For debugging this, I'd suggest to disconnect the feedback look, and then try to figure out if the problem is around the gain device (lower MOSFETs) or the Aleph current source (upper MOSFETS):
- Disconnect the feedback: remove/disconnect R4.
- Turn on the amp, try to get the DC output voltage to a lowish value.
- Check gain devices: what is the gate voltage of Q7 and Q8? Does it drift? Does the drift correlate with the DC output drift?
- Check the Aleph current source: what is the gate voltage of Q5 and Q6? Dos it drift? Does the drift correlate with the DC output drift?
Hi Russellc, thank you for chiming in.
Tracing the boards makes some sense now. In the configuration I've got them in R27 is essentially open cct.
Placing a short as you suggest puts the 100K into play.
Tracing the boards makes some sense now. In the configuration I've got them in R27 is essentially open cct.
Placing a short as you suggest puts the 100K into play.
I didn't suggest anything, I copied and pasted from 6L6's build guide, which sounds like you didn't read, but carry on.
Russellc
OK, I've placed a shorting link across the R27 position and now the bias is adjusting as we would expect.
The DC offset is still very sensitive.
The DC offset is still very sensitive.
Multi or single turn pot? As I recall from first firing mine up, because of location of capacitors the offset and and Bias or sort of slow to react to the pot.
Unless by sensitivity of the offset you mean you turn, then it reacts going past where you want then when you turn down the same thing happens?
If so turn just a bit and wait. Then turn a little more. If you mean its just hyper sensitive, I don't know.
I would take the time to read builder's guide. Doesn't take as long as it seems, a lot is just banter or irrelevant. skim that and note problems similar to yours, or points you are not clear on and take notes with post number.
Sounds like you are getting closer!
Russellc
I'm using a 10T pot, even if I think about touching it the DC offset shoots off by about 10V. It will adjust but is far too sensitive. That is why I'm going to try a 25T pot in that position.
If you mean R27, isn't that supposed to be 100K? Its been a while since I built mine and I know some things change, I bought a kit from 6L6 a long time ago for my Aleph J. (I'm looking at schematic at beginning of build guide...)
Russelc
Russelc
It's going to drift. It can't have reached any kind of thermal stability yet. 25T would be nice if you have it around, but that's your call. Tiny adjustment - wait for DMM to catch up... tiny adjustment.
With that said, if it's moving 10V something is wrong that a pot swap is not going to fix. Not 10mV, but 10V!? That is not due to the #of turns in the pot and its relative sensitivity.
With that said, if it's moving 10V something is wrong that a pot swap is not going to fix. Not 10mV, but 10V!? That is not due to the #of turns in the pot and its relative sensitivity.
Last edited:
Suggested course of action -
1) Confirm Blk Dynamite's inquiry about the cap values
2) Confirm Russellc's inquiry about your target voltage for bias setting
3) Confirm my inquiry re: actual voltage change at DC offset when you bump pot. (Make sure inputs are shorted)
4) If you can get DC offset to below ~100mV you're good for now. Test other channel. You can finalize tweaking after cooking.
You've got a ton of great help, but it probably feels like people shouting in your ear. So, I'll bow out for a bit.
You're close to making great music. I love that amp!
Congrats on the progress. Looking forward to seeing your success.
1) Confirm Blk Dynamite's inquiry about the cap values
2) Confirm Russellc's inquiry about your target voltage for bias setting
3) Confirm my inquiry re: actual voltage change at DC offset when you bump pot. (Make sure inputs are shorted)
4) If you can get DC offset to below ~100mV you're good for now. Test other channel. You can finalize tweaking after cooking.
You've got a ton of great help, but it probably feels like people shouting in your ear. So, I'll bow out for a bit.
You're close to making great music. I love that amp!
Congrats on the progress. Looking forward to seeing your success.
Last edited:
The potentiometer at R27 is meant to adjust bias current not offset. Without the R27 jumper in place, it's not going to work. You should at least install a wire jumper at the empty R27 position on your board. You may also install a 10k resistor at that position instead, which will allow more variation in adjustment with the trimmer. (That's what I did.)
Yes, I've read the build guide.
I can get the DC offset to sit at about 1V but if I try to get it to zero it becomes unstable.
I can get the DC offset to sit at about 1V but if I try to get it to zero it becomes unstable.
It looks like you may have installed both the R7' trimmer and R7 fixed resistors. That will make it difficult to set the offset correctly. Remove the fixed resistor, and the trimmer should let you adjust the offset very close to zero.
The output monitoring multimeter is just waving at me in slow time. I can get the DC offset down to almost zero but it's constantly climbing to about 1V. The bias is now steady at 400mV over R16.
The soldering around Q2 looks a bit dodgey on the top side. Solder should completely wet the plated through holes and be visible on both sides of the board.
While you are working on the board, use alcohol to clean up the excess flux. That will make things easier to inspect.
While you are working on the board, use alcohol to clean up the excess flux. That will make things easier to inspect.
Good stuff 😀
I'll let TA take over - he's far more experienced than I, but the board looks properly stuffed. 100k pot at R27A. Jumper (now) at R27, and 1k2 at R24.
My suggestion is don't remove or change a thing yet, but YMMV.
See TA's suggestion, but until you run out of travel on the pot, that's not necessarily needed (I think). Has it reached a "click" yet?
Climbing could just mean it hasn't gotten to temp equilibrium yet, but climbing to over a volt is hmmmmm.... ?
Are you able to get the DC offset to settle on both sides of 0V to some extent even if it's not perfectly stable? In other words, can you get both a positive and negative offset? Make sure red DMM lead is on OUT or red speaker terminal and blk DMM lead is on neg speaker terminal (or GND). Try swinging it to one side and the other of 0V0. Make sure to note polarity of offset.
I'll let TA take over - he's far more experienced than I, but the board looks properly stuffed. 100k pot at R27A. Jumper (now) at R27, and 1k2 at R24.
My suggestion is don't remove or change a thing yet, but YMMV.
See TA's suggestion, but until you run out of travel on the pot, that's not necessarily needed (I think). Has it reached a "click" yet?
Climbing could just mean it hasn't gotten to temp equilibrium yet, but climbing to over a volt is hmmmmm.... ?
Are you able to get the DC offset to settle on both sides of 0V to some extent even if it's not perfectly stable? In other words, can you get both a positive and negative offset? Make sure red DMM lead is on OUT or red speaker terminal and blk DMM lead is on neg speaker terminal (or GND). Try swinging it to one side and the other of 0V0. Make sure to note polarity of offset.