I’m planning a BA3 complimentary build, and am interested in experimenting with the difference between adding imbalance/distortion in the FE (P3) versus the OS (changing source resistor values.) I have the BA3 built as a pre right now and have enjoyed making tweaks to P3. I’ve also messed around with altering source resistor values in my F6, though I have more exploring to do there.
Question(s): what would be a good starting point for source resistor values if I want to introduce a bit of negative phase H2? Thought it could be fun to use sockets in the source resistor positions to make playing with values a bit easier. Or is this whole idea moot since I can adjust imbalance in the front end?
If you want to introduce H2 in the output stage, you need to intentionally imbalance the source resistors... for example using 0.47 in the lower half and 0.56 in the upper.
I don’t suggest this with the BA-3, but it could be done. Absolutely give it a try in the F6, as the single device per phase makes it simpler.
(In fact, if you look at the F6, you’ll see that it’s already been done!!
Why? To sweeten it up a little, I suspect...)
However, in the case of the BA3 since it has the adjustable front-end, with P3, is where you want to do the distortion trick. It’s simpler, easier, and more straightforward. And you can swap phases if you like. Just use 0.47ohm source resistors in both sides of the output stage and smile. :
I don’t suggest this with the BA-3, but it could be done. Absolutely give it a try in the F6, as the single device per phase makes it simpler.
(In fact, if you look at the F6, you’ll see that it’s already been done!!
Why? To sweeten it up a little, I suspect...)However, in the case of the BA3 since it has the adjustable front-end, with P3, is where you want to do the distortion trick. It’s simpler, easier, and more straightforward. And you can swap phases if you like. Just use 0.47ohm source resistors in both sides of the output stage and smile. :
You can control the bias to whatever you want because of the adjustability designed into the circuit. So heat has nothing to do with it.
The original circuit, way back when Nelson gave away a bunch of test mule chassis at Burning Amp fest 2009(?) and they had unmatched devices, so the 1R source resistors helped equalize things a little better, as well as this being a zero feedback design, so a little more source degeneration isn't a bad thing.
I'd use 0.47 and most of the recent builders I've been talking to with BA# have done so with great results. But it really doesn't matter that much in the big picture.
The original circuit, way back when Nelson gave away a bunch of test mule chassis at Burning Amp fest 2009(?) and they had unmatched devices, so the 1R source resistors helped equalize things a little better, as well as this being a zero feedback design, so a little more source degeneration isn't a bad thing.
I'd use 0.47 and most of the recent builders I've been talking to with BA# have done so with great results. But it really doesn't matter that much in the big picture.
Could use some advice. I'm worried I did a dumb.
I am building balanced complementary BA3 monoblocks.
I started to bias the FE on one of them last night. Started with variac and everything looked good so went to up to 0.75v slowly. I actually have a BUNCH of DMMs so I was doing both channels (or in the case of balanced: phases) at the same time.
I noticed that as soon as I got offset to 0 on one phase, the other would go up. I was puzzled so I read the balanced BA3 thread more, and did a search and found out that you actually need to bias the FE in the nominal SE/unbalanced configuration (R5s @ 100ohm going to ground instead of both R5s connected to each other @ 2x100ohm). No problem, I un-bridged R5 and put them back to ground.
Now I am trying to bias them after this change, but when I power up with variac, if I get to full mains (120v) I am see 60+ VDC across R10 and R11. I'm an idiot and I forgot to dial P1 and P2 back to zero before re-attempting bias with the new configuration. Upon noticing the high VDC I immediately powered off, and turned P1 and P2 back to zero. But still, when I power up with variac I am seeing WAY too much VDC on R10 and R11 even before full mains. And when I turn P1 and P2 I don't see it having ANY affect on VDC AT ALL.
I didn't see/smell any magic smoke. No blown fuses. I rechecked PSU output and it's nominal 32v. I don't see anything toasted either.
Does anyone have any advice? Did I fry my Q3 & Q4s?
Just to re-confirm: I want my DMM probes on both sides of R10 and both sides of R11 and then DMM + on R12 before cap and - on ground? So that's 3 DMMs, one for R10, one for R11 and one for R12/ground (offset).
Thanks in advance for all the help, and sorry for being dumb!
I am building balanced complementary BA3 monoblocks.
I started to bias the FE on one of them last night. Started with variac and everything looked good so went to up to 0.75v slowly. I actually have a BUNCH of DMMs so I was doing both channels (or in the case of balanced: phases) at the same time.
