Good article. One minor mistake: the GEC KT88 was not derived from the TT21 transmittter valve, but the other way round. Two GEC engineers unofficially pursuaded a couple of production line girls to put a few KT88 innards through the final stages of production of a top cap valve (TV line output?), because they thought that a top-cap version of the KT88 might be a good Tx valve. They were right. After tests confirmed this, the TT21 became official.
Glad to know it was in someway helpful.
What I want to know is in Table 1 from the Mullard publication comparing the total distortion of both the EL34 & EL84 where the pentode mode produces ~ twice the total distortion of both triode and UL modes, did the DUT for each tube type use a regulated screen or RC or voltage divider tap?
What I want to know is in Table 1 from the Mullard publication comparing the total distortion of both the EL34 & EL84 where the pentode mode produces ~ twice the total distortion of both triode and UL modes, did the DUT for each tube type use a regulated screen or RC or voltage divider tap?
I happen to agree with you Wavebourn although I would not go so far as to say UL is nonsense,
Actually, I did not say that it is a nonsense. I said,
Everything reveals in comparison.
Interesting that they seem to talk only about P-P. Ultra Linear can be used single ended, too. (or did I miss that somewhere in the text?)
Also, they assumed only negative G1 drive, limiting all musings around "on onset of grid current". And of course, did not observe an alternative proposed by Schade long time before: feedback applied between cathode and the 1'st grid.
He did not write in that famous article about regulated screen grid voltages. What he proposed, to use local feedback around 6L6 tubes, from plates to control grids, to make them usable for audio applications. In his example he applied feedback in series with secondaries of phase splitting transformers that drive first grids. He mesaured results and provided them in the table for comparison.
I am going from memory here but am sure I have somewhere amongst my files a schematic with I think Schade's feedback scheme. IIRC it used a 12AX7 front end, 12AU7 as driver and I think 1625 or maybe 807 outputs. I believe it used a capacitor along with the resistor in the feedback circuit as a filter I suppose. I must look for it. I will post it in any case. It was, to me unusual.
It is my opinion, by and large, that several of our amplifier designs are not so much different output stages but more like different methods of application of feedback.
It is my opinion, by and large, that several of our amplifier designs are not so much different output stages but more like different methods of application of feedback.
Last edited:
I considered Schade for my SE amp, it's a very neat system because it requires no capacitor in the feedback loop and it avoids the output transformer - a device entirely unsuited to feedback IMO.
If I use just Schade for my new PP project, I'd perhaps use a low impedance driver (SRPP) into a series resistor to mix the feedback into - anyone done this? I'm aiming for as much feedback as practical into a GU50 with G2=250V and a 400V(ish) plate.
If I use just Schade for my new PP project, I'd perhaps use a low impedance driver (SRPP) into a series resistor to mix the feedback into - anyone done this? I'm aiming for as much feedback as practical into a GU50 with G2=250V and a 400V(ish) plate.
I did something similar on the KT-88 I am working on now (well PA is for the project is done unless I find a problem later on). I used normal common cathode 6CG7 driver running about 6mA with 11K from driver plate to the feedback junction. FB resistor is on the order of 120-140k (I forget where I finally ended up) IIRC grid leak resistor was 220k. I did this mainly with the idea of reducing the variation in FB ratio with driver current.
Seems to work fine as in sounds nice but unfortunately I haven't the equipment to generate any test results nor the time (need to complete the project soon) to do a lot of listening tests with different layouts. Kegger over on the other forum said he might do some experimentation with that but I will probably have to pester him a little bit.
Or this one? The capacitor is just a DC blocking device. The gNFB loop didn't require an external capacitor since the feedback signal line is 75R RF coax, and its inherent capacitance is enough to stabilize the amp.
Attachments
No, its not that one Miles, (I also have yours filed away )
I found the schematic but its not Schade, not UL either, I don't think it belongs in this thread. It is pretty old but a goodie I think, could use modernization.
Was that by design or did it work out that Way? Thats neat "killing two birds with one stone".
Reminds me of a direct coupled "RH" amp where the feedback resistor is also the plate resistor of the driver, another schematic filed away (more like a stack of ~2500 sheets of paper on a shelf). One day I should really sort that stack but then I really would not be able to find anything!
)I found the schematic but its not Schade, not UL either, I don't think it belongs in this thread. It is pretty old but a goodie I think, could use modernization.
Was that by design or did it work out that Way? Thats neat "killing two birds with one stone".
Reminds me of a direct coupled "RH" amp where the feedback resistor is also the plate resistor of the driver, another schematic filed away (more like a stack of ~2500 sheets of paper on a shelf). One day I should really sort that stack but then I really would not be able to find anything!
Do you mean C9 and C10?
I was thinking that just like a coupling capacitor effects the sound, so would a feedback capacitor - hence the neatness of the Schade method which makes sure the feedback signal is nice and accurate...
Thanks Wavebourn, I'll check that on the scope when it's hooked up - I want the amp to clip on the GU50s rather than the harder SRPP driver.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.