!
I believe the Quad II is a AB design, and cathode biased as well.
The difference in C/A ratio windings could be the reason that I like
the sound of a Quad more than a Mac.
Also, that the Quad operates in AB, and is capable of driving complexe
loads better than a Mac, IMO.
>> As long as we don't look at the frequency extremes, and drive
>> the stage into overload, all will function relatively well.
>> At extremes, and overload, all kinds of phaseshifts will occur,
>> especially when global feedback is applied also, as do
>> the Mac and the Quad.
>> (And quite heavily, I might add).
>NO no!! The Quad has not heavy feedback! It does not n e e d that much,
>because it is a brilliant and wery well executed design!The Mac is also a
>brilliant
>design, but although good x-formers, not so brilliantly executed...
The Mac does.........
And the Quad as well, compared to your beloved SET... :-)
>I n c o r r e c t , as earlier mentioned.The UCC is in i t s e l v e s
>inherently
>stable. As I said:
Providing all goes well, matched tubes, symmetrical
driving, etc.
Things that our MacFanatic believes don't apply to the Mac......
>I'm sorry, I don't take your point? Both the Mac, and the Quad x-formers
>are wery well executed.
The Quad *amps* perform better, IMO, because of the transformer
(ratio, and overdimensioned compared to the Mac).
>> And cathode bias isn't silly, I never said that.
>
>No you didn't, but I do! It's s a f e but sonicvice ....huh,-!....
??? The Quad has cathode bias...........
>> I may have been overreacting on his part, and I would welcome a
>> discussion of the merits of all mentioned circuitry.
>>
>Well, this should be a start then. I did not notice that this initially was
>some
>debate over bias-curcuits and safety, linked with various amp-topics,-
>Guess
>I was a sloppy reader in the beginning, never to notice. I thought it was
>about
>some of the real interesting topics we've been dealing with here!
>>
It turned into an interesting discussion, IMO, and I'd like
to thank mr. Russ Sadd for his valuable contribution.
As well as all others who contributed, guess I learned
something along the way, too! :-)
Sander deWaal
On Mon, 19 Jan 1998 02:34:40 +0000, André Jute <an...@indigo.ie>
wrote:
>> > Compared to the average THD of a SET, which may be
>> > 5 % at decent listening levels, rather low.
>> Oops,-not nescessarily so! I believe that people who are serious about
>> SET-amps, as I am myselves, of course use appropriate speakers with
>> them! Speakers with a sensitivity that allows you to have the same (,or
>> even better) dynamic range, compared to the 'ordinary' 86-94 dB/W
>> stuff,- Now, this can be done, it's quite possible, but a little labourious
>> and,-eh,-space consuming & wery expensive....But you simply can
>> not evaluate the merits of the SET-amp, as a lone-standing apparatus!!
>>
>> Every amp should be evaluated integrated with the proper use of it.
>> If not, the discussion becomes MEANINGLESS........ ;-)
>Whoever said SET amps are operated at 5% distortion either has an agenda
>or has never heard one. The tubes and possibly a few amps may have been
>specified at 5% distortion in days of yore, but modern designers know
>the standards of perception are higher now, in fact have known since the
>1940's. It is no big deal to design an SE amp with zero negative
>feedback and 1% or 2% of distortion at 20W or 25W. But as others have
>pointed out, that is not where you will play it. The point is in any
>event not what a bunch of EEs can measure, but what a critical,
>experienced listener can hear, and he hears less distortion in SET than
>in anything else because SET sounds more like the real thing.
I wrote that, mr. Jute.
I based this on measurements I made on several 300B and
2A3 amps, as well as literature.
Further, it is my belief that a SET can add space to a recording,
and a certain "pleasuring round-off" of some digital recordings.
However, for LP and most CD's, I *prefer* PP amps.
I listened extensively to various SET -based systems, and
(almost) always, they were throwing a too big soundstage.
The exception was a Audion 300B SET on Ensemble PA-1
speakers.
>>Speakers with a sensitivity that allows you to have the same (,or
>> even better) dynamic range, compared to the 'ordinary' 86-94 dB/W
>> stuff,- Now, this can be done, it's quite possible, but a little labourious
>> and,-eh,-space consuming & wery expensive....
>
>The politically correct code for "costs a lot money and is very
>inconvenient", young master Lien, is "SET owners have to be truly
>committed". (Where's Zip when we need him to straighten out some
>foreigner so that he doesn't frighten off the punters with blunt
>language?)
