This is what I keep wondering as I read this board. Swapping boards from one side to the other could reveal a problem not related to the board itself.
The process of elimination may lead to unexpected places.
I actually went through this with my Jensen modified Bulwark boards, after blowing the first set of Elna coupling caps. Took days to find the final failed components, simply because they didn’t fit my ideas of where they should have been.
I actually went through this with my Jensen modified Bulwark boards, after blowing the first set of Elna coupling caps. Took days to find the final failed components, simply because they didn’t fit my ideas of where they should have been.
As a noob, I can advise you on such a debugging method. Maybe it's worth going back to the beginning:
1. disconnect everything from the power supply
2. Check the voltages on the secondary windings of the transformer
3. Check the voltages across the bridge rectifiers
4. Check the voltage at the output from the power supply.
5. Disconnected daughter cards and without them check "good" amplifier channel by taking measurements. Disconnect the good one and connect the "bad" amp and measure again. Do this according to MJ's advice
6. If everything is ok, connect both amplifier boards and take the measurements.
If anything is wrong, don't go to the next step.
1. disconnect everything from the power supply
2. Check the voltages on the secondary windings of the transformer
3. Check the voltages across the bridge rectifiers
4. Check the voltage at the output from the power supply.
5. Disconnected daughter cards and without them check "good" amplifier channel by taking measurements. Disconnect the good one and connect the "bad" amp and measure again. Do this according to MJ's advice
6. If everything is ok, connect both amplifier boards and take the measurements.
If anything is wrong, don't go to the next step.
The logic problem is that the right channel works and has always done so. This suggests the PSU must be OK. Both VDCs re constantly checked and register 25.88 +/- to the two boards inputs. Am using a very powerful rectifier (60 amp) that you recommended but hope that is not an issue.
By "load" you mean?
By "load" you mean?
Post 5395 is where you mention three transformers, but might have meant three transistors. Which makes me ask do you have the correct transistors? Lots of great advice here. Hopefully you will build a dim bulb tester to help you find the problem. My earlier advice about measuring all possible is because this fault may be like TA's front end problem and not be near the smoke.
TA: No load and the smoking still happens. It's the same with a meter attached as load or without -- smoking at 1:44.
KevinHeem: Still haven't built one yet and am trying to work through the concept. Hoped that the new pbc board would solve the issue, but it hasn't.
Have been wondering if the very powerful rectifier (60 amp) could, somehow, be contributing to the issue but this seems unlikely as the right channel never smokes.
Obviously, there is the question of the chassis and heatsink and I have looked everywhere for any difference between the left and right sides (and builds) and can find none. They are exactly the same. The heatsink rails are treated the same way and never touch the pcbs. these are connected to the heatsinks with the same 7 standoffs and the mosfets are also connected exactly the same way.
KevinHeem: Still haven't built one yet and am trying to work through the concept. Hoped that the new pbc board would solve the issue, but it hasn't.
Have been wondering if the very powerful rectifier (60 amp) could, somehow, be contributing to the issue but this seems unlikely as the right channel never smokes.
Obviously, there is the question of the chassis and heatsink and I have looked everywhere for any difference between the left and right sides (and builds) and can find none. They are exactly the same. The heatsink rails are treated the same way and never touch the pcbs. these are connected to the heatsinks with the same 7 standoffs and the mosfets are also connected exactly the same way.
Search "dim bulb tester" on YouTube. Plenty of videos that will walk you right through from concept to completion.
Should have been build a DBT before you put together that fancy test bench. Don't mean to beat up on you...I did the same thing.
Should have been build a DBT before you put together that fancy test bench. Don't mean to beat up on you...I did the same thing.
Signal lost: Yes, my mistake meant 3 transistors. Now the 2 identical transistors are ordered from Mouser under the 595-LM385LPR-1-2 designation but on the transistor itself, they say 95ME2HE, 385-12 T1. Hope these are the correct ones. The optoisolator is always checked carefully for polarity.
KevinHeem: Have spent a lifetime of doing things out of order and now, in retirement, am regressing back to my youth. Have reached about 23, I think...
KevinHeem: Have spent a lifetime of doing things out of order and now, in retirement, am regressing back to my youth. Have reached about 23, I think...
@Craigl59
This post from a while back comes to mind:
https://www.diyaudio.com/forums/pass-labs/241729-aleph-illustrated-build-guide-357.html#post5738959
I have not tried this myself but it could be a way to test your "bad" board without having to pull it in and out of the chassis.
This post from a while back comes to mind:
https://www.diyaudio.com/forums/pass-labs/241729-aleph-illustrated-build-guide-357.html#post5738959
I have not tried this myself but it could be a way to test your "bad" board without having to pull it in and out of the chassis.
Got one measurement tonight. Started the unit both boards running and checked the voltage to each. The VNeg was steady at -25.4 for both boards. But the VPos started out around +25 volts then decreased down -- stopped measuring at around 22 vdc, also for both boards. Is this because there is no load? Would have thought that they would all be the same. Let me know.
Kevin Heem: Read the posts; thanks, have got some 47ohm 3 watt resistors lying around. Think I should try this with, say, the old board and check to see if there is no smoking? If so, does that suggest the heatsink is not "sinking." (LOL). Whoops! Checked and they are 1/4 watt. So no go here yet.
Last edited:
Not trying to spend your money for you, but if you had a second Tack Life power supply you could connect the two as a bipolar supply and bypass your external PSU altogether.
You mean one as + and one as -? Like the Tack Life a bunch but only use it to power transport motors.
Something is not right on your “+” supply if it is more than 1V different then your negative and even worse if it fluctuates with no load attached.
Take it very slow and break into pieces - as been said before
1) transformer AC voltages
2) rectified PSU +\-DC voltages
3) then +\-DC voltages with amp boards attached
4) if steps 1-3 are correct, then check bias and offset
bullittstang: Will do this tomorrow morning. Have checked the PSU outage to the amp splitter without boards attached and it is very even and steady at right around +/- 25.8 vdc. Only when the boards are connected does the phenomenon noted above occur.
Best to get some rest and go again with fresh eyes.
Be sure and check your grounds, that could pull your voltage if not solid.
Be sure and check your grounds, that could pull your voltage if not solid.
The power supply is only converting voltage with no load attached. It becomes a power supply when connected to a load. So you are only checking that it converts to the correct voltage not that it can supply power.