For the circuit in #80, the time constants C1, R23 & C5, R29 are to close to each other. If NFB were used there is danger of LF instability. For a long tailed phase invertor the relation Rk ( mu +1 ) >> rp + Rl must be satisfied. This one barely makes it, but R21 has been made slightly larger to help that. A CC in the tail would be a better fix. There is no cap between the input jack & the first grid. A drop of 25V across only half of the OPT winding with a plate current of 50 mA indicates 500R. That is very large, eeven a low cost replacement OPT would do better.
All these points make me wonder if the circuit was ever built.
All these points make me wonder if the circuit was ever built.
I've had something similar running on & off the bench for about a year. An honest 7W at clipping in Class A PP without exceeding plate dissipation. The 6EM7 dissimilar triode has about the same plate dissipation as the 6AV5. But includes another high mu triode for a voltage amplifier. Connected as a differential stage here with a 120K tail to -150V, results in a very good phase splitter. Still needs a front end. The spectrum shews SMPTE Intermediation Distortion at 3W,
Attachments
Um no. not 30W. However it is a decent amplifier considering when it was designed. No way you get 30W out of 2a3 P-P. Nope.
How so? It just measured 20W on the bench today.
Perhaps you know something the rest of us don't.
Shew us your proof, with a captioned schematic.
And some real calculations.
Well, you can, and the Brook did, with CF drivers and sliding auto-bias.
https://www.tubecad.com/2007/12/15/High-Quality Audio Amplifier With Automatic Bias Control.pdf
Here is a 20wpc triode EL34 amp that is extremely easy to build. I'm lucky to have the LS-57 outputs, but a high-quality 5K p-p transformer should work as well. You might have to back off the feedback a bit. I've built a number of Williamson amplifiers and this is one the best-sounding to date. A really lovely amplifier. I used the stock Heathkit Williamson design he indicates in the vid, but did a little tweaking to the shelf filter and feedback values for stability and smoother response. Oh, and that's not me in the video, I just copied the design. ;-)
Last edited:
Shelf Filter?
Did you mean the dominant high frequency pole, series RC from a plate to ground,
or alternatively, a series RC across the plate load resistor?
Thanks!
Did you mean the dominant high frequency pole, series RC from a plate to ground,
or alternatively, a series RC across the plate load resistor?
Thanks!
Yes, the low-pass filter, some people call it a high-frequency shelf network, across the plate resistor of the input tube to limit the ultrasonic HF response. They became pretty much standard in Williamson amps after it was found they were prone to oscillate due to the extended HF response.
Last edited:
I believe . . .
The purpose of the dominant pole in an amplifier that has global negative feedback, is to make the total amplifier gain go to unity at a high frequency, before the output phase at that frequency becomes 180 degrees.
Because when the negative feedback loop shifts by 180 degrees, it becomes positive feedback.
Positive feedback, and gain more than unity . . . makes the amplifier an oscillator.
Often the fastest changing high frequency phase shift in an amplifier is the phase shift from the output transformer primary to the secondary (2 pole low pass filter). 12dB/octave
Often, the output transformer is also a high pass filter to low bass frequencies (1 pole high pass filter). 6 dB/octave
Build an amplifier, you may get an oscillator.
Build an oscillator, you may get an amplifier.
That happens more often than you think!
Just my opinions
The purpose of the dominant pole in an amplifier that has global negative feedback, is to make the total amplifier gain go to unity at a high frequency, before the output phase at that frequency becomes 180 degrees.
Because when the negative feedback loop shifts by 180 degrees, it becomes positive feedback.
Positive feedback, and gain more than unity . . . makes the amplifier an oscillator.
Often the fastest changing high frequency phase shift in an amplifier is the phase shift from the output transformer primary to the secondary (2 pole low pass filter). 12dB/octave
Often, the output transformer is also a high pass filter to low bass frequencies (1 pole high pass filter). 6 dB/octave
Build an amplifier, you may get an oscillator.
Build an oscillator, you may get an amplifier.
That happens more often than you think!
Just my opinions
Last edited:
Yes, I understand, thanks. My point was that the amp, as built in the video, was unstable and wouldn't tolerate any capacitance across a fixed load. I swapped in a more aggressive shelf filter and did some tuning to make it very stable.
In any event, it's a lovely amp and capable of 18-20 watts. Since the OP was shifting his thinking to a triode-wired EL34 amp I thought I mention this classic Williamson approach--relatively simple and with a very natural tone. I have not tried any other output transformers with it. Possibly a consultation with Jack Eliano at Electraprint would result in an affordable alternative that would tolerate the amount of feedback used.
Hi. Pregnant pause at the moment. Have order some, ehem, Lundahl 1679s. Not really cheep anymore, but I sold some stuff and had a bit of extra cash. Also they are relatively more huge than I had allowed for in the original design, so have stopped building as I will work out where everything will fit, when they turn up... Hope they turn up. Was supposed to be 2 weeks, but nothing as yet.
I am building the the circuit in post 72, as it is a little un usual with a transformer input, and promises low distortion. I have enjoyed all the other circuits, and quite like the idea of the non symmetric triode amp, and introduced me to some other nice valves, like the CV6040, which is a 12E1 in a smaller bottle.
I am building the the circuit in post 72, as it is a little un usual with a transformer input, and promises low distortion. I have enjoyed all the other circuits, and quite like the idea of the non symmetric triode amp, and introduced me to some other nice valves, like the CV6040, which is a 12E1 in a smaller bottle.