Heatsink question
- Thread starter Betomon
- Start date
I agree!
Okay so there was some strange behavior when testing.. read on..
A: -85.6mV
G: .894V
E 1V
J 1V
L -263mV
F .91V
K -1.51V
H 2.15V
So, something interesting is going on around the A path. I tested by going to the leads of R44. If I test on the lead touching C13, it tests ~87mV. If I test it on the other lead, the one going to C22, it causes protection to come on immediately (audible clicking of relay and "standby" light comes on as soon as I touch the lead with my meter). The meter was hooked up to ground on the negative lead as usual.
I repeated that test on the good board, and, it shows 0mV at R44 by C13 side, and 23mV at R44 by C22 side.. and no protection. Something is very wrong in the bad board then?? But what??
For kicks I measured the emmiters of Q14 and Q13.. they both were right around -86mV - I did replace Q14 though. I suppose it's possible I screwed something up?
Q4 is still wrong, you have to measure the voltage on the base of Q4. But already E=1V and J=1V, check the base-emitter whether it's short.
Betomon
Active Member
Betomon
Active Member
So the readings I took before but at 162Ohm on R34. Okay will do. And I'll check Q4 again
Q4 is fine, you gave me the wrong number on the first try. Double check the numbers before you post. I wasted a lot of time on this because when the Q4 still wrong after changing, I had to re-analyze the whole thing in case I was wrong.
we spent a lot of time chasing around and around because of the missing voltage. You need to make sure you double check carefully.
we spent a lot of time chasing around and around because of the missing voltage. You need to make sure you double check carefully.
Last edited:
Betomon
Active Member
You're absolutely right I sincerely apologize. I get a bit confused with the mirroring of this board, and having to flip it around to measure from the underside. I will triple check all measurements now.
Betomon
Active Member
Okay, I measured again, this time checking.. Here we go:
At maximum resistance across VR36 (168-169)
A: -83mV (Both emitters show the same, as does point of C13 and R44 which are on the same path.
G: 1.486V
J 1.27V
E .67V
L .64V
K -1.22V
F -.61V
H 1.46V
No discernible current/voltage through the big resistors..
At maximum resistance across VR36 (168-169)
A: -83mV (Both emitters show the same, as does point of C13 and R44 which are on the same path.
G: 1.486V
J 1.27V
E .67V
L .64V
K -1.22V
F -.61V
H 1.46V
No discernible current/voltage through the big resistors..
Again, this really does not make sense again.
Check "M" and "N" for me to verify.
Please double check all the numbers written on the schematic here. Verify the numbers. BLUE is for R34=1ohm, GREEN is for R34=162ohm.
You need to be more careful in reading the numbers and verify. This should have been done and fixed long time ago if you have not missed the reading and the voltage. You measure, you cannot assume anything, that you have voltage, you have connectivity or you have ground, you have to verify step by step. Like if I ask for the voltage of collector of Q4, you automatically measure the collector of Q28 and R33 that they are have the same voltage to prove they are connected together. You always check the voltage of components that connected the the rail or ground to prove that it is actually connected.
This should be fixed in two or three go around. Here we are, how many pages we gone through already, still have no amp to show for. I am running out of patience in the thread, I spent so much time on this analyzing just to find you missed something. Read my response and what I ask you to measure carefully. Every reading I have to ask two to three times. Like I asked you to read both R34 = 1ohm and 162ohm, you gave me only 1ohm first, then you miss read one voltage. I am not going to work on this much longer.
This might sound harsh, it does not matter whether you are mechanical, electrical or any engineering, you have to read things carefully, measure carefully, double verify. Use common sense to take precaution and be careful. You just make too many mistakes, too many incomplete answer.
I suggest you go back and read this thread, step by step, how I come up with the resistor value for you, how I experiment to see whether your heatsink is good enough. I am not going to repeat over and over. It's up to you now.
Sorry if I offended you, but this is true in real life job and work.
Check "M" and "N" for me to verify.
Please double check all the numbers written on the schematic here. Verify the numbers. BLUE is for R34=1ohm, GREEN is for R34=162ohm.
You need to be more careful in reading the numbers and verify. This should have been done and fixed long time ago if you have not missed the reading and the voltage. You measure, you cannot assume anything, that you have voltage, you have connectivity or you have ground, you have to verify step by step. Like if I ask for the voltage of collector of Q4, you automatically measure the collector of Q28 and R33 that they are have the same voltage to prove they are connected together. You always check the voltage of components that connected the the rail or ground to prove that it is actually connected.
This should be fixed in two or three go around. Here we are, how many pages we gone through already, still have no amp to show for. I am running out of patience in the thread, I spent so much time on this analyzing just to find you missed something. Read my response and what I ask you to measure carefully. Every reading I have to ask two to three times. Like I asked you to read both R34 = 1ohm and 162ohm, you gave me only 1ohm first, then you miss read one voltage. I am not going to work on this much longer.
This might sound harsh, it does not matter whether you are mechanical, electrical or any engineering, you have to read things carefully, measure carefully, double verify. Use common sense to take precaution and be careful. You just make too many mistakes, too many incomplete answer.
