OTL Asylum

OTL, Output Transformerless Amplifier User Group.

Return to OTL Asylum


Message Sort: Post Order or Asylum Reverse Threaded

Circlotron with 12AT7, 6SN7, 6080 & feedback

77.43.10.3

Posted on May 14, 2009 at 09:17:46
With reference to my last post of March 28 2009 (comparison of Modified Futterman vs. Real Futterman by using the very same tubes 12AT7, 6SN7, 4X 6080), I have just found the corresponding (and missing) circlotron schematic.
It resembles somewhat the GRAAF GM20, but in this case there is a differntial DC feedback too.
With respect to the previous Futterman/Modified Futterman cases the preliminary LT Spice simulations of this specific topology looks very good (it exhibits almost no second harmonic at all and only a quite normal third harmonic level).
Due to DC coupling the amplitude and phase margins are excellent.
LT Spice model is available on request.

 

Hide full thread outline!
    ...
Circlotron schematic, posted on May 14, 2009 at 09:21:49



And this is the schematic.

 

RE: Circlotron schematic, posted on May 17, 2009 at 09:56:50
Lew
Audiophile

Posts: 10911
Location: Bethesda, Maryland
Joined: December 11, 2000
Thanks, Cellai. Can you estimate the output impedance with the feedback? Also, can you provide a larger more readable version of the schematic? I cannot tell what's going on at the input stage. Also, I think I see a coupling cap between the driver stage and the output stage. Does the driver stage have a constant current source? Would be nice. Ralph says that the 6080 is not as viable as the 6AS7 in his circlotron amplifiers. Finally, a Futterman with a triode output stage is not exactly a Futterman, since he used pentodes. Could call it (the previous design, not this one) a modified Futterman, I guess.

 

Dickie, D. P., Jr. & Macovski, A. (1954 June). “A Transformerless 25-Watt Amplifier for Conventional Loudspeak, posted on May 17, 2009 at 11:02:27
LineArrayNut
Audiophile

Posts: 1953
Location: Tenn.
Joined: December 10, 2002

Dickie / Macovski it's been posted b4



I *like* the loudness button!

 

RE: Dickie, D. P., Jr. & Macovski, A. (1954 June). “A Transformerless 25-Watt Amplifier for Conventional Loudspeak, posted on May 17, 2009 at 16:36:46
Lew
Audiophile

Posts: 10911
Location: Bethesda, Maryland
Joined: December 11, 2000
Altho I still cannot really see Cellai's schematic, the two (yours vs his) appear to be a bit different in the input stage. In any case, there is a lot not to like about the one you posted: too many gain stages, too many coupling caps, many NFB loops (if I am reading the schematic correctly), electrolytics in the signal path, etc. (What is the purpose of that 56K/40uF network between the plate of the second 12AT7 section and the cathode of the first section?) I realize you are not necessarily endorsing it but merely supplying a response to my question. Thanks for that.

 

Circlotron output impedance, posted on May 18, 2009 at 00:22:00
LT Spice model says 2.5 Ohm (with feedback).

Larger schematic and other answers have been sent to your mail.

 

Let me get my scanner up and going..., posted on May 18, 2009 at 08:00:47
LineArrayNut
Audiophile

Posts: 1953
Location: Tenn.
Joined: December 10, 2002
... moved my comp the other day...

and I'll post the 1212 designs totem pole and the M60MkII.2 schemas TV always wants - if that's ok w/ Ralph
I *like* the loudness button!

 

Lessee if a gif file will work., posted on May 19, 2009 at 09:39:19
Ralph
Manufacturer

Posts: 4769
Location: Minnesota
Joined: April 24, 2002



This is the M-60 MkII.3.

 

What a luxury!, posted on May 20, 2009 at 00:20:32
I am truly amazed.
So many premium tubes and power resistors for such a streamline topology (with relatively low gain and no feedback at all)!
I have never seen or touched one but, as a DYI builder (and now, let me say, an amateur non-professional designer), I think I could never afford such a luxury.
It looks a bit too expensive even for DIY people.
Just for sake of manufacturing economy I hope you will decide some day to explore the possibilities of a topology closer to this last published Circlotron (I am not jealous, I do not wish to enter this business and the credit will be yours). I am quite shure the absolute performance will not suffer.
As a matter of fact some feedback will be certainly beneficial to gain some bits of performance of your already excellent Circlotron amplifiers and some reduced tag price will certainly allow to 'not-so-rich' customers to purchase and listen to this elite OTL amplifiers.