Zip stated :"tubes are for boobs".
So much for comments from a salesman, who notably sells Jadis! :-)
>buying a Tannoy Royal Westminster. I am not just cracking jokes: I
>switch between Lowthers and electrostats at will with almost all my amps
>except the smallest; for instance I often drive -63 (86dB) with Quad
>II, and in a smaller room my standard setup used day in and day out was
>12W QII/ESL63 until I started rolling my own.
Agreed, I play a lot with 15 W in 86 dB/w/m (PP).
But I have a small listening room, and I rarely listen at levels
above 75 dB.
>(A corollary is: "everyone *knows* big SE amp *can't* be any good".
>Crap. They have just never heard a really big SE amp, among other
>reasons because output transformers for really big SE amps are very rare
>and expensive.)
I agree that a 50 W SET amp would be very good, in terms of
dynamics, bass control and spaciousness.
However, the more accurate approach of the PP triode amp
is more like I would hear it.
It can be done relatively inexpensive, too.
Steve Jones
Sander deWaal wrote:
>
> I have those figures from my own experience (hey I used
> to build SET's, too!), and reports from magazines.
> If you're able to design a SET without global FB, with .5%
> THD, you're certainly a better designer than those at AN,
> Cary, AI, etc! :-)
> 2nd harmonics, can create a certain "aura" around instruments,
> and can cause a huge soundstage that's not there in reality.
A few small caveats, here.
With sensitive speakers the 2nd harmonics are below an audible level.
But, if 'reality' is your bag (with stereo), how do you rationalize the
center 'phantom image'? In nature, we *never* hear a single source from
two independent sounds.
It's ALL an illusion, Sander! :-)
-Steve Jones
Steve Jones
André Jute wrote:
>
> One misconception, generally subsumed by the less thoughtful of the SE
> trendies, is that a Lowther is 96dB to 100db sensitive in its box and
> the room. It is utterly untrue.
As a dedicated follower of fashion (NOT!), which of the Lowthers are you
refering to:
PM6C
PM2C
PM6A
PM7A
PM2A
PM2A mkII
PM3A
PM4A
PM5A
DX2
DX3
DX4
(please note, these are only the 'full range' models)
And which cabinet and room did you measure yours in?
> A Lowther, in almost any box but the
> most space consuming and expensive, is an approximately 90db speaker
> with what appears to be random peaks up to 100db.
Bullshit. The mid-band sensitivity of my PM7As in Medallion II
enclosures, in room, is 100dB+. -3dB is at (around) 45Hz (at the
listening position) and at (around) 15kHz. This using a Ratshack SPL
meter and Stereophile test CD (and accounts for the 'around' note).
> If you're going to
> spend that much money and allocate that much space, you may as well go
> for a decent horn speaker, a Tannoy or one of the American types.
No bitch about Tannoys (GREAT speakers!). Which American speakers are
you refering to?
-Steve Jones
Bob C
Steve Jones wrote in message <34C378...@vcd.hp.com>...
>Bullshit. The mid-band sensitivity of my PM7As in Medallion II
>enclosures, in room, is 100dB+. -3dB is at (around) 45Hz (at the
>listening position) and at (around) 15kHz. This using a Ratshack SPL
>meter and Stereophile test CD (and accounts for the 'around' note).
>
>No bitch about Tannoys (GREAT speakers!). Which American speakers are
>you refering to?
>
>-Steve Jones
Steve;
You are correct about the sensitivity of lowthers, I OWN them also. He is
just repeating some info regarding bare drivers. This was in a German
Magazine. Please understand that you are dealing with "fantasy island" It is
dangerous when
"some people" give opinions on subjects they no nothing about.
Bob
Anne Merete Driveklepp
> >
> > André Jute wrote:
> >" That said, I find these remarks a bit elitist:"
Do you think I care about your findings ? Why bother, it's just
the laws of nature,-..... But then again,-.........
> >" ......................from whose rib I carved this new thread."..)
> >
> > Andre
(!!).......I guess the rest of us are in for some real changes.......
T.Lien
>
>
>
André Jute
Sander deWaal wrote:
> >Whoever said SET amps are operated at 5% distortion either has an agenda
> >or has never heard one. The tubes and possibly a few amps may have been
> >specified at 5% distortion in days of yore, but modern designers know
> >the standards of perception are higher now, in fact have known since the
> >1940's. It is no big deal to design an SE amp with zero negative
> >feedback and 1% or 2% of distortion at 20W or 25W. But as others have
> >pointed out, that is not where you will play it. The point is in any
> >event not what a bunch of EEs can measure, but what a critical,
> >experienced listener can hear, and he hears less distortion in SET than
> >in anything else because SET sounds more like the real thing.