I suggest you go back and read this thread, step by step, how I come up with the resistor value for you, how I experiment to see whether your heatsink is good enough. I am not going to repeat over and over. It's up to you now.
Sorry if I offended you, but this is true in real life job and work.
Last edited:
I go to a new post, right now, I see M and N change when you adjust R34. This CANNOT be happening. This is only two 1N400x diode drop with 3mA running through it.
Also voltage across Q7 change with R34. You need to remeasure the whole thing again, compare the BLUE and GREEN numbers to verify those are correct before I do anything else. Those are the numbers I got from you from the last few posts.
Also voltage across Q7 change with R34. You need to remeasure the whole thing again, compare the BLUE and GREEN numbers to verify those are correct before I do anything else. Those are the numbers I got from you from the last few posts.
Betomon
Active Member
Okay I'm gonna go through everything. Don't worry you have not offended me at all. I'll take my time based on what you've said and try to figure this out. I truly appreciate all the help.
Hopefully once the amp is stable I can post back results..
Betomon
Active Member
OK so I spent more time analyzing things, and I came up with an issue centered around Q6. So I changed it. Afterwards, I'm getting proper bias readings.. But with a large offset. So I changed q5 to try to see if it made a difference. Helped but there is still an offset, or, difference, of about 7 mv. R69 is at 20mv and r78 is at 12 as am example.
I also changed Q13 just to match the other side.. Plus I figured it couldn't hurt if I didn't screw up.
Then, I started taking readings and came across an issue where voltages were changing upon different startups.. This may explain why my numbers were inconsistent.. So I looked at a couple of caps and sure enough they were leaky.
One was totally destroyed once I took it out.. C22. it had green crud at the bottom, and the lead was barely on. This is part of the protection circuit but is in line with A reading of before, so maybe it had a larger effect.
Because of that leaky cap, I suspected the others to be bad so I changed them. Others on the power supply also (because I was getting inconsistent rail Voltages.. 48-53V.. Upon startups..)
After all this, the amp behaves much better. I now get constant voltages, proper relay delay on both channel, and pretty close readings vs the good board. Also, it now does not shut down at high volumes. I have tested a dummy load and a spare woofer just for kicks. It has meat like never before!
I may finally be approaching the point of the initial question? I sure hope so!
Thanks for your patience with me Alan
I don't think Q6, Q5 and Q13 is the problem. But if you rather change all the transistors instead of finding the problem, it's your choice. If you measure the voltage at all the points you should be able to see where the problem is, I don't think leaky is the problem in the amp section.
C22 is the relay circuit, we never get to it because we never find the problem. I am glad you fix it.
C22 is the relay circuit, we never get to it because we never find the problem. I am glad you fix it.
ghazzer
Senior Member
Please note that line voltage is not a constant. PERIOD.
I am going to assume that you aren't connected to a sub-station with a dedicated line (sarcasm).
When line voltage changes, guess what? Your rail voltage will change accordingly. When your rail voltage changes so will every other voltage. The level of change is usually a function of the number of people and businesses on your leg, time of day, etc. Think about what happens at about dusk: a whole bucking funch of sensors say "Hey. It's getting dark. Time to turn on all the security lights." The electric company does a decent job of monitoring the line voltage and making adjustments to keep it within tolerance, but it is rarely 120.000VAC (or 115.000VAC).
One habit I try to keep is to record the rail voltage at the top of the notes I keep for every session of troubleshooting, and check it from time to time and write each value in the margin. That way I am not surprised when the drop across R123 reads 4.5vdc and later reads 4.9vdc.
Just another thing you have to keep track of - - -
I am going to assume that you aren't connected to a sub-station with a dedicated line (sarcasm).
When line voltage changes, guess what? Your rail voltage will change accordingly. When your rail voltage changes so will every other voltage. The level of change is usually a function of the number of people and businesses on your leg, time of day, etc. Think about what happens at about dusk: a whole bucking funch of sensors say "Hey. It's getting dark. Time to turn on all the security lights." The electric company does a decent job of monitoring the line voltage and making adjustments to keep it within tolerance, but it is rarely 120.000VAC (or 115.000VAC).
One habit I try to keep is to record the rail voltage at the top of the notes I keep for every session of troubleshooting, and check it from time to time and write each value in the margin. That way I am not surprised when the drop across R123 reads 4.5vdc and later reads 4.9vdc.
Just another thing you have to keep track of - - -
No, not true, the last problem is the bias spreader adjustment affect the rest of the circuit. I don't care where the rail is 52V or 60V or 40V, it should not happen like that. This kind of transistor circuit run on current mode, you set up the current, changing the voltage does not change much within the operating limit.
To be exact, the current of the whole power amp section is set by CR1, 2 5 and 6. It does not change much regardless of the rail voltage. The voltages I asked to measure change less than a few mV even if you vary the rail from 40 to 80V.