I noticed that as soon as I got offset to 0 on one phase, the other would go up. I was puzzled so I read the balanced BA3 thread more, and did a search and found out that you actually need to bias the FE in the nominal SE/unbalanced configuration (R5s @ 100ohm going to ground instead of both R5s connected to each other @ 2x100ohm). No problem, I un-bridged R5 and put them back to ground.
Now I am trying to bias them after this change, but when I power up with variac, if I get to full mains (120v) I am see 60+ VDC across R10 and R11. I'm an idiot and I forgot to dial P1 and P2 back to zero before re-attempting bias with the new configuration. Upon noticing the high VDC I immediately powered off, and turned P1 and P2 back to zero. But still, when I power up with variac I am seeing WAY too much VDC on R10 and R11 even before full mains. And when I turn P1 and P2 I don't see it having ANY affect on VDC AT ALL.
I didn't see/smell any magic smoke. No blown fuses. I rechecked PSU output and it's nominal 32v. I don't see anything toasted either.
Does anyone have any advice? Did I fry my Q3 & Q4s?
Just to re-confirm: I want my DMM probes on both sides of R10 and both sides of R11 and then DMM + on R12 before cap and - on ground? So that's 3 DMMs, one for R10, one for R11 and one for R12/ground (offset).
Thanks in advance for all the help, and sorry for being dumb!
Hi. I'm a total beginner in the diy scene, and was looking at building a BA-3. I'm a little confused by what combinations of chassis size, rail voltage, and output power correlate with each other.
Since the "3U 400mm Mini Dissipante" has the same heatsink area as the ultra-common "4U 300mm Dissipante," would it be wrong to use the shorter and narrower chassis in a build confirmed working with 4U 300mm? My gut tells me there's nothing wrong, but I find it odd that the cheaper, and (to me) more appealing form-factor chassis is rarely used.
If I used a 3U 400mm or a 4U 300mm (which from my current understanding are equivalent in heat dissipation capacity), could I use a 24V Antek Toroidal for approximately 32V rails, or would I be thermally limited to an 18V transformer for 24v rails?
While 6L6's initial post states 32V rails produce about 40w into 8 ohms, I see mention of around 50w into 8 ohms. Is the circuit simply under-rated, or does this higher power figure require more output devices than the standard 6 per channel?
Sorry for all the crazy questions, and I apologize ahead of time if these are redundant. Thank you.
Since the "3U 400mm Mini Dissipante" has the same heatsink area as the ultra-common "4U 300mm Dissipante," would it be wrong to use the shorter and narrower chassis in a build confirmed working with 4U 300mm? My gut tells me there's nothing wrong, but I find it odd that the cheaper, and (to me) more appealing form-factor chassis is rarely used.
If I used a 3U 400mm or a 4U 300mm (which from my current understanding are equivalent in heat dissipation capacity), could I use a 24V Antek Toroidal for approximately 32V rails, or would I be thermally limited to an 18V transformer for 24v rails?
While 6L6's initial post states 32V rails produce about 40w into 8 ohms, I see mention of around 50w into 8 ohms. Is the circuit simply under-rated, or does this higher power figure require more output devices than the standard 6 per channel?
Sorry for all the crazy questions, and I apologize ahead of time if these are redundant. Thank you.
So I've just completed a BA-3 build, and I'm already stuck on biasing the front end. I iteratively increase p1 and p2, until I start to see voltage across r10, but p1 is the only pot that does anything at all. p2 makes small, insignificant changes in both dc offset on r12 and bias voltage. P1 will bring r10 up to 1v, but with 10v dc offset. Meanwhile, r11 barely can get up to 20mV. This is the same for both channels. I have built this gain stage EXACTLY as per 6l6s guide, using +/- 32V PSU. Can someone please lend me some advice? I'd be glad to provide any measurements/info you might need to help me. Thanks in advance.
R3/R4 at 60mV means the jfets are conduction 6mA so they are working
though that current is a bit lower than the 8mA on the schematics.
You probably had 8mA idss parts to begin with.
Please check parts placement: Q3 is p channel mosfet (2sj313)
Q4 is n-channel (2sk2013)
In your easier post, did you accidentally switch P1 and P2?
P1 should control the voltage you see across R11 while
P2 should do the same for R10.
You wrote that the highest you got across R11 was 20mV. Assuming
you did that by adjusting P1, did that voltage start at 0 and then
only hit 20mV at the end of the trimmer's adjustment range?
What Vgs voltage do you measure for Q4 when you hit 20mV across R11?
Perhaps you can take some and post some clear photos of just the
front end board and see if anyone spots something.
though that current is a bit lower than the 8mA on the schematics.
You probably had 8mA idss parts to begin with.
