Can anyone recommend a proven Parallel Push Pull tube power amplifier circuit that uses standard octal tubes?
Why is Parallel Push Pull not more common? Are the problems with strapping plates of multiple output tubes just too hard to overcome? Surely someone has attempted to address the impedence imbalance issues.
Any ideas on who has done more development in Parallel Push Pull?
Any ideas on who has done more development in Parallel Push Pull?
If you parallel tubes, you need to get them to share current equally. This usually means some degeneration resistors in the cathodes (some, maybe 25%, proportion of 1/gm nominal for the R's), and individual bias adjustments. Matched tubes (gm) would be a very good idea too. Grid stoppers definitely needed, for both g1 and g2, and likely some VHF chokes (Wire wound on a 2 Watt, 470 Ohm or so resistor) in the individual plate circuits. Check the amp for parasitic oscillations with a high BW scope when finished, some adjustments maybe required.
Then you can just consider the bunch as a single tube with N x gm and N x the input capacitance, and with N x the current and Watts capability. You can use the same plate curves graph for operating point selection, just use N x times the plate current indicated.
My guess is not many amplifiers got designed this way, since choosing the correct tube would be easier and cheaper (back then at least), and one has no control over what grade of tube matching or tube type the user might put in the thing, maybe causing instability at VHF or DC. Manufacturers don't want a nightmare product to support. Servicing could be an issue too, trying to find which tubes are causing problems. Probably have to change them all at once to keep them matched. And if they don't get changed all together each time, you face an N x failure/repair rate. Only a DIY amp, with a big box of cheap tubes available, and a matching tester, could tolerate this.
Then you can just consider the bunch as a single tube with N x gm and N x the input capacitance, and with N x the current and Watts capability. You can use the same plate curves graph for operating point selection, just use N x times the plate current indicated.
My guess is not many amplifiers got designed this way, since choosing the correct tube would be easier and cheaper (back then at least), and one has no control over what grade of tube matching or tube type the user might put in the thing, maybe causing instability at VHF or DC. Manufacturers don't want a nightmare product to support. Servicing could be an issue too, trying to find which tubes are causing problems. Probably have to change them all at once to keep them matched. And if they don't get changed all together each time, you face an N x failure/repair rate. Only a DIY amp, with a big box of cheap tubes available, and a matching tester, could tolerate this.
Last edited:
Also, while I'd certainly take Smoking-Amp's advice about all the matching and adjusting and protection, plenty of people simply parallel the tubes and nothing more w/out any trouble.
(see pmillett.. "807" Push-Pull amp)
I built an amp off his driver PCB, also w/ parallel output tubes. He convinced me to just parallel them, w.out kludging in another set of bias adjustments. I've been using the amp a couple times a week for over a year now, w/out any trouble or bias shift or anything. That amps all to the max, too.. Highest voltages I can get away w/ for plate and screen and drivers, always playing it loud on relatively inefficient speakers. (My power supplies are all regulated, tho)
(see pmillett.. "807" Push-Pull amp)
I built an amp off his driver PCB, also w/ parallel output tubes. He convinced me to just parallel them, w.out kludging in another set of bias adjustments. I've been using the amp a couple times a week for over a year now, w/out any trouble or bias shift or anything. That amps all to the max, too.. Highest voltages I can get away w/ for plate and screen and drivers, always playing it loud on relatively inefficient speakers. (My power supplies are all regulated, tho)
Last edited:
If you plan to simply parallel output tubes, make sure that you don´t push them up to their max. power dissipation. If the current is not shared equally, some tube plates might glow red 😀
PPP is not a good topology because in most cases you don´t even need that much power. Buy decent speakers first and you will see that even 1W out of eg. a EL95, can make your neighbours, call the police on you 😉
PPP is not a good topology because in most cases you don´t even need that much power. Buy decent speakers first and you will see that even 1W out of eg. a EL95, can make your neighbours, call the police on you 😉
- Status
- Not open for further replies.