My best wishes to all.

 

RE: What a luxury!, posted on May 20, 2009 at 08:09:00
Lew
Audiophile

Posts: 10911
Location: Bethesda, Maryland
Joined: December 11, 2000
Dear Luca, I am afraid I cannot agree with you, nor am I really sure why you think the Atma circuit is so much more expensive to build than, for example, the one you posted. Can you be more explicit about cost? Cost is much more determined by the purchasing decisions of the DIYer than by the circuit he is building. There are some real advantages to a fully balanced topology that I would not want to give up. Gain of the Atma circuit can always be increased by using higher transconductance tubes in the input voltage amplifier stage. (I, for example, am using 6900s, which are milspec versions of the 5687 but with even higher Gm.) Moreover, it is not difficult to add NFB to the Atma circuit. Ralph used to offer it as a switch-able option on his production ampifiers (and maybe he still does), and Kevin Covi has done it in a published schematic. Have you heard an Atma-sphere amplifier? I am really surprised at your negative reaction. With all respect, I also think it is a bit presumptuous for you to state categorically that a simpler circuit would sound just as good, unless you have done a comparison in your own system. Mods to the circuit, including changes to the dual-differential cascode and the elimination of the driver cathode-follower stage were discussed on the old Atma-sphere Owner's Group website. Several alternatives have been tried. I for one am using constant current source(s) in the cathode(s) of the driver stage tube(s). But the basic design is very sound. You could build it with 10-cent parts or with 10-dollar parts; it's your choice.

 

RE: What a luxury!, posted on May 20, 2009 at 08:14:23
Ralph
Manufacturer

Posts: 4769
Location: Minnesota
Joined: April 24, 2002
Years ago we used to sell a kit of the earlier version of this amp. It did rather well until we discovered that some people were selling kits they assembled as if they were factory assembled.

We've been avoiding negative feedback as it is detrimental to the overall sound. Certainly it limits the number of speakers that can be used with the amp (although there are plenty which work quite well with the M-60 and its smaller brother, the S-30).

What we have found is that if an amplifier with no feedback cannot work with a speaker, then it is impossible for that speaker to *ever* sound like real music- it may sound OK, like a good hifi, but it will never cross that threshold into true greatness. So we have limited ourselves to those speakers wherein true musical fidelity, true greatness, true music experience **is** possible.

For more information about this see

 

DIY Budget, posted on May 20, 2009 at 08:55:19
My personal approach is 'design to cost' which means, see what HW you can purchase at the local surplus stores (or on e-bay auctions), design a working circuit around it, get out the most from that. For the two OTL amplifiers I have posted so far (one in this section of the audioasylum with six tubes/channel, the other in the DIY section with 5 tubes/channel) I spent less than 400 Euros in total (including the tubes of course). I had chosen 6080 tubes only because they are cheap and readily available and most of the boxes had been hand crafted from the chassis of some damaged VHS recorders.
The real fun for me is cultivating this hobby using what I can purchase at low cost. In other words, if I had to use three double triode 6SN7 tubes just for the input differential amplifier I would give up. To me 6SN7 and its Russian counterpart 6H8C are exactly the same if they cost the same.
Sorry if I seemed somewhat 'negative', this was not my intention.

 

RE: DIY Budget, posted on May 20, 2009 at 09:00:30
Lew
Audiophile

Posts: 10911
Location: Bethesda, Maryland
Joined: December 11, 2000
No problem. Your goals are very different from mine. I feel that I am saving a great deal of money just by owning these amps and that upgrades and modifications to the circuit that I make myself are "cost-no-object", because the results are such a good bargain compared to what non-DIYers pay for "the next big thing".

 

NFB or not NFB, posted on May 20, 2009 at 09:27:24
Thank you very much for your clarifications.
I do not agree with you when you say that negative feedback is generally detrimental to overall sound but I respect your opinion.
I would rather think that some limited feedback (8-10 dB) is beneficial to get the best sound, even with those speakers that sound natural when matched to the right no-feedback amplifier.

 

RE: NFB or not NFB, posted on May 20, 2009 at 11:06:46
Ralph
Manufacturer

Posts: 4769
Location: Minnesota
Joined: April 24, 2002
In the old days we built amplifiers that had 8 db of feedback- switchable to zero. As soon as the feedback was turned on, the collapse of the soundstage, the added artificial sheen in the highs along with reduced impact, were all clearly audible on a wide range of speakers. It was the sort of thing that anyone without any training or expertise could hear- the sort of thing that you hear in 5 seconds. In other words, it was **very** audible.

