What about this one. Is this the same one you mentioned?
https://www.mouser.com/ProductDetail/Vishay-Dale/NH010400R0FE02?qs=/pO/jbY2KwU7VcfCR/vOpg==
https://www.mouser.com/ProductDetail/Vishay-Dale/NH010400R0FE02?qs=/pO/jbY2KwU7VcfCR/vOpg==
That part is non-inductive and, therefore, an excellent find by you. It does require mounting to a metal chassis, for full dissipation capability.
If I wanted to parallel the output tubes for more power it is as simple as adding one, choosing a 3K OT and making sure the PT can supply the required current?
I assume its not that simple but that is why I am asking.
I assume its not that simple but that is why I am asking.
Yes, and I assume just use NFB from one of the output tube plates? Or mix them with 2 NFB resistors?
The plates of the paralleled output tubes connect together and the same Rfb departs from this connection.
For parallel single ended (PSE), a different O/P trafo, whose primary impedance is 1/2 that used with a single tube, is appropriate. Increased power handling capability will be necessary, if a GNFB loop is to eventually be added.
The tubes used in a parallel pair should be well matched for transconductance (gm) and cathode current. If closely matched tubes are not available, "mere" separate RC cathode bias networks rate to be inadequate. A change is required in how the O/P tubes are biased that incorporates a negative supply and individual bias trim pots., to allow of "exact" quiescent current setting/matching.
The issue of comparatively high SE O/P "iron" expense, alone, makes the Cit. 5 knock off a (IMO) much better choice for the increased power O/P 2X 1625s would yield.
The tubes used in a parallel pair should be well matched for transconductance (gm) and cathode current. If closely matched tubes are not available, "mere" separate RC cathode bias networks rate to be inadequate. A change is required in how the O/P tubes are biased that incorporates a negative supply and individual bias trim pots., to allow of "exact" quiescent current setting/matching.
The issue of comparatively high SE O/P "iron" expense, alone, makes the Cit. 5 knock off a (IMO) much better choice for the increased power O/P 2X 1625s would yield.
Eli, The CXSE I would choose is for the single tube output is rated at 25W. So would I be alrighty invested in O/P iron for a parallel tube option? Is there more to consider or is it not wise to run the CXSE near their rating?
Last edited:
I'm late at this very busy party, but here are a few comments regards the RH807 schematic. And apologies to those who may have already made these observations.
The 807 won't do much with an unbypassed 10K screen resistor. Better would be a 100R CC resistor to avoid parasitic oscillations from developing.
The internal NFB resister from plate to plate reduces the internal impedance of the amplifier. The amp will be more susceptible to power supply ripple.
In any case the internal NFB doesn't amount to much, the 100K FB resister is driving back into the 807 G1 resister in parallel with the 1/2 ECC81 plate resistance of ~10K.
Two LC filter sections to provide channel isolation is overkill. Better spend the money saved by utilizing a single LC section on good gapped OPTs.
The circuit with this lineup has plenty of gain to support full loop NFB from the OPT secondary. I've built several like this, never with an expensive OPT. And used up to 20 db NFB with no instability problems at all. And got measured results to confirm that.🙂 Like what's in this one, 6AQ5/6AU6, built in 1968, still running in the workshop every day.
Attachments
By 100R screen resistor do you mean 100 ohm? (sorry I am somewhat new to this forum).
Wouldn't that result in too high a screen voltage for this tube?
Wouldn't that result in too high a screen voltage for this tube?
300V on the screen of an 807 is OK. But if say a 6V6GT was to be used, something less would be better for long life.
For a cap input power supply, 250-0-250 would result in about 270V after losses in the choke. And a reasonable first cap of say 20 microF is a good choice. Some use monster input caps, that puts a lot of stress on the PT. It is the resulting RMS currents set up by large caps that causes transformer heating & failure. 5V4G rectifier is a better choice than the 5U4 shewn in the SE807RH schematic, altho it may not look a impressive.
All of this kind of information is found in many of the old engineering texts from the heyday of vacuum tubes.
