Push Pull 2A3

I just saw this thread. Looks like a great project Joe! Keep us posted on how it’s coming along.
Thanks Dave, your help and advice has been invaluable to me! I'll be sure to update this thread, and I'm sure I'll have quite a few questions about implementation.

I've stocked up on some of the Jensen electrolytics that you recommended even if I end up going with the metallised Film in oil cans as it shouldn't be hard to make a switch. Also got the other parts you recomended like riken resistors and already had a lot of AN magnetic 2W but will prob go with the silver ones at some point. Also decided that I'll try the choke input Psu first as the guy who I spoke to was adamant that it sounded better for this amplifier after trying both ways.

Hearing the WE124 vs another WE Set amplifier (might have been a WE91) was an eye opener, I nearly decided on a WE124 but its quite complex and prob not suited for a first attempt.
 

Salectric

Senior Member
Thanks Dave, your help and advice has been invaluable to me! I'll be sure to update this thread, and I'm sure I'll have quite a few questions about implementation.

I've stocked up on some of the Jensen electrolytics that you recommended even if I end up going with the metallised Film in oil cans as it shouldn't be hard to make a switch. Also got the other parts you recomended like riken resistors and already had a lot of AN magnetic 2W but will prob go with the silver ones at some point. Also decided that I'll try the choke input Psu first as the guy who I spoke to was adamant that it sounded better for this amplifier after trying both ways.

Hearing the WE124 vs another WE Set amplifier (might have been a WE91) was an eye opener, I nearly decided on a WE124 but its quite complex and prob not suited for a first attempt.
Thanks Joe!

A choke input PS should work nicely. If you recall I tried both cap input and choke input on my 46 amps, and I preferred choke input. Choke input also lets you use a 5R4GY as well as a 5U4GB. I preferred the 5R4GY but at least you have the flexibility to use either one with choke input.

In thinking some more about your circuit, my only concern is whether the gain will be adequate. If it is, I think you will be very happy with it.

Also, you should probably try Paul’s recommendation about the 3 gridstopper resistors. I tried that initially with my 417a using 3 carbon comp resistors tied together at one end. Unfortunately I just don’t like how cc resistors sound so I gave up on that arrangement rather quickly. However 3 Riken carbon film resistors should sound much better. As you know, I ended up with a single Riken connected to just one of the grids and I snipped off the tube socket tabs for the other grids. While that has worked fine for me, I can certainly appreciate that it might not work as well in all applications.
 

Thermionics

Post Whore In Training
I have push-pull 6B4G amps sitting in my listening room that are AC heated and nice and quiet. A single hum pot will do nicely for an amp like this. A single ended 2A3 amp can be built to be pretty quiet, and the push-pull output stage will knock it down more.

If you have hum, the first thing to do is make sure it's hum and not buzz. After that, the circuit can be split up a bit to see where it might be coming from.
Do you make house calls, Paul? :D
 

marantzfan

Administrator
Staff member
Is that in Arizona? :D
Seattle-ish.

Funny, I don’t know if I ever told you this Paul, but the day you brought me to Bottleheadquarters, I had a little problem getting back. ;)

Took the fairy back to Seattle and me and my BeePre went to Pike Brewing for a beer. I had some time before I had to catch my flight back.

After finishing a couple beers, I took the train towards what I thought was the Airport. It wasn’t until I heard them announce that Renton was the next stop that I asked someone “do you know how much farther to the airport?”

Sufficed to say, I had to reschedule my flight for the next day and stay another night.
 

paulbottlehead

Active Member
Ha ha, yeah, the train is a little slow. You'll have to come back and visit soon when the new space is setup. If you had called me we could've put you up for the night.
 

marantzfan

Administrator
Staff member
Ha ha, yeah, the train is a little slow. You'll have to come back and visit soon when the new space is setup. If you had called me we could've put you up for the night.
I hope to get back up there one of these days. You and Dan were extremely gracious hosts and I felt like a kid in a candy shop!
 
I have a breadboard 6B4G PP amp based loosely on the Poindexter Musical Machine originally. It sounded really nice but when I changed it to IT's then I converted to 6E5P triode strapped differential drivers. Sounds great to me. Magnequest AS-250 M4 outputs.
 

mred

Senior Member
I have a breadboard 6B4G PP amp based loosely on the Poindexter Musical Machine originally. It sounded really nice but when I changed it to IT's then I converted to 6E5P triode strapped differential drivers. Sounds great to me. Magnequest AS-250 M4 outputs.
I have the same AS-250 transformers for my 2a3 amps.
 
The more time I have spent on this the less confident i am about this build.

I've had the chance to do a bit more work on this and have concentrated on the PSU as I think the signal section is more or less done.

The original schematic I based this amp on did not have any values for B++ and there were reportedly some errors in the PSU schematic anyway so I planned to come up with a better one, I realised this was ambitious at the time but I think it was actually more complicated than I thought!

While researching PSUs used in other similar push pull amplifiers I noticed that the "Seth" 2A3 Push Pull amplifier (which has a similar signal section) has a build guide as a PDF gives the B++ to the 2A3s as 320v - so as I understand it the output transformer would be given around 325-330v as it drops a bit of voltage. one difference is that the output transformers used are 5K while mine are either 6K or 6.6K *, i know this will make a difference but I'm not sure what exactly.

I have tried to work out a choke input PSU in PSUD, there is so much im not sure about here I dont really know where to start... I suppose the first thing is all the variables to input for the power transformer, I have just entered the secondary voltages and crossed my fingers that is enough to be getting on with.

I have played with the secondary values for the power transformer and it looks like Id need a 450 - 0 - 450 transformer to result in a B++ of about 330v (assuming that is in the correct "ballpark". I wont be using a dropping resistor for the B+ (first stage) but couldn't work out how to add a VR tube in PSUD, I think that part is fine.

I have attached a screen shot from PSUD. If anyone who knows what they are doing could take a look over it and let me know if I'm on the right track - it would be much appreciated!

*having read some conflicting reports about the output transformers I intend to use (some say 6K and others 6.6K) I understand I need to know the correct value so i'll try and get to the bottom of it in the next week or two as I should be able to get some help with this from someone who knows what they are doing.

PP2a3PSU.jpg
 

paulbottlehead

Active Member
While researching PSUs used in other similar push pull amplifiers I noticed that the "Seth" 2A3 Push Pull amplifier (which has a similar signal section) has a build guide as a PDF gives the B++ to the 2A3s as 320v - so as I understand it the output transformer would be given around 325-330v as it drops a bit of voltage. one difference is that the output transformers used are 5K while mine are either 6K or 6.6K *, i know this will make a difference but I'm not sure what exactly.
Higher impedance tends to go with higher plate voltage and less operating current. For a 6 or 6.6K transformer, 300V plate to cathode and 50mA of plate current is the change I would make.

I have tried to work out a choke input PSU in PSUD, there is so much im not sure about here I dont really know where to start... I suppose the first thing is all the variables to input for the power transformer, I have just entered the secondary voltages and crossed my fingers that is enough to be getting on with.
If you put in the secondary voltage, rated AC current, and regulation, you'll get pretty close. I tend to use 10% as my regulation value unless the power transformer is very high end.

I have played with the secondary values for the power transformer and it looks like Id need a 450 - 0 - 450 transformer to result in a B++ of about 330v (assuming that is in the correct "ballpark".
For a choke input filter, this looks somewhat reasonable. You could use less voltage with a cap input filter, but the AC current drawn from the power transformer will go up.

I wont be using a dropping resistor for the B+ (first stage) but couldn't work out how to add a VR tube in PSUD, I think that part is fine.
If you're going to use a VR tube, you need a current limiting element between it and B+, which would either be a dropping resistor or a constant current source. None of this is easily modeled in PSUD.


*having read some conflicting reports about the output transformers I intend to use (some say 6K and others 6.6K) I understand I need to know the correct value so i'll try and get to the bottom of it in the next week or two as I should be able to get some help with this from someone who knows what they are doing.
6K vs. 6.6K isn't the end of the world, but if they had been 8K or 10K, you'd be making nearly the same power as a single ended 2A3 amp, and that would be a little bit funny.
 
Higher impedance tends to go with higher plate voltage and less operating current. For a 6 or 6.6K transformer, 300V plate to cathode and 50mA of plate current is the change I would make.


If you put in the secondary voltage, rated AC current, and regulation, you'll get pretty close. I tend to use 10% as my regulation value unless the power transformer is very high end.


For a choke input filter, this looks somewhat reasonable. You could use less voltage with a cap input filter, but the AC current drawn from the power transformer will go up.


If you're going to use a VR tube, you need a current limiting element between it and B+, which would either be a dropping resistor or a constant current source. None of this is easily modeled in PSUD.