> I wrote that, mr. Jute.
Whoa, Sander. We are on the same side, so we should be on first name
terms even when we disagree. And here I cannot see that we disagree too
much.
> I based this on measurements I made on several 300B and
> 2A3 amps, as well as literature.
Then the point T Lien and I made in different ways is valid: if your
measurements reflected actual use instead of some notional maximum
power, you should get more sensitive speakers or downsize your
expectation of SPL or start loving a new kind of music!
Maybe we should campaign for a more relevant measure of tube amps. I
noticed that Svetlana had gotten the point for their SV572-10/-30/-160
for which in Class A2 operation at their recommended operating points
(750V and 130-145mA) they quote the distortion at 1W as well as more
traditionally the maximum power at 5% distortion. The numbers at 1W are
respectively 0.05%, 0.45% and 0.68% (taken into 8ohm). Amazingly, they
neglect to do this for the SV572-3, which the more traditional among us
can instantly see as a tube in the 211/845 that is easier to use than
the other big transmitting tubes.
How much distortion is measured is irrelevant; vanishingly low
distortion is a melted sand agenda, nothing to do with tubes. How much
can be perceived or distinguished is what we should worry about. The
question is, what level of distortion is audible? I suggest that it is
likely that an experienced, critical listener may not know the sound is
distorted until it reaches 2% of THD.
> However, for LP and most CD's, I *prefer* PP amps.
> I listened extensively to various SET -based systems, and
> (almost) always, they were throwing a too big soundstage.
No such thing! <G> I keep my ESL 11ft appart, centre to centre (and I
know people who keep them even further apart). I sit over twenty feet
from them. I am a keen concert-goer and my favourite seats are six rows
back, just to one side of the conductor. *For me* 11ft is therefore
just enough to give the impression of a window on a lifesize soundstage
many feet behind the plane between the speakers. (The corollary test is
of course that piano sonata should emanate from a point source slightly
to the left of centre-stage. But of course you are right: SET does give
one a bigger soundstage than one can measure with a tape--so does
closing your eyes in a live concert in the Concertgebouw!)
> >>[[[From T Lien: ]]]Speakers with a sensitivity that allows you to have the same (,or
> >> even better) dynamic range, compared to the 'ordinary' 86-94 dB/W
> >> stuff,- Now, this can be done, it's quite possible, but a little labourious
> >> and,-eh,-space consuming & wery expensive....
> >
> >[[[From Andre: ]]]] The politically correct code for "costs a lot money and is very
> >inconvenient", young master Lien, is "SET owners have to be truly
> >committed". (Where's Zip when we need him to straighten out some
> >foreigner so that he doesn't frighten off the punters with blunt
> >language?)
>
> Zip stated :"tubes are for boobs".
> So much for comments from a salesman, who notably sells Jadis! :-)
I was referring to a notorious occasion about two or three years ago
when someone accused Zip of selling outrageously overpriced rubbish and
Zip, instead of in his usual manner going into the morals of this
person's mother, merely dismissed him with a remark something like, "The
amps I sell are only for the truly committed," and someone else (not me,
I swear) pointed out that this translated into English as, "You don't
have the money to be taken seriously."
> >(A corollary is: "everyone *knows* big SE amp *can't* be any good".
> >Crap. They have just never heard a really big SE amp, among other
> >reasons because output transformers for really big SE amps are very rare
> >and expensive.)
>
> I agree that a 50 W SET amp would be very good, in terms of
> dynamics, bass control and spaciousness.
> However, the more accurate approach of the PP triode amp
> is more like I would hear it.
> It can be done relatively inexpensive, too.
Well, we're talking here about amps to be build by people who, by
definition (otherwise they wouldn't know why the higher power is
necessary!), already have a very high-end setup, or to be sold to such
people. The amp would have to be at least as good as what was already
standing in their listening room. Therefore it would most likely be
expensive. At quite a low financial level, for the same power the
difference in cost between PP and SE OPT and IST becomes an irrelevantly
small percentage of the total cost. The decision between PP and SE would
not be influenced by cost but by personal preference.
Besides various types of SE amps, I have PP amps in pentode, triode
coupled and DHT triodes and love them all, so I am not making a case for
any particular choice here. (I don't want to be confused with
"Macfanatic"!--my mind is permanently open to persuasion.)
