• WARNING: Tube/Valve amplifiers use potentially LETHAL HIGH VOLTAGES.
    Building, troubleshooting and testing of these amplifiers should only be
    performed by someone who is thoroughly familiar with
    the safety precautions around high voltages.

Operating Points of Fi 2A3 / 3 Tube 3 Watt Direct Coupled Amplifier (Lipman)

Found this schematic in a post.

852385cf-d118-4916-a824-06ca0d703bea-jpeg.840323
This was from me :)
Yes, I reverse engineered a Don Garber amp, generating this schematic, and found it matched with Lipman. FYI - I recently painted a chassis which will be used to build my version of this classic.
 

Hearinspace

Member
Paid Member
2008-06-03 5:18 am
Thanks for posting the link. I was wondering about a possible relationship with the Loftin-White circuit and had started looking around for more information but there was such variety in circuits posted online that I began to think "Loftin-White" was being used as a general synonym for direct coupling.
Some implied that capacitive or R/C coupling of the two cathodes was part of what it means to be a true L-W circuit. I get the sense from reading the article you posted and also articles from 1930 issues of Radio News (see January through July here) that it is mainly the elimination of any capacitive coupling by setting bias voltages with resistance. Would that be its defining characteristic?

Another question that comes out of reading about the Loftin-White circuit is the emphasis on stability frequently mentioned together with its immunity to the destructive effects of pulling the input tube. Based in large part on what I learned from the Paraglows years ago, I have built a number of directly coupled circuits and have never found stability to be a problem ( Nor have irresistable impulses to pull a tube in the middle of Beethoven's 5th! : )

Thank you !
 
Last edited:

Hearinspace

Member
Paid Member
2008-06-03 5:18 am
My fault! I knew I should have thought about the wording a little more but wasn't sure how to do it without it getting cumbersome and I was in a hurry.
I meant the difference between the drawing you posted and the Lipman. R-5 in Lipman's is in the same position as R7 in yours. R-7 in Lipman's doesn't have an equivalent in yours, making it (in my amateurish view) fundamentally different. My thinking is that if yours is a part for part representation of the Fi then the Fi isn't a Lipman.
I drive my friends nuts like this but I'm not good at understanding / categorizing complex things and have to make these distinctions in order to stay sane. Sorry!
 
Thanks for posting the link. I was wondering about a possible relationship with the Loftin-White circuit and had started looking around for more information but there was such variety in circuits posted online that I began to think "Loftin-White" was being used as a general synonym for direct coupling.
Some implied that capacitive or R/C coupling of the two cathodes was part of what it means to be a true L-W circuit. I get the sense from reading the article you posted and also articles from 1930 issues of Radio News (see January through July here) that it is mainly the elimination of any capacitive coupling by setting bias voltages with resistance. Would that be its defining characteristic?

Another question that comes out of reading about the Loftin-White circuit is the emphasis on stability frequently mentioned together with its immunity to the destructive effects of pulling the input tube. Based in large part on what I learned from the Paraglows years ago, I have built a number of directly coupled circuits and have never found stability to be a problem ( Nor have irresistable impulses to pull a tube in the middle of Beethoven's 5th! : )

Thank you !
Like the WE Model 91, the term Loftin-White has been diluted to mean little more than direct-coupled single ended. (It happens everywhere. Any alcoholic drink in a conical glass has become a "martini" ...)

I do think the multi-tapped resistor across the power supply is the most distinctive feature. The Lipman circuit abandons three of the taps (large driver cathode resistor with negative supply, screen voltage tap, and driver B+ greater than cathode voltage for a higher plate load) so it's not really the same thing. Only the tap to the 2A3 cathode remains.

That resistor chain does improve the stability of bias points as the tubes age, or are replaced with similar-but-different tubes. This has always been the issue with direct coupling. It was a very clever idea at the time. But it wastes power - today we do it with trimmers and servo circuits.
 