To be exact, the current of the whole power amp section is set by CR1, 2 5 and 6. It does not change much regardless of the rail voltage. The voltages I asked to measure change less than a few mV even if you vary the rail from 40 to 80V.
Last edited:
Betomon
Active Member
You're absolutely correct.. Thank you for that tip, I will keep it handy next time I measure.
BTW Alan I haven't given up on getting this right.. I've just been totally swamped with my day job. I did have the chance to test voltages, and only noticed negligible differences between boards. I know with certainty that there exactly a 7mV difference between R68 and R78. Goes up and down with VR36 always keeping that delta.
I vaguely remember there being a different voltage drop between two diodes (comparing boards).. one of the ones you mentioned.. but, I have to go back and gather my notes. Or better yet start from scratch with new notes, perhaps tomorrow or once I find more than a couple of hours of time..
Betomon
Active Member
Today I spent the morning measuring voltages around the amp.. and.. I have some interesting findings. I began measuring resistors to see if anything seemed off when compared to the good board. That led me to find the whole "positive" side to be off.. And it appears the problem lies around Q7. See below:
Bad board / Good board - Both with R36 fully counter-clockwise - Line Voltage this session 120.5V Rail voltage this session 51.5V (I turned off the unit in the middle of testing to flip board upside down, then on again. The rail voltage changed about .6V) Measurements taken with the ground lead attached to the chassis ground.
R28 -.56V / -.71V
R27 -.65V / -.63V
R34 1.07V / .76V (!)
R35 -.57V / -.72V
R75 .47V / 0mV (!)
R74 .47V on the side touching E of Q12, and -17.7mV on the other side, which connects to the E of Q11. This seemed very odd to me?? R74 on Good board was -160mV
R83 -17.1mV / -160mV
R26 .65V / .68V
Because of the above I began measuring the associated transistors again:
Emitter of Q12 .47V / 160mV
Emitter of Q11 -.15.2mV / -160mV
Collector of Q12, Q11 - Exactly the rail voltage on both
Base of Q12 1.07V / .77V
Base of Q11 -.57V / -.72V
Collector of Q6 1.08V / .77V
Collector of Q5 -.57V / -.72V
Base of Q7 113mV / -69.3mV
Collector of Q7 1.08V / .77V
Emitter of Q7 -.57V / -.72V
So, that explains the difference in bias. I'm currently trying to figure out why/how Q7 is putting out such different voltages? Could it be because of Q11/Q12 not matching? Those are new transistors.. Not sure why or how any of this is happening? All the resistors i've checked (and I've checked many) are at the right resistance.
Read post 180, This is not what I asked. Your reading is COMPLETELY USELESS!!! Your problem is why things change when you adjust the bad channel.
You don't read my instructions, I cannot help you anymore. this should be fixed in 10 posts if you could have answered my question and not giving wrong reading. When we got close, you refused to follow post 180 and you went and change all the transistors against what I told you not to. This is at post 198 and going no where!!! I think I gave you very clear and simple instructions. I think I spent a whole lot more time on this thread than any other ones I helped. AND is going nowhere. You are on your own now. I am not going to start doing calculation for you, it's a new ballgame again now that you change everything.
The reading above is normal other than the bias is set too low.
You don't read my instructions, I cannot help you anymore. this should be fixed in 10 posts if you could have answered my question and not giving wrong reading. When we got close, you refused to follow post 180 and you went and change all the transistors against what I told you not to. This is at post 198 and going no where!!! I think I gave you very clear and simple instructions. I think I spent a whole lot more time on this thread than any other ones I helped. AND is going nowhere. You are on your own now. I am not going to start doing calculation for you, it's a new ballgame again now that you change everything.
The reading above is normal other than the bias is set too low.
Last edited:
Betomon
Active Member
I read your instructions and I did follow post 180. I took measurements for r36 fully CCW and CW.. I guess you mean I never went back to double check all of them? 189 and 190 is what I didn't do fully..
I thought I found a bad transistor which is why I changed it.. As for the other one, like I said I just had it and wanted to use it. I saw no harm in that. I guess you did. I never refused to do anything you told me..
What calculations do I need to do? E=IR and their counterparts aren't getting me to this problem.. I'm not sure the direction the current is flowing to know where the extra 7mV are coming from.. I understand you lost all patience with me.. Sorry to bother you
Read from #180 to #190. I asked you to measure R34 at 1ohm AND 160ohm again to verify the few points. We were close to nailing the problem. M and N should never change no matter what you set R34. I asked you to double check and report back, but instead, you change everything. It's very hard to work with you like this. Comparing the good and bad channel at that point is irrelevant.
When you start blind changing, you might introduce new problem like you lost the voltage that threw me way off for a long time.
But now you change everything. The voltage change in your post #197 is normal, if you can just verify M and N don't change with adjusting R34, you might be good to go.
I won't be available for a while.
When you start blind changing, you might introduce new problem like you lost the voltage that threw me way off for a long time.
But now you change everything. The voltage change in your post #197 is normal, if you can just verify M and N don't change with adjusting R34, you might be good to go.
I won't be available for a while.
Last edited:
