Yesterday I was checking the setting of the trimpots on my ACA (which has been working fine for a few months now since I built it) and when I finished noticed that one channel had gone silent. All the wires and solder joints look and feel solid I fear I might have shorted something with the alligator clips of the multimeter. Most likely the neighboring pole of the MOSFET Q1.There is no apparent damage to anything. I also had connected a home made headphone adapter to the speaker outputs and noticed that it turned stereo into mono so it is also an alternative posibility that it (the headphone adapter) has an internal short and shorted the amp in turn.
Where do I begin troubleshooting this? If I burned out something and can anyone suggest the most likely candidate? The mosfet? I'm really pissed at myself.
Where do I begin troubleshooting this? If I burned out something and can anyone suggest the most likely candidate? The mosfet? I'm really pissed at myself.
Here is a chance to sharpen your skills, and members will jump in with help.
First thing is to measure the DC voltage at every point on the schematic you
can get to.
First thing is to measure the DC voltage at every point on the schematic you
can get to.
ACA has opposite polarity on the outputs, because its inverting. Common mistake is to use headphone adapter, which has one common ground. But in reality, you have shorted the outputs.
At the end of the 1.6 build guide is a troubleshooting section. There is an annotated schematic that shows what voltages should be present at multiple points. Follow this guide and you will find the part that needs to be replaced. Just a part or two and all is fixed. We all have broken our little treasures along the path to happiness. Fixing them makes the sadness go away.
Okay! I will get on the voltage measuring but in the mean time I can report that while the led is lit on the left side (the dead side) of my amp there is no heat being generated while idling. Thanks @Nelson Pass for chiming in on my little catastrophe, I'm honored, and @adason I think you put your finger on what I may have done. If you are correct what kind of damage would result?
@Signal Lost thanks for the additional guidance!
@Signal Lost thanks for the additional guidance!
Here are the voltages of my measurements on the dead side of my ACA at the points indicated on the troubleshooting section of the ACA 1.6 build guide. Quite a few are at variance with the ranges specified in the troubleshooting image. Just for forms sake I did the measurements on the "live" side of the amp and all those were more or less in the specified ranges (only variances of a few tenths of a volt and I had not let the amp fully warm for a 1/2 hour as specified.
Q1 S= 0, D=0, G=7.3
Q2 S=20.6, D=23.7, G=23.5
Q4 D=23.7
R10= 5.5
R9=7.3
R15 B=20.6 E=20.6
R5 = 23.5
I eagerly await your diagnosis. If you need measurements from "all" points I can do that too but I got the idea from the troubleshooting section that these were the important ones.
Q1 S= 0, D=0, G=7.3
Q2 S=20.6, D=23.7, G=23.5
Q4 D=23.7
R10= 5.5
R9=7.3
R15 B=20.6 E=20.6
R5 = 23.5
I eagerly await your diagnosis. If you need measurements from "all" points I can do that too but I got the idea from the troubleshooting section that these were the important ones.
In context: {attached}
Two pairs of points look "impossible". The bottom of R15 is either zero or 20V?? (Or am I mis-reading?) The voltages around R1 R2 R3 R4 imply 35 Amps of current which is so not-likely. Without a look, I am wondering if you maybe burned the copper off the board? Or all four R1-R4?
Attachments
Last edited:
Open connection to the drain of Q1 or Q1 is open? (Mosfets usually short unless a bond wire melts first)
@prr I double checked and there is also 20v at R3 and R4 so would the logical conclusion be that the ling to Q1 S is broken ie most likely a bad solder joint or less likely a break in the cirucuit? I think I will detach the board from the heat sink and see what I can see if you agree. Or could an internal problem in Q1 cause the 0v reading on S and D? I'm thinking not, but I'm a novice. Thanks for your help!
> see what I can see if you agree.
I am always in favor of bright light and strong magnifier. See what you can see.
I am always in favor of bright light and strong magnifier. See what you can see.
You also can ask your self "how would the voltage on Q1 become the numbers you have measured?"
if the black probe is on ground and you touch the red probe to it it will read 0V because they are connected directly (no parts in between). If you get the same readings at other locations it could mean they are connected directly (shorted).
My instructors drilled into us "follow the voltage" after we had verified the continuity of adjacent items. 0v on Q1 can mean it is connected to ground or it is not connected to the voltage supply
if the black probe is on ground and you touch the red probe to it it will read 0V because they are connected directly (no parts in between). If you get the same readings at other locations it could mean they are connected directly (shorted).
My instructors drilled into us "follow the voltage" after we had verified the continuity of adjacent items. 0v on Q1 can mean it is connected to ground or it is not connected to the voltage supply
The gate on q1 is being told to conduct and the 20.6v points show that it is not. The q1 circuit is open, either inside or outside the fet.
Problem solved! When I lifted the board from the heat sink and did some voltage testing I noticed that the center post (the drain) of Q1 went from 0v to 12.3v when I poked at it with the probe. I also noticed that the amount of solder was not as generous as it could be so a little touching up with the soldering iron got the juice flowing reliably. While I had it off the heatsink I also trimmed a few wire stems that were a little over long on the bottom side of the board and may have presented a danger of touching the heat sink when mounted.
So in summary what I think happened was that the solder joint on the drain pin of Q1 was good enough to make the amp work for 3 months but when I checked the balance by attaching an alligator clip to that pin it jarred it enough to break the connection.
Eureka, the sweet music is flowing again! Thank you @PRR, @Nelson Pass, @Signal Lost and @kevinkr!!!
So in summary what I think happened was that the solder joint on the drain pin of Q1 was good enough to make the amp work for 3 months but when I checked the balance by attaching an alligator clip to that pin it jarred it enough to break the connection.
Eureka, the sweet music is flowing again! Thank you @PRR, @Nelson Pass, @Signal Lost and @kevinkr!!!
Congratulations. You are now a member of FAB.
I thought about it and realized that it could not have been open inside the Fet.
I thought about it and realized that it could not have been open inside the Fet.
You should see the Terminology they hit you with in plant taxonomy
I actually did not understand what you meant by “open“ regarding Q1 and what the consequences of that were. And while I’m confessing ignorance, what is FAB?Thanks again for your help and indulging a biologist who kind of enjoys being in over his head in a different subject.🙂
Bruce Barbour
PS I’ve watched your interviews with Steve Guttenberg and I like your take on your role in the industry as an “entertainer“, right on! There is a bit too much seriousness in certain parts of the audio file world.
I actually did not understand what you meant by “open“ regarding Q1 and what the consequences of that were. And while I’m confessing ignorance, what is FAB?Thanks again for your help and indulging a biologist who kind of enjoys being in over his head in a different subject.🙂
Bruce Barbour
PS I’ve watched your interviews with Steve Guttenberg and I like your take on your role in the industry as an “entertainer“, right on! There is a bit too much seriousness in certain parts of the audio file world.
FAB
Fearless Amplifier Builder
ZM being Omnispecial, again - so sole CAB around
Chicken Amplifier Builder
Fearless Amplifier Builder
ZM being Omnispecial, again - so sole CAB around
Chicken Amplifier Builder