But "the more accurate approach of the PP triode amp" seems to me open
to interpretation. I know the engineers will tell us what they can
measure. But like all professional opinions, this one is liable to be
taken to absurd conclusions, indeed should be taken to absurd
conclusions to discover its limits of sense and sensibility. When we
pull back from the absurd edge, we discover that the measurements do not
tell everything that is important to the story, that there is a large
margin left to be decided by taste. My own opinion--emphasize
*opinion*--is that, when music is reproduced at concert hall SPL,
single-ended triodes give the closest approach to the live event of any
topology I have heard; at lower levels the SET advantage becomes even
more clearcut. Your *opinion* is that PP triodes do it better. I respect
that*, but would as little elevate it to a generality as my preference
for SE amps of adequate power for the speakers to be driven with the
music selected.
Finally, could I just, on hand of some private letters I received which
pointed out that I overlooked something, add a small bit to my previous
letter. There is an advantage to low-power amps, insofar as they
ruthlessly honour the KISS principle. All other things being equal,
fewer parts means clearer sound. I don't dispute that; I merely thought
it was axiomatic and therefore didn't say it.
Andre
*and indeed, when I finish with the SV572 in SE, intend to turn to their
use in PP
--
André Jute
Steve Jones wrote:
>
> AndrČ Jute wrote:
> >
> > One misconception, generally subsumed by the less thoughtful of the SE
> > trendies, is that a Lowther is 96dB to 100db sensitive in its box and
> > the room. It is utterly untrue.
>
Thanks. I would consider it an insult be accused of being fashionable.
>, which of the Lowthers are you
> refering to:
> PM6C
> PM2C
> PM6A
> PM7A
> PM2A
> PM2A mkII
> PM3A
> PM4A
> PM5A
> DX2
> DX3
> DX4
> (please note, these are only the 'full range' models)
Wonderful what you learn when you mix with experts. I didn't even know
there are some Lowthers that are not "full range" (insofar as you can
call a 14kHz speaker "full range"). Which are the non-"full range" ones,
Steve?
The poorest was the 6C, the best the 2A and 4A. We didn't test the DX
types.
As for the famous tests taken by a respected German magazine Klang &
Ton, we failed precisely to duplicate their results, perhaps because we
don't have an anechoic chamber (instead we climb 60ft up a fire tower
into still free air). Our results for Lowthers in free air showed the
same sort of spread but at a lower level. Considering differences in
environment, equipment and skill levels, this seemed to me entirely to
validate their results.
> And which cabinet and room did you measure yours in?
The factory Acousta was the least satisfactory of the standard type
cabinets we tried. An optimized Fidelio was by far the best. We used two
of my listening rooms (one smaller than the other but the bigger one
"harder"), plus the very much "harder" workshop in an attempt to match
the factory claims. For comparison, we measured ESL-63 as 85dB v the
factory claim of 86dB; we have several times had lifelong practitioners
of the art of noise measurement from the airport as our guests and they
too measure the -63 at 85dB.
> > A Lowther, in almost any box but the
> > most space consuming and expensive, is an approximately 90db speaker
> > with what appears to be random peaks up to 100db.
>
> Bullshit. The mid-band sensitivity of my PM7As in Medallion II
> enclosures, in room, is 100dB+. -3dB is at (around) 45Hz (at the
> listening position) and at (around) 15kHz. This using a Ratshack SPL
> meter and Stereophile test CD (and accounts for the 'around' note).
If you say so, Steve; I bow to your expertise. But we had a hard time
getting more than 97dB out of the well-broken-in PM6A (that we used for
a key reference because some of the others were new) in a whole variety
of cabinets. Since your Medallion boxes are new on the market, perhaps
you would care to give us the full table so we can see how flat they are
between 45Hz and 15kHz.
Let me rephrase my remark, which you characterize as "bullshit", in a
form more suited to your expectation: "In any box but the most
expensive, and in any room but the hardest and most sonically confusing,
the Lowther is a 95db speaker with random response variations of plus
*and* minus 5db, so that its response falls randomly between 90 and
100db over the approximate band 40c/s to 15kc/s." The meaning is exactly
the same as in my earlier post: I have just put it in the jargon speaker
people use. Do you still want to call it "bullshit"?
Or even more simply: "The Lowther response is too ragged to be made flat
by any box or room except at expense not justified by the intrinsic
sound quality of the speaker."
All the same, I'll use Lowthers until I can afford Tannoys. Then I'll
move up and forget I was ever so poor I had to use Lowthers.
My point wasn't to attack either the Lowther driver or its users--they
are themselves masters of the art of negative publicity, more of which
piles on every time they indulge in a flame war--but that the Lowther as
the second or third or fourth or fifth best horn (depending on your
geographic perspective) and the only one that is remotely "affordable",
is one of the *compromises* that those who insist on low power SE amps
must make.
> > If you're going to
> > spend that much money and allocate that much space, you may as well go
> > for a decent horn speaker, a Tannoy or one of the American types.
>
> No bitch about Tannoys (GREAT speakers!). Which American speakers are
> you refering to?
I meant the classic Altecs and the JBLs, the smaller cinema types, that
the lowpower American SE fanatics cream over. I have experience of
Tannoys but very little of these American types, so I'm going on
hearsay.
Andre
--
macfa...@hotmail.com
In article <69tfav$3sb$1...@newsd-153.iap.bryant.webtv.net>,
gat...@webtv.net (Randy Nachtrieb) wrote:
> First off, I am curious as to what sort of speakers(or source) you
> have which are so demanding that they make virtually all amplifiers
> clip, regardless of their power.
Just normal Paradigm loudspeakers, medium sensitivity.
> Secondly, as to lack of clipping, I believe you are confusing your
> MC-275 with the modern Mac amps containing "Power Guard", which
> alleviates clipping but squashes dynamics.
No sir. THe MC275 DOES clip but only when I am reaching SPLs of 110 dB or
more. I do my regular listening at 105 dB peaks and about 80 dB average.
Try the Mercury 1812 release with the MC275 the cannons will scare you to
death! Also, loud horn passages are reproduced full tilt without
compression or audible distortion. Try Sinatra-Basie on reprise.
The "dynamic compression"
> problem is probably the reason why McIntosh contacted me several years
> ago about the possibility of replacing "Power Guard" with a circuit of
> mine known as "Dynamic Waveform Preservation".
Apparently your circuit did not cut it since McIntosh is still using the
Power Guard. It愀 nice trying to compete with Sid Corderman but...
I also have a MC2205 power guard amp (my subwoofer amp) and let me tell
you that this solid state rig can outperform many highly regarded tube
amps. The Power Guard feature does work you know. It愀 not perfect but It
sure gets the job done and the massive physical package is very
impressive.
Your assertion that UL amps create loudspeaker pumping is absurd,
> the UL circuit contains no physical mechanism for this
I was referring to motorboating or low frequency pulses.
> Yes, my speakers (whose design is original) would most likely make
> your Dynas & even your Macs sound quite good, it is however my
> experience, that people with your mind set do not listen with their
> ears. Rather their opinion is based upon some arbitrary agenda or
> fetish. Perhaps in your case the obsession with the MC-275 was brought
> about by the sub-conscious knowledge that your parents named you after
> its designer, Mr. Nestorovic.
Ha ha ha, this is plain silly. Just imagine I said you were named after
Randall Smith the designer of the Mesa Boogie guitar amp!!!
you must be kicking yourself for getting rid of all those MC275s!
Regards,
Nestor J Salguero
Sander deWaal
On Mon, 19 Jan 1998 21:31:46 +0000, André Jute <an...@indigo.ie>
wrote:
>> I wrote that, mr. Jute.
>
>Whoa, Sander. We are on the same side, so we should be on first name
>terms even when we disagree. And here I cannot see that we disagree too
>much.
I'm just told by my mom to be polite to older people :-)
But I should have known, tube people are different......
>> I based this on measurements I made on several 300B and
>> 2A3 amps, as well as literature.
>
>Then the point T Lien and I made in different ways is valid: if your
>measurements reflected actual use instead of some notional maximum
>power, you should get more sensitive speakers or downsize your
>expectation of SPL or start loving a new kind of music!
The last suggestion is absolutely out of the question! :-)
I said before, that my speakers are 86 dB/w/m, and I love
them, so the only solution for me would be, to build something
around a 845, 211, SV572PSE, or 300B PSE.
I was condidering the VV 52B for a moment, but that beauty
retails for about $750 or so....... :-(
>How much distortion is measured is irrelevant; vanishingly low
>distortion is a melted sand agenda, nothing to do with tubes. How much
>can be perceived or distinguished is what we should worry about. The
>question is, what level of distortion is audible? I suggest that it is
>likely that an experienced, critical listener may not know the sound is
>distorted until it reaches 2% of THD.