For a cap input power supply, 250-0-250 would result in about 270V after losses in the choke. And a reasonable first cap of say 20 microF is a good choice. Some use monster input caps, that puts a lot of stress on the PT. It is the resulting RMS currents set up by large caps that causes transformer heating & failure. 5V4G rectifier is a better choice than the 5U4 shewn in the SE807RH schematic, altho it may not look a impressive.
All of this kind of information is found in many of the old engineering texts from the heyday of vacuum tubes.
I want to go SS power supply. Can a varistor be used between the PT and the first cap to limit the inrush?
What about the B+ being on before the output tube heaters are fully on. Is that really a thing? Or not with this level of power?
Cathode stripping starts to become an issue above 500 V. The thermistor is good enough, especially in an amp with a SS rectified negative bias supply. The negative rail turns on "instantly", while a slight delay occurs in the B+ rail. That slight delay allows the bias supply to electrostatically shield the O/P tube cathodes. 😀
The 1625 is a fantastic tube. Remember, a 1625 is an 807 with a 12V filament, and an 807 is essentially a 6L6G with a different base and a plate cap.
I've been very, very happy with the push-pull 1625 amp that I built. It sounds amazing, and although it really isn't that important, the 1625 is a very pretty looking tube.
One suggestion though: Even the newest 1625s are pushing 60 years old. It is a wise idea to "recondition" them by running only the filament for about 24 hours before using them in an amplifier. This helps to clean up the vacuum and minimize the chance of arcing. Really not a big deal to do, but something that I suggest doing with any power tubes that haven't been used for a long time. As an example, the 715C was designed as a pulser tube for WWII radar systems (Absolutely no use for audio, it is horribly non-linear). It is rated to hold off something like 20 kV, but most WWII surplus 715Cs will only hold off about 1500 at best. If you run the filament for 24 hours, they'll take the full 20 kV they were designed for.
As mentioned, cathode poisoning shouldn't be an issue here. Essentially the problem comes down to depletion of the electron cloud, allowing positive ions to bombard the cathode. More information can be found in Care and Feeding of Power Grid Tubes.
I've been very, very happy with the push-pull 1625 amp that I built. It sounds amazing, and although it really isn't that important, the 1625 is a very pretty looking tube.
One suggestion though: Even the newest 1625s are pushing 60 years old. It is a wise idea to "recondition" them by running only the filament for about 24 hours before using them in an amplifier. This helps to clean up the vacuum and minimize the chance of arcing. Really not a big deal to do, but something that I suggest doing with any power tubes that haven't been used for a long time. As an example, the 715C was designed as a pulser tube for WWII radar systems (Absolutely no use for audio, it is horribly non-linear). It is rated to hold off something like 20 kV, but most WWII surplus 715Cs will only hold off about 1500 at best. If you run the filament for 24 hours, they'll take the full 20 kV they were designed for.
As mentioned, cathode poisoning shouldn't be an issue here. Essentially the problem comes down to depletion of the electron cloud, allowing positive ions to bombard the cathode. More information can be found in Care and Feeding of Power Grid Tubes.
Final question (Just kidding)...
Construction. I am planning on spinning a PCB for the PS. Was thinking about doing the same with the amplification side as well. What are peoples thoughts on this versus PP wiring?
This would be apart from things I currently think I know:
-PP makes component swapping and experimentation easier
-PP makes heat management easier
-PCB reduces chances of wiring errors.
Construction. I am planning on spinning a PCB for the PS. Was thinking about doing the same with the amplification side as well. What are peoples thoughts on this versus PP wiring?
This would be apart from things I currently think I know:
-PP makes component swapping and experimentation easier
-PP makes heat management easier
-PCB reduces chances of wiring errors.
Why is a 10W rating called out for the cathode resistor? By my calcs, 50ma through a 400 ohm resistor dissipates 1 W.
Good question! A hefty reserve of capability that allows for poor air circulation, etc. is one thing, but overkill is another. The originator of the RHnn amps is banned from this site, for (IIRC) contentious behavior.