6K vs. 6.6K isn't the end of the world, but if they had been 8K or 10K, you'd be making nearly the same power as a single ended 2A3 amp, and that would be a little bit funny.
Thankyou very much Paul, thats a really big help and now that you have said it I have realised it matches what a chap I spoke to who built a similar amplifier with the same output transformers did (B++ 300v) although he used fixed bias (I don't think I will) . I'll rearrange the power transformer with the aim to get 300v B++.

I guess I can get in touch with sowter and finally get the chokes ordered and take a look at what Hammond have re power transformers. I guess 200ma would do for the power transformer or should I be looking for something with a higher rating? I think the choke I was planning on ordering is also 200ma, the 300ma one is something like 4kg!
 

paulbottlehead

Active Member
If you are not using fixed bias, then you will need more than 300V of B+ for the 2A3s. You will need to add the DC voltage drop through the output transformer and the bias voltage from filament to ground. This will make your B+ much more like 380V.

Your PSUD simulation shows that a 200mA rated AC winding would be plenty of current for a monoblock but a little shy for a stereo amp. I would be getting the power supply chokes from Hammond and the power transformer from Sowter. Do also be aware that a 6.3V heater winding with a center tap can be full wave rectified to heat a 2A3, so there's no need for 2.5V windings.

The (new) Dynaco PA-060 with a 5R4 rectifier and a small-ish cap input filter would be a tempting choice.
 
If you are not using fixed bias, then you will need more than 300V of B+ for the 2A3s. You will need to add the DC voltage drop through the output transformer and the bias voltage from filament to ground. This will make your B+ much more like 380V.

Your PSUD simulation shows that a 200mA rated AC winding would be plenty of current for a monoblock but a little shy for a stereo amp. I would be getting the power supply chokes from Hammond and the power transformer from Sowter. Do also be aware that a 6.3V heater winding with a center tap can be full wave rectified to heat a 2A3, so there's no need for 2.5V windings.

The (new) Dynaco PA-060 with a 5R4 rectifier and a small-ish cap input filter would be a tempting choice.
Aha, now I'm back to being unsure😂 oh dear

tbh I was hoping that I could deal with the heater filiment /biasing stuff later as I really don't understand it yet. My friend here in London had told me that I should go for coleman regulators (aparantly there is one for push pull), I wasn't sure if this also took care of the biasing or not (had been in the pub for a couple of hours and it was all going over my head), there was also talk about 3 pots, one each tube for hum and another for balance. I guess I've got a lot more reading to do.

I had been advised the Hammond chokes weren't up to the task and that the Sowter ones specifically for choke input are what I needed?
 

paulbottlehead

Active Member
OK, so what I would advise that you do is to just build the amp for triode strapped 6AV5s. When you decide you want to change things around, you can make the changes to go to 6B4Gs or swap sockets for 2A3s.

I do not advise using Coleman regulators unless you're reasonably experienced. They do not take care of biasing, they are just for heating the tubes.

With ~380V or so of B+ and a 6.6K load, you're going to run 50mA, and that will be about 60V of bias. That means you have a 600 ohm bias resistor for each pair of tubes with 100mA flowing through it to provide the 60V of bias to the cathode/filament. The plate to cathode voltage sits on top of this, which brings you to 360V, then you have losses in the output transformer that get you to 380V or so.

If you want to use a Hammond choke as a choke input element, go for 10H/500mA and you should be just fine.
 
OK, so what I would advise that you do is to just build the amp for triode strapped 6AV5s. When you decide you want to change things around, you can make the changes to go to 6B4Gs or swap sockets for 2A3s.

I do not advise using Coleman regulators unless you're reasonably experienced. They do not take care of biasing, they are just for heating the tubes.

With ~380V or so of B+ and a 6.6K load, you're going to run 50mA, and that will be about 60V of bias. That means you have a 600 ohm bias resistor for each pair of tubes with 100mA flowing through it to provide the 60V of bias to the cathode/filament. The plate to cathode voltage sits on top of this, which brings you to 360V, then you have losses in the output transformer that get you to 380V or so.

If you want to use a Hammond choke as a choke input element, go for 10H/500mA and you should be just fine.
Thanks for the suggestion, I'll certainly look into it, I have already got the tubes and sockets for the 2a3 though.

Ill give the coleman regulators a miss then as it does look like it's a complication I could do without. I'm certainly not experienced building tube amplifiers from scratch (only kits).

Thanks again Paul.
 
Top