I think that IMD is a far better way of judging an amp
than THD.
Elsewhere I said, that the maximum THD could probably
be 5%, providing that it contains not too much odd order components.
>No such thing! <G> I keep my ESL 11ft appart, centre to centre (and I
>know people who keep them even further apart). I sit over twenty feet
>from them. I am a keen concert-goer and my favourite seats are six rows
>back, just to one side of the conductor. *For me* 11ft is therefore
>just enough to give the impression of a window on a lifesize soundstage
>many feet behind the plane between the speakers. (The corollary test is
>of course that piano sonata should emanate from a point source slightly
>to the left of centre-stage. But of course you are right: SET does give
>one a bigger soundstage than one can measure with a tape--so does
>closing your eyes in a live concert in the Concertgebouw!)
That would depend on the spot one's sitting, eh? :-)
What I meant was the effect of a double bass of about
2 meter wide.......
>> Zip stated :"tubes are for boobs".
>> So much for comments from a salesman, who notably sells Jadis! :-)
>
>I was referring to a notorious occasion about two or three years ago
>when someone accused Zip of selling outrageously overpriced rubbish and
>Zip, instead of in his usual manner going into the morals of this
>person's mother, merely dismissed him with a remark something like, "The
>amps I sell are only for the truly committed," and someone else (not me,
>I swear) pointed out that this translated into English as, "You don't
>have the money to be taken seriously."
I 'm only polluting the RA# waters since early '97, so.....
But it sure sounds like Zip, if he is to carry SET amps next month,
they'll "blow away" everything he heard before *grin*
>Besides various types of SE amps, I have PP amps in pentode, triode
>coupled and DHT triodes and love them all, so I am not making a case for
>any particular choice here. (I don't want to be confused with
>"Macfanatic"!--my mind is permanently open to persuasion.)
So am I, when I'm finally off my hobby horse, and have
a discussion with more open-minded people :-)
>But "the more accurate approach of the PP triode amp" seems to me open
>to interpretation. I know the engineers will tell us what they can
>measure. But like all professional opinions, this one is liable to be
>taken to absurd conclusions, indeed should be taken to absurd
>conclusions to discover its limits of sense and sensibility. When we
>pull back from the absurd edge, we discover that the measurements do not
>tell everything that is important to the story, that there is a large
>margin left to be decided by taste. My own opinion--emphasize
>*opinion*--is that, when music is reproduced at concert hall SPL,
>single-ended triodes give the closest approach to the live event of any
>topology I have heard; at lower levels the SET advantage becomes even
>more clearcut. Your *opinion* is that PP triodes do it better. I respect
>that*, but would as little elevate it to a generality as my preference
>for SE amps of adequate power for the speakers to be driven with the
>music selected.
I'm sorry, I had a poor choice of words here.
I meant that I prefer the PP approach, partly because of my speakers
need a little more power than the average 300B SET provides, and:
possibly, I'm biased towards PP because I *know* they have lower
distortion (though predominantly odd order, which is like cursing in
church) ......
I have no experience with high power SET's, maybe they're
the thing for me......
Hmmmmmm, I must consider that VV 52 or SV 572 approach after all.
>Finally, could I just, on hand of some private letters I received which
>pointed out that I overlooked something, add a small bit to my previous
>letter. There is an advantage to low-power amps, insofar as they
>ruthlessly honour the KISS principle. All other things being equal,
>fewer parts means clearer sound. I don't dispute that; I merely thought
>it was axiomatic and therefore didn't say it.
I agree with that.
That's why I'm always fooling around with phase splitter designs,
and possibly the reason that I don't like complex designs like
ARC ..... :-)
>Andre
>*and indeed, when I finish with the SV572 in SE, intend to turn to their
>use in PP
Could you post the results here, please?
I sure woulkd welcome your comments on the differences
between the two!
Randy Nachtrieb
Hi Nestor,
According to the McIntosh people I spoke with, the intended reason
for Power Guard is actually to protect tweeters from burning out
(nonsense) & had nothing to do with sound quality, though they are quite
aware of the sonic disadvantages. Atleast when Bob Carver ripped it off,
he had the good sense to only put it in his PA amps, wth a defeat
switch. My circuit was only intended to improve sound quality & did not
coincide with their technical objective.
Ulta-linear operation does not promote motorboating.
Given the system you are running, the MC-275 is most likely a very
good choice sonically, as it does not sound like you are asking all that
much from it dynamically. Though IMHO you could do as well or better for
far less money.