Global negative feedback enhances the odd-ordered harmonics that the ear uses to gauge the loudness of sounds- we are very sensitive to less than 0.01% enhancements! General Electric proved this in the 1960s. So if you really want the system to sound relaxed, to have the same sort of ease with power that real music has, you have to get rid of negative feedback.

 

Scientific Method (invented by Galileo Galilei of Pisa - Italy, 17th century), posted on May 21, 2009 at 00:52:49
I do not know so much about old days, I would rather remind that we live in the 21st century now.

I have learned about tubes and tube applications not before than a few years ago (and only because a colleague of mine is very fond of them).

At the time of my University studies I regret to say that tubes had already been discontinued.

Unfortunately time goes by swiftly, but nowadays you do not have to actually build all possible tube circuits to test their relative performance, there is Spice simulator that works for you.

I work for an Aerospace firm (in the past I had also worked as Applications Engineer for Tektronix, the well known scope manufacturer); in my company we design high speed digital and analog equipment for telecommunication applications and all electrical designers (junior & senior) say 'Spice does not lie'.

So I have tried to put to the Spice test your statements (and Spice appears to me as the only viable approximation of true scientific method).

And as I have only available the Spice model of my Circlotron I have used this one. I have chosen to compare the harmonic distortion of this particular Circlotron, with and without feedback, at a current level of about 1.4A @ 8 Ohm (350 mA/output triode, which is probabily a condition on the onset of clipping/compression for middle aged 6080 tubes).

These are the LT Spice results expressed in dBc @ 1KHz(-40 dBc = 1% distortion, -60 dBc = 0.1% distortion, -80 dBc = 0.01% distortion).

NO FEEDBACK: 2 KHz -85 dBc, 3 KHz -25 dBc.

WITH FEEDBACK: 2 KHz -35 dBc, 3 KHz -50 dBc.

In other words, second harmonic is much better without feedback than with feedback but if you look at the third harmonic level it is much worst without feedback (I would say intolerable, -25 dBc can not be considered hi-fi by audiophiles).

We can certainly agree that such results do not necessarily apply to other Circlotrons but my message is: be careful, if you want to convince somebody of your points you must be much much more specific.

All amplifiers exhibit vanishingly small distortion at low level, nobody could claim the superiority of an amplifier based on low level listening sessions only. Your statements claim the subjective superiority of OTL-NOFEEDBACK without enough supporting evidence.

If the above results applied to what you said before('as the feedback was turned on, the collapse of the soundstage') I might rather argue that based on Spice simulations it could have not been third harmonic to destroy the soundstage but the second harmonic.

And this is the reason why I will never accept at face value the opinions of whoever 'king of the jungle' I meet.

The case of Galileo Galilei is always present in my mind.

Thank you for your time and my best wishes to all.

P.S. LT Spice model of my Circlotron is available on request - Every audiophile must be put in the positon of independently checking the above findings, otherwise we would abjure the scientific method!

 

RE: Scientific Method (invented by Galileo Galilei of Pisa - Italy, 17th century), posted on May 21, 2009 at 10:18:45
Lew
Audiophile

Posts: 10911
Location: Bethesda, Maryland
Joined: December 11, 2000
Talking of scientific method: (1) Since Ralph was referring to the performance of his own amplifiers, perhaps it would behoove you to create a Spice model of an Atma-sphere amp and test that before making any claims. (2) I cannot recall how much current is used with the 6AS7s in a typical Atma amp, but I am pretty certain it is less/lower than 350mA per triode. In fact, I think it is more like 150mA per triode or 75mA per each section of a dual-section 6AS7. (I cannot say because my amps use 6C33C triodes.) (3) Results could vary depending upon how and how much feedback is applied.

If you search the internet for opinions on the sound of Atma-sphere amps which are generally used and preferred without any NFB, I don't think you will read any subjective opinions which are consistent with the notion that the amps produce significant 3rd order harmonic distortion. Further, since cumulative experience in high end audio suggests that simple measurements of harmonic distortion, especially using a computer model, are at best inadequate to predict listening pleasure, you need to add listening tests before drawing any further grandiose conclusions. Galileo actually did experiments in the real world; he did not rely on predictive models. Finally, if you were to audition an Atma amp and find that you preferred the sound with NFB present, that's cool too.

