I'm contemplating building a PP 2A3 amplifier and I have a good 5 VDC SMPS source. Is it possible to wire a pair of 2A3 filaments in series? I don't have experience with DHTs [obviously]. Thanks in advance!
I wouldn't recommend it. You end up with the tubes DC bias off by +2.5V on one and -2.5V on the other. You also need to have some level of decoupling (even with a DC supply), depending on whether you plan to use cathode or fixed bias. Also note that the 2A3 is not spec'd for series filament operation, unlike many older tubes destined for the classic 5-tube Superhet-type AM table radios as all of these were indirectly heated cathodes.
For a push-pull amplifier, you might want to look at using a pair of 6B4G tubes. These are basically a 2A3 with a 6.3V filament. You can run the filaments in parallel with a clean DC supply and likely do better.
For a push-pull amplifier, you might want to look at using a pair of 6B4G tubes. These are basically a 2A3 with a 6.3V filament. You can run the filaments in parallel with a clean DC supply and likely do better.
The Amity amplifier has the 2A3 filaments in parallel.... pin 1 to pin 1 and pin 4 to pin 4. The fact that it's a push-pull circuit provides some phase cancellation. However, this might not matter as most later 2A3 tubes are actually dual filaments and grids within a single, albeit separate plate structure. In single-ended circuits, you need to balance out the hum, or use a DC filament supply. I've actually found the older (original) single plate 2A3 triodes to be quite difficult to null out the hum using AC on the filament. In any case, depending on the pair of DHTs you have, some pairs will work out better than others.
Attachments
My reason for building the amp was a surplus of NOS 2A3 tubes. But it seems like I would have to jump through a lot of hoops to use them on a PP design. I think your 6B4G suggestion is a great one. Thank you!
If you have a good selection of 2a3, go for it.
A good circut using a transformer for phase splitting and a good opt will yield a very great sounding amp with se like mid range and great bass control.
A good circut using a transformer for phase splitting and a good opt will yield a very great sounding amp with se like mid range and great bass control.
I'm not sure that changing to the 6B4G alters anything. I would not use series filaments or a switching supply with either tube. Directly-heated tubes are very sensitive to filament supplies and you'll hear that SMPS clearly. All you need for a 2A3 PP amp is a single good 2.5 volt transformer. They can be wired in parallel, but you won't be able to adjust the bias for the tubes individually so they will have to be well-matched to avoid any significant imbalance. Look at some of the old PP 2A3 amp designs. They can share a single filament supply, but the better examples provide individual supplies for each tube. Nonetheless a shared filament transformer is one way to go to avoid those "hoops."
Many thanks for your suggestions. May I ask you a few more questions?
I was initially in favor of the 6B4G are the octal sockets and the ability to use a repurposed Eico ST-70 power transformer's 6.3V filament supply. The nominal secondary is 720V CT with 117V input, which could be lowered with a choke input design. But then I'd need 2.5V . . . .
I'm contemplating a PP design with fixed bias (maybe @ 45V?) for optimal performance.m. What would the target B+ be in that case (e.g. 320V)?
Maybe the best solution would be for Heyboer to wind a custom tranny with (4) 2.5VCT windings; a bias tap; and a HV secondary (either doubling or not).
Thoughts?
Thanks in advance.
I was initially in favor of the 6B4G are the octal sockets and the ability to use a repurposed Eico ST-70 power transformer's 6.3V filament supply. The nominal secondary is 720V CT with 117V input, which could be lowered with a choke input design. But then I'd need 2.5V . . . .
I'm contemplating a PP design with fixed bias (maybe @ 45V?) for optimal performance.m. What would the target B+ be in that case (e.g. 320V)?
Maybe the best solution would be for Heyboer to wind a custom tranny with (4) 2.5VCT windings; a bias tap; and a HV secondary (either doubling or not).
Thoughts?
Thanks in advance.
Are you talking about a stereo amp? If so, you would need at least two 6.3 volt taps taps for a stereo 6B4G amp. You can't use the same filament supply for both channels, you will get crosstalk between them.
In terms of B+, it depends on how you plan to run the tubes and how much power you want. 2A3 and 6B4G are electrically the same tube. For Class A, you want -45 volts bias and 250 volts from cathode to plate. With fixed bias, that would mean about 260 volts B+ (you will lose 10 volts across the ouput transformer). That will get you about 6 watts in push pull. For Class AB, you will need -62 volts bias and a supply of about 310 volts. This will get you about 15 watts per channel in PP.