The difference between our relative statements regarding names, is
that I am not infatuated with Mesa-Boogie amplifiers, it's merely a
coincidence that I Iive two blocks from their Hollywood shop, really,
honest :).
Yes, I do wish I still had atleast one of those MC-275's I got rid
of, I could use the extra three grand.
Regards,
Randy Nachtrieb
André Jute
Sander deWaal wrote:
> I think that IMD is a far better way of judging an amp
> than THD.
> Elsewhere I said, that the maximum THD could probably
> be 5%, providing that it contains not too much odd order components.
I'll go with IMD, but my case is that the 5% or whatever is not an
absolutely fixed number for all time, but is generally lowered over time
by increasing experience, not only for individuals but for the
audio-aware community in general, starting right on the street with the
boomboxers.
> >No such thing! <G> I keep my ESL 11ft appart, centre to centre (and I
> >know people who keep them even further apart). I sit over twenty feet
> >from them. I am a keen concert-goer and my favourite seats are six rows
> >back, just to one side of the conductor. *For me* 11ft is therefore
> >just enough to give the impression of a window on a lifesize soundstage
> >many feet behind the plane between the speakers. (The corollary test is
> >of course that piano sonata should emanate from a point source slightly
> >to the left of centre-stage. But of course you are right: SET does give
> >one a bigger soundstage than one can measure with a tape--so does
> >closing your eyes in a live concert in the Concertgebouw!)
>
> That would depend on the spot one's sitting, eh? :-)
> What I meant was the effect of a double bass of about
> 2 meter wide.......
Sure. I wrote "the Bechstein sounds about four times true size, as if
the pianist has a nine-foot reach" and struck it because I am fed up
with the spotty louts in the peanut gallery throwing bricks and calling
me a liar every time I use a colourful phrase, and the more idiotic of
our resident EEs demanding with menaces that I prove it with
measurements, and "Bob C" Deutschmann wittering xenophobically that I
am quoting *foreign* sources.
> >*and indeed, when I finish with the SV572 in SE, intend to turn to their
> >use in PP
>
> Could you post the results here, please?
> I sure woulkd welcome your comments on the differences
> between the two!
It'll be a while, Sander, because I have my amplifier book to get out,
and for the moment I am concentrating on the SV572 in SE, and we're
still testing the last of the 300B "Lundahl" protos (listening to the
high-power version with KR VV 300B blue glass right now), but yes, I
intend to publish the results. I'm publishing so much, soon I will have
nothing to sell and will have to beg on the street... However, if you're
in a hurry, Svetlana's netsite has a huge amount of information,
including some very clever circuits by Eric Barbour, David Wolpe and a
particularly nice circuit from Japan, and don't overlook that an 811 is
virtually a low-anode-voltage SV572, so that the circuits are for all
intents and purposes interchangeable, so also look for 811 circuits on
the Svetlana site; another Dutchman who comes here, Mattisj de Vries,
has a fascinating netsite with, among other goodies, an 811 circuit that
could easily be simplified to be less extravagant and also easier to
build in an SV572 version--but be sure you understand the drive
requirements of the 572 before you start making changes to published
circuits.
Andre
tom schlangen
hi Andre,
> As for the famous tests taken by a respected German magazine Klang &
> Ton,
BTW, i've subscribed that magazine since years. in the past they did
some reviews on tube amps, too. furthermore, years ago they had a
short series of articles describing restauration of vintage tube gear.
and lately they had a series written by Holger Stein about the
development of his modular preamp design.
unlike other `hifi' mags, K&T is not printed on high gloss paper, very
thin, and only comes out once a month, but i've learned a _lot_ from
it, especially on speaker design and DIY, and from the
`basics/principles' articles, too.
highly recommended if one is into DIY-speakers (from astonishing good
el cheapo to high endian).
tom
Steve Jones
André Jute wrote:
>
> Wonderful what you learn when you mix with experts.
I'm no expert, but I do know what I write about.
> I didn't even know
> there are some Lowthers that are not "full range" (insofar as you can
> call a 14kHz speaker "full range").
Which are those?
> Which are the non-"full range" ones,
> Steve?
C45 and C55 drivers. Might help if you knew what you were writing about.
> As for the famous tests taken by a respected German magazine Klang &
> Ton, we failed precisely to duplicate their results, perhaps because we
> don't have an anechoic chamber (instead we climb 60ft up a fire tower
> into still free air). Our results for Lowthers in free air showed the
> same sort of spread but at a lower level. Considering differences in
> environment, equipment and skill levels, this seemed to me entirely to
> validate their results.