 

RE: Scientific Method (invented by Galileo Galilei of Pisa - Italy, 17th century), posted on May 21, 2009 at 21:22:07
Tre'
Industry Professional

Posts: 17260
Location: So. Cal.
Joined: February 9, 2002
"at a current level of about 1.4A @ 8 Ohm (350 mA/output triode"

I think cellai@space.it is talking about driving the amp (tubes) to this level. He's not talking about idle current. He's talking about where the current reaches at the peak when the test signal is applied. 1.4amps total, 350ma. per tube (two triodes in each tube).

With each of the 8 triodes (two per tube) on one side of the push pull at 175ma. you have 1.4amps across the speaker because the other set of tubes would be at zero (cutoff or very close to it).

If there is enough drive signal to drive each triode from 75ma. all the way up to 175ma. then there is more than enough drive signal on the other set of triodes to drive them from 75ma. (idle current)to zero ma. each.

This is what I've been trying to explain since the days of the ASOG, once the current in one set of tubes doubles from idle (ie; 75ma to 150ma. each) the current in the other set of tubes goes to zero (assuming linear tubes) and you are out of Class A.

1.4amps flowing across 8 ohms is 7.84 watts RMS. (15.68 peak watts)

You have to drive an amp to test it for harmonic distortion. Whether for real or in a simulator.

Tre'
Have Fun and Enjoy the Music
"Still Working the Problem"

 

Yes, I meant 1.4 A peak., posted on May 21, 2009 at 23:15:39
Sorry, I did not make myself clear, I meant 1.4 A peak.
Each triode of a 6080 could likely exceed even 500 mA peak, but only in pulse mode, that is, for a very short time (let's say some tens milliseconds). I say this because during an integration test a mishap had occurred and I witnessed an ultra rapid fuse (500 mA) in series with a 6080 anode to blow. Unfortunately high currents, even within anode power dissipation limits, greatly reduce reliability and tube life.

Thank you for your clarification.

 

Scientific method and computer models, posted on May 22, 2009 at 00:38:39
Maybe in the future I could decide to create a Spice model of the Atma Sphere amplifier. In any case I will never publish any finding without prior approval of the manufacturer; it is not my business.

By the way, I am quite shure that the Atma Sphere amplifiers are highly respected because they sound excellent (and not because they claim some ultra low harmonic distortion). This is not the point.

The point is that if you claim that no-feedback is ALWAYS better than with-feedback you MUST PRODUCE scientific supporting evidence.

Ralph says: "It was the sort of thing that anyone without any training or expertise could hear- the sort of thing that you hear in 5 seconds. In other words, it was **very** audible" to me".

There are a lot of people all around the world that claim to speak to the deads, listen to the aliens and/or even smell paradise flavours when they talk with God. They generally fall into three categories: Prophets, Saints and mads.

If you abjure science everything becomes a matter of faith.

I understand that a lot of asylum inmates have faith in some more or less esoteric audio thing (OTL, OFC cables, vynil, special caps, dampening rings, FET, Bipolars, etc.).

Unfortunately I do not believe in any particular esoteric audio thing, I simply love music and all the 'economic' means for getting a subjectively pleasant sound reproduction and/or subjectively positive experience (for example my son loves guitar distortion and he likes so much amplifiers with significat odd order harmonics).

However when you decide to stay with Science, as 'scientific' evidence supporting your statements you may decide to use a sort of 'wine tasting' panel (where a group of named individuals with 'golden ears' convene at a laboratory to perform several blind listening sessions under controlled conditions and according to a pre-defined protocol), you may choose to study the correlation of electrical measurements (harmonic distortion, transient harmonic distortion, wavelet analysis (used in mainly sismology and medicine)) or the like.

If you do not produce scientific evidence you can only say:
"All Customers of Manufacturer X products believe that 'technical solution y' greatly enhance their listening pleasure. 'Technical solution y', that produces the reduction of correlated 'measurement z1, z2, z3', etc., is believed to influence the subjective perception of 'parameter W1, W2, W3, etc.'".

So that any informed potential Customer or amateur is free to decide and if he wants to enter the circle of the believers.

My best wishes to all.