jdrouin

Sorry for the late reply, am on the other side of the world, so your awake, am asleep and vice-versa. Compression is merely a term to describe the bunching of the curves to the right of the operating point. If you look at the right side curves, you'll see them getting closer together and to the left of the operating point, further apart. Now if the curves were evenly spaced, then you would have no distortion. However in our case as the musical signal moves along the loadline, left and right, it for a better term compresses, as the signal becomes more negative and, expands as it becomes more positive, hence the generation of distortion and in the case of tubes mostly second harmonic distortion. Where it becomes particularly bad is when the Vgk curves rapidly change curvature when operating currents are low. You'll see below 0.2mA for the 6SF5 the curves bend greatly and above 0.5mA they straighten and the spacing becomes more uniform. So when choosing an operating point, best to avoid that non-linear region. That is the standard engineering approach and quite sound. Now in the black arts of tube design, this added distortion maybe used to advantage as the 6SF5 inverts the signal (and its distortion) being a common cathode stage. Now the 2A3 generates distortion as well. The trick is to get an input tube that has a distortion profile matching the output tube's. By adding the inverted distortion to the distortion, hey presto they cancel, at least in theory. Perhaps this is what Don Garber found with the 6SF5. The 6SF5 is similar to the 12AX7, 6SL7. 12BZ7 tubes in distortion profile and I've had no luck with them using a 2A3. Mine you I haven't tried every combination and no doubt someone on diyaudio will pull a rabbit out of a hat. I can say that the 12AT7 and the 6J6 worked this magic with the 2A3. Anyway, enough from me, hope this helps. For what its worth, I have never been able to make Garber's amp design work, always thought there were deliberate mistakes in the component values to throw copiers off.
Ejam, thank you for that considered reply. I've understood from the beginning that we want to design so that the space between the grid lines along the load line are even. However, I'd never heard the term compression used to describe the non-linear areas even though I've been doing this hobby for several years now, so thank you. It's also interesting to think about using the distortion generated in the input and output stages as a mutually canceling dynamic.

Last week I breadboarded a 2A3 amp using 1/2 a 6SL7 as a simple voltage amp in each channel. It sounds surprisingly good. If I'm able to find another 6SL7 cheaply enough, I'll try an SRPP input stage.

Otherwise, I'm going to try breadboarding the Lipman/Fi circuit using a single rectifier tube with separate power supply rails in each channel, like Garber did on the stereo version. I'll use minimum 500V caps since the B+ will need to be close to 450V. However, I don't currently have three 22uF caps available per channel, as Garber/Lipman used. Do we think it would be a problem if I use a 15uF, 47uF, and 100uF in each channel?
 
Ejam, thank you for that considered reply. I've understood from the beginning that we want to design so that the space between the grid lines along the load line are even. However, I'd never heard the term compression used to describe the non-linear areas even though I've been doing this hobby for several years now, so thank you. It's also interesting to think about using the distortion generated in the input and output stages as a mutually canceling dynamic.

Last week I breadboarded a 2A3 amp using 1/2 a 6SL7 as a simple voltage amp in each channel. It sounds surprisingly good. If I'm able to find another 6SL7 cheaply enough, I'll try an SRPP input stage.

Otherwise, I'm going to try breadboarding the Lipman/Fi circuit using a single rectifier tube with separate power supply rails in each channel, like Garber did on the stereo version. I'll use minimum 500V caps since the B+ will need to be close to 450V. However, I don't currently have three 22uF caps available per channel, as Garber/Lipman used. Do we think it would be a problem if I use a 15uF, 47uF, and 100uF in each channel?
Jdrouin

Thanks for your reply. Just to qualify, when I said I had no luck with the 12AX7 etc, I was referring to a direct coupled circuit. Personally, the best results I had was using a 5695 in a SRPP, LC coupled (grid choke) to a 2A3. I started off with Gordon Rankin's Baby Ongaku and modified it by replacing the 12AT7 with the 5695 and the RC coupling with LC. Very nice result on my system. Regarding the 6SL7, good tube in my experience. However, I usually like more that 1mA driving the 2A3 hence my 5mA for the 5696 after listening. Some say 20mA is a minimum, and you get into that bigger is better thing. Remember audio is a recipe not a race.

BTW, I reckon Garber went for the 22uF caps as a sonic choice. A conversation I had with him 20 or more years ago was he wanted a fast power supply by using smaller caps (what ever that means). I suspect that he used 22uF caps as they don't present current in-rush problems for your tube rectifier. Try your combination and if your happy with it, problem solved. The 22uF 630VDC Solen are not too expensive (around $12 each) or I use the ASC 20uF 440VAC oil caps for ($15 each) if your not happy.

Just before I sign off, a word of advice from 30 years of doing this, trust your ears and not others and when you make something that sounds good, sit back, relax and enjoy the music, otherwise you'll spend way too much time and money chasing your tail like I did.