The Eico power tranny will not work for stereo Class A. There would be a total of 240 mA just for the ouput tubes. That's too much for the Eico tranny. It was meant for Class AB operation, where they static current is much lower. It might work for Class AB.
The problem with Class A operation is the low B+, which severely limits the driver stage, so you have to use a choke load or interstage transformer. With Class AB you have up to 400 volts B+ so you can use a conventional resistoir-cap coupled driver.
Below is a really nice, simple PP 2A3/6B4G amp that is good for about 10 watts. Ignore the green modifications, they aren't necessary, and the amp is fine without them. I have built this and it's lovely. In place of the 76 and 6A6, I used 6P5 and 6N7, which are the same tubes with octal bases. Both are easily available. You can also use 6B4Gs. Your Eico power tranny should work fine for this in stereo. Just add an extra 6.3 volt tranny for the second channel 6B4Gs or two 2.5 volt trannies for 2A3s.
The only change I might make is to increase the 400 ohm cathode resistor to 700 ohms to reduce the idle current and push the amp more into AB territory. Your Eico power tranny might be happier with this. But you can try it both ways.
In terms of B+, it depends on how you plan to run the tubes and how much power you want. 2A3 and 6B4G are electrically the same tube. For Class A, you want -45 volts bias and 250 volts from cathode to plate. With fixed bias, that would mean about 260 volts B+ (you will lose 10 volts across the ouput transformer). That will get you about 6 watts in push pull. For Class AB, you will need -62 volts bias and a supply of about 310 volts. This will get you about 15 watts per channel in PP.
The Eico power tranny will not work for stereo Class A. There would be a total of 240 mA just for the ouput tubes. That's too much for the Eico tranny. It was meant for Class AB operation, where they static current is much lower. It might work for Class AB.
The problem with Class A operation is the low B+, which severely limits the driver stage, so you have to use a choke load or interstage transformer. With Class AB you have up to 400 volts B+ so you can use a conventional resistoir-cap coupled driver.
Below is a really nice, simple PP 2A3/6B4G amp that is good for about 10 watts. Ignore the green modifications, they aren't necessary, and the amp is fine without them. I have built this and it's lovely. In place of the 76 and 6A6, I used 6P5 and 6N7, which are the same tubes with octal bases. Both are easily available. You can also use 6B4Gs. Your Eico power tranny should work fine for this in stereo. Just add an extra 6.3 volt tranny for the second channel 6B4Gs or two 2.5 volt trannies for 2A3s.
The only change I might make is to increase the 400 ohm cathode resistor to 700 ohms to reduce the idle current and push the amp more into AB territory. Your Eico power tranny might be happier with this. But you can try it both ways.
Last edited:
Again, your kindness is extraordinary. Thank you!
That looks like a Japanese schematic?
Any harm in using the amp without the green changes? I might keep the 5Z4 sub and a choke of at least 5H, along with the 6.6K tranny. Maybe sub a 6J5 I have on hand for the 76 / 6P5?
That looks like a Japanese schematic?
Any harm in using the amp without the green changes? I might keep the 5Z4 sub and a choke of at least 5H, along with the 6.6K tranny. Maybe sub a 6J5 I have on hand for the 76 / 6P5?
Yes, it is a Japanese design. Personally, I would keep the 5AR4 for the slow warm-up time, especially since the input stage and driver are direct-coupled. A 6J5 would work fine, but you might want to skip the cathode-bypass cap since the gain is a little higher or include it and add the feedback. Watch the plate voltage, again because of the direct-coupling. You will probably need to reduce the cathode resistor to around 2500 ohms. 6.6K OPT would work, as would a 5H choke. You can also experiment with the feedback--leave it out or include it. If you decide to use it make sure the primary leads to the output tubes are wired in the correct phase. The amp will tell you if they are not. ;-) It will howl with displeasure, literally. ;-)
I decided to have a custom power transformer with separate 2.5 VAC CT supplies for each of the (4) 2A3s. If I ground the center taps of the 2.5 VAC windings, can I eliminate the 100 ohm hum balance pots?
You don't want to ground the filament supply, obviously, but you can attach the cathode resister and bypass cap to the CT of the filament supply, yes.