 

RE: Scientific Method (invented by Galileo Galilei of Pisa - Italy, 17th century), posted on May 22, 2009 at 07:21:16
Lew
Audiophile

Posts: 10911
Location: Bethesda, Maryland
Joined: December 11, 2000
Tre', you get no argument from me on what you say. Paul Speltz made some interesting measurements of what goes on in the circlotron when the current in one phase exceeds the limit; you might be able to find it if the ASOG archives are still accessible. I hesitate to try to recapitulate his hypothesis.
Cellai, Please stop draping yourself in the holy grail that is "science". I am a scientist by profession, too, albeit a molecular biologist, not an EE. No thoughtful person would disagree with what you say about the audio "faith". There are usually two polarized sides in debates about audio measurements: (1) engineers who believe ONLY in measurements, vs (2) subjectivists who believe ONLY in what they hear. I am somewhere in between, because I do not believe we know exactly what to measure that is a good correlate of the sound of "reality". Lets stick to the specifics. I only claim that your measurements of your circuit do not necessarily apply to the Atma-sphere circuit, which has a very different input and driver stage. Even if there were a correlation, I do not accept SPICE as the arbiter of what sounds good. SPICE and other measurements of different kinds plus listening tests, with the latter taking precedence, would be my choice for judging amplifiers. You have done one experiment; you should know your results would not get published in a reputable journal without additional supporting evidence, if that was the hypothetical goal. I do not disagree with your point that Ralph made a broad generalization about NFB that may not be true in all cases for all amplifiers. I think he meant to refer only to his products.

 

Galileo?, posted on May 22, 2009 at 08:26:59
Lew
Audiophile

Posts: 10911
Location: Bethesda, Maryland
Joined: December 11, 2000
By the way, while I admire Galileo greatly, it was bothering me a bit to credit him entirely for originating the scientific method. I was thinking of Copernicus, Leonardo da Vinci, and others who went before him. A Google search on "scientific method" gave the following information:

The Book of Optics (Arabic: Kitab al-Manazir‎; Latin: De Aspectibus or Opticae Thesaurus: Alhazeni Arabis) was a seven-volume treatise on optics, physics, mathematics, anatomy and psychology written by the Iraqi Muslim scientist Ibn al-Haytham (in Europe Latinized as Alhacen or Alhazen) in 1011–21, when he was under house arrest in Cairo, Egypt.
The book had an important influence on the development of optics, as it laid the foundations for modern physical optics after drastically transforming the way in which light and vision had been understood, and on science in general with its introduction of the experimental scientific method. Ibn al-Haytham has been called the "father of modern optics",[1] the "pioneer of the modern scientific method,"[2] and the founder of experimental physics,[3] and for these reasons he has been described as the "first scientist."[4]

So perhaps Ibn Alhazen should be your reference. Apparently, like Galileo, he was persecuted for his work.

 

RE: Scientific method and computer models, posted on May 27, 2009 at 13:59:54
Ralph
Manufacturer

Posts: 4769
Location: Minnesota
Joined: April 24, 2002
We have had the distortion of the amplifier analyzed several times over the years and as one might expect, we actually have distortion analyzers on site (although I don't put a lot of stock in them though as they tell you very little about the amp's behavior with real-world signals!). As the amp is fully differential, it did not make a lot of 2nd harmonic, as expected. The distortion at full power is primarily third harmonic, about a quarter of that seen in SETs at full power. The 4th (also considered a low order, like the 2nd and 3rd) is negligable. Overall the amp is capable of less than 1-3% THD at full power if properly set up.

It is the nature of the lower ordered harmonics to 'color' the presentation of an amplifier. If these are kept to a low level, the amp will be perceived as 'transparent' and uncolored.

It is the higher orders (5th and above) that are the major concern where loudness cues lie (as shown by General Electric back in the 1960s). These must not be enhanced by the amplifier if the system is to lack loudness cues at high volume. So far the Spice analysis described has not shed any light on this.

It is my opinion that it is better to work with the actual hardware rather than a simulation. For one thing, it can be analyzed by real-world analyzers.

 

RE: Circlotron schematic, posted on April 13, 2016 at 03:57:45
Posts: 1
Joined: April 13, 2016
Hi dude, I have 20 6as7 and wanna try your amplifier. May I have a larger schematic , since the size not let me read componente values?
Best regards,

Odilon Carvalho

 

RE: Circlotron schematic, posted on April 13, 2016 at 09:16:30
I am amazed to see that some old posts of mine (with my old moniker) resurface from time to time.
I will do my best to find my old notes of the time (this schematic was modified as a result of sims) and to share them.

Just in case give also a look here.

http://www.audioasylum.com/cgi/vt.mpl?f=tubediy&m=245607

 

Page processed in 0.032 seconds.