Please check parts placement: Q3 is p channel mosfet (2sj313)
Q4 is n-channel (2sk2013)
In your easier post, did you accidentally switch P1 and P2?
P1 should control the voltage you see across R11 while
P2 should do the same for R10.
You wrote that the highest you got across R11 was 20mV. Assuming
you did that by adjusting P1, did that voltage start at 0 and then
only hit 20mV at the end of the trimmer's adjustment range?
What Vgs voltage do you measure for Q4 when you hit 20mV across R11?
Perhaps you can take some and post some clear photos of just the
front end board and see if anyone spots something.
I may have confused p1 with p2 with regards to which one was having an effect on voltage across mosfet source resistors. Not at home right now, but I'll check on this as soon as I am able.
As for parts placement, I have double-checked that everything is correctly installed, p-channel, n-channel, proper orientation, etc...
And yes, the 20mV I saw across R11 only came at the very end of the pot travel, it remained 0.000 until a few turns before the end. I have some 1k pots I can swap in for the 500ohm, would it be worthwhile trying that?
This is my first time attempting anything like this, so please bear with my apparent confusion. I will update with answers to your other questions and clearer photos later today.
Thanks so much for the help, I truly appreciate your time.
As for parts placement, I have double-checked that everything is correctly installed, p-channel, n-channel, proper orientation, etc...
And yes, the 20mV I saw across R11 only came at the very end of the pot travel, it remained 0.000 until a few turns before the end. I have some 1k pots I can swap in for the 500ohm, would it be worthwhile trying that?
This is my first time attempting anything like this, so please bear with my apparent confusion. I will update with answers to your other questions and clearer photos later today.
Thanks so much for the help, I truly appreciate your time.
IT WORKED. You are the man, Dennis. I can't thank you enough. I simply replaced the 500ohm p2 with a 1k device. I guess it just went against my intuition to need different resistances on the two halves of this circuit, but I assume this makes up for differences in the manufacture of the semiconductors?? Again, please forgive my ignorance. This has, and surely will be, the start of a long and fun learning process. What a great feeling to be able to balance this stage and get on my way to a fully functional amplifier.
Stay tuned (if you dare) for many more difficulties to come as I proceed!
Stay tuned (if you dare) for many more difficulties to come as I proceed!
Happy you got it working. 
There are variations in the parts, first within each polarity. Then there are
differences between the polarities. Even with complementary parts you
will need to match and select to get both sides to behave identically. In
your case I think it's a combination of the slightly low current for the
jfets combined with the higher turn-on voltage for the 2sk2013 that
caused the issue you encountered.
Have fun with the rest of your build.
There are variations in the parts, first within each polarity. Then there are
differences between the polarities. Even with complementary parts you
will need to match and select to get both sides to behave identically. In
your case I think it's a combination of the slightly low current for the
jfets combined with the higher turn-on voltage for the 2sk2013 that
caused the issue you encountered.
Have fun with the rest of your build.
Power supply for BA-3
Hi all,
So I've got the BA-3 almost right where I want it, and it sounds great. My only problem now is (hopefully) a simple one. But, as a novice in this world, I think I still need a little help from the group. I'm using a Hammond 600VA toroid with dual 24v secondaries and the DIYaudiostore PSU v3.0 built almost exactly as per the 6L6 build guide. With line voltage at 120, the rails are +/- 36.5v. Taking heed of the warning about destroying the JFETs on the front end board, I'm running the amp at 111vac through my variac, which results in exactly 32vdc rails.
Can someone please offer some guidance for how best to deal with this issue? My intuition tells me to increase the resistance value of the 3 watt resistors on the PSU board (I'm using 0.47ohm), but I'm not sure by how much, and whether that will be effective. Any help would be greatly appreciated. Thanks, all.
Hi all,
So I've got the BA-3 almost right where I want it, and it sounds great. My only problem now is (hopefully) a simple one. But, as a novice in this world, I think I still need a little help from the group. I'm using a Hammond 600VA toroid with dual 24v secondaries and the DIYaudiostore PSU v3.0 built almost exactly as per the 6L6 build guide. With line voltage at 120, the rails are +/- 36.5v. Taking heed of the warning about destroying the JFETs on the front end board, I'm running the amp at 111vac through my variac, which results in exactly 32vdc rails.
Can someone please offer some guidance for how best to deal with this issue? My intuition tells me to increase the resistance value of the 3 watt resistors on the PSU board (I'm using 0.47ohm), but I'm not sure by how much, and whether that will be effective. Any help would be greatly appreciated. Thanks, all.