Sure, for free air measurements. You listen to the raw drivers with your
head in the clouds, too?
> The factory Acousta was the least satisfactory of the standard type
> cabinets we tried. An optimized Fidelio was by far the best. We used two
> of my listening rooms (one smaller than the other but the bigger one
> "harder"), plus the very much "harder" workshop in an attempt to match
> the factory claims. For comparison, we measured ESL-63 as 85dB v the
> factory claim of 86dB; we have several times had lifelong practitioners
> of the art of noise measurement from the airport as our guests and they
> too measure the -63 at 85dB.
I agree with the above. What's the point of it?
> Let me rephrase my remark, which you characterize as "bullshit", in a
> form more suited to your expectation: "In any box but the most
> expensive, and in any room but the hardest and most sonically confusing,
> the Lowther is a 95db speaker with random response variations of plus
> *and* minus 5db, so that its response falls randomly between 90 and
> 100db over the approximate band 40c/s to 15kc/s." The meaning is exactly
> the same as in my earlier post: I have just put it in the jargon speaker
> people use. Do you still want to call it "bullshit"?
You're waffling there, Andre. Your prior comments were bullshit, and
your attempts at covering your ass are disingenuous (at best).
> Or even more simply: "The Lowther response is too ragged to be made flat
> by any box or room except at expense not justified by the intrinsic
> sound quality of the speaker."
"Simplify, but don't simplify too much" -A. Einstein
Wise words (his, not yours).
> All the same, I'll use Lowthers until I can afford Tannoys. Then I'll
> move up and forget I was ever so poor I had to use Lowthers.
Sorry if it's what the 'common' folks use. BTW, I wouldn't mind a pair
of Tannoys either!
> I have experience of
> Tannoys but very little of these American types, so I'm going on
> hearsay.
Sadly, like most everything you post.
-Steve Jones
Sander deWaal
On 21 Jan 1998 17:48:13 GMT, t.sch...@ndh.net (tom schlangen)
wrote:
>BTW, i've subscribed that magazine since years. in the past they did
>some reviews on tube amps, too. furthermore, years ago they had a
>short series of articles describing restauration of vintage tube gear.
>and lately they had a series written by Holger Stein about the
>development of his modular preamp design.
>
>unlike other `hifi' mags, K&T is not printed on high gloss paper, very
>thin, and only comes out once a month, but i've learned a _lot_ from
>it, especially on speaker design and DIY, and from the
>`basics/principles' articles, too.
>
>highly recommended if one is into DIY-speakers (from astonishing good
>el cheapo to high endian).
I agree.
It's one of the most versatile magazines when DIY
speakers are concerned.
They were also raving about the "Raphael" amps,
if I recall correctly. :-)
Strange thing is, how Andre Jute knows about this mag,
I believe Andre is from the US, right?
Lucian Rusu
By the way, how can you get a subscription to this magazine?
I lived in Vienna for a few years, have a number of issues of "Klang und
Ton" and I have also appreciated the no-nonsense approach of this revue.
Also, are there any other European magazines (ideally, publishing more
schematics and details of tube amps) worth mentioning?
Sander deWaal
On Thu, 22 Jan 1998 08:25:13 -0500, Lucian Rusu <ru...@nortel.ca>
wrote:
The Dutch magazine 'Audio & Techniek" occasionally
publices DIY items, with tubes also.
Of course, this mag is in Dutch :-)
Mail to : aud...@worldonline.nl , the editor is John
van der Sluys.
André Jute
Sander deWaal wrote:
> Strange thing is, how Andre Jute knows about this mag,
> I believe Andre is from the US, right?
I live in Ireland, Sander, because they give artists and inventors a
taxfree life.
I know about Klang und Ton because I have friends everywhere. I shall
review the Holger Stein's pre-amp when all the modules have arrived--the
bits I have (control module for volume and switching, active tube
amplification) are absolutely brilliant. I gave up work on my own
modular pre-amp when I saw the concept of his...
Andre
Kamaruddin
Sander deWaal wrote:
> On 21 Jan 1998 17:48:13 GMT, t.sch...@ndh.net (tom schlangen)
> wrote:
> I believe Andre is from the US, right?
>
> _
> Sander deWaal
> postm...@pegasus.demon.nl
He's from Ireland, right?
One Eye Jack