Gotcha. I've got a plethora of 3k-4k resistors I can try. Based on the info (0.015v across the emitter resistors at idle), does this sound like the problem?
I won't know how this pans out until tomorrow afternoon.. I'll post back after I've got the resistor value narrowed down (if in fact this does reduce current draw).
I won't know how this pans out until tomorrow afternoon.. I'll post back after I've got the resistor value narrowed down (if in fact this does reduce current draw).
Yes, the bias current is too high. It would probably produce clean audio at this point but would very quickly overheat.
Replaced said transistor with a 3.6k. Didn't have any effect on either the current draw or voltage reading between the emitter resistors. I'm puzzled, I double checked the solder connections to make sure they're okay.
I took a couple more pictures of the area if this helps any.
http://www.wicks.com/Unrelated/DSCF0035.JPG
http://www.wicks.com/Unrelated/DSCF0037.JPG
One peculiarity I noticed... When I first power up the amp using the 10A output of my 12V power source, it starts near 7A current draw and slowly creeps upward of 12-13A before I notice the outputs getting too hot and need to cut power.
I took a couple more pictures of the area if this helps any.
http://www.wicks.com/Unrelated/DSCF0035.JPG
http://www.wicks.com/Unrelated/DSCF0037.JPG
One peculiarity I noticed... When I first power up the amp using the 10A output of my 12V power source, it starts near 7A current draw and slowly creeps upward of 12-13A before I notice the outputs getting too hot and need to cut power.
I threw a 2.7k resistor in parallel with the 3.6k, and that put the kibosh on all that current draw nonsense. It now idles happily at less than 2 amps, the amp meter is not sensitive enough to tell me exactly how much.
I figure the resistance value is now around 1.6k.. My next step is to clip the 3.6k out and see how it handles 2.7k.
I'm looking for the highest value that will maintain 0.000v on both ends of the emitter resistors when the amp is pushed near thermal shutdown, correct?
I figure the resistance value is now around 1.6k.. My next step is to clip the 3.6k out and see how it handles 2.7k.
I'm looking for the highest value that will maintain 0.000v on both ends of the emitter resistors when the amp is pushed near thermal shutdown, correct?
Yes. If the current remains at or near 0.000v at all operating temps, it's OK. Ideally, you'd find the value that would get just a VERY slight increase (0.002 max) at the thermal cutoff temperature but for a bass amp, it's not that critical.
When heating the amp up, remove the signal and check the idle current every couple of minutes. After it reaches thermal, remove the signal and let it cool (no fan, just let it cool down slowly). Check the idle current every few minutes to make sure that the idle current remains low throughout the cooling period. If it does, it's OK.
When heating the amp up, remove the signal and check the idle current every couple of minutes. After it reaches thermal, remove the signal and let it cool (no fan, just let it cool down slowly). Check the idle current every few minutes to make sure that the idle current remains low throughout the cooling period. If it does, it's OK.
With a 2.7k resistor installed the amp started out pulling ~2A, then within a minute or so was pulling 5-7A. Will try a lower value. Is this normal to require such a low value resistor?
I've never seen an amp require such a change in the value of the resistor but I've never used these transistors as substitutes.
Are you 100% sure that all transistors are tightly clamped to the heatsink?
Are any of the clamps bent?
Are you 100% sure that all transistors are tightly clamped to the heatsink?
Are any of the clamps bent?
I'm positive they're mounted securely. It's a fairly simple design, the transistor itself rests on the insulating pad, and a simple metal shim is mounted on top (the thickness prevents bending of any kind), straddling two transistors with a screw in the center to clamp the transistor to the heatsink (base) of the amp.
I'm running out of resistor options as my lowest one is 2.7k-- Doesn't prevent the amp from drawing excessive current at idle. The highest parallel combo I'm able to make is just over 2k, which works to maintain 0.000v across the emitter resistors at idle.
I just had another event unfold in the long process of this amp rebuild. One of the 2200uf rail capacitors has just erupted in a plume of smoke. The amp was idling for several minutes, no other bizarre behavior was noticed.
I'm running out of resistor options as my lowest one is 2.7k-- Doesn't prevent the amp from drawing excessive current at idle. The highest parallel combo I'm able to make is just over 2k, which works to maintain 0.000v across the emitter resistors at idle.
I just had another event unfold in the long process of this amp rebuild. One of the 2200uf rail capacitors has just erupted in a plume of smoke. The amp was idling for several minutes, no other bizarre behavior was noticed.
Sometimes the clamps get bent from over-tightening and they only apply pressure to one side of the transistor which can prevent it from mating to the heatsink.
Were any of the other caps hot when the vented cap failed?
Was the current draw high when the cap failed?
Were any of the other caps hot when the vented cap failed?
Was the current draw high when the cap failed?
There was no excessive current draw. The needle was well under 2A. Even with the failed cap venting the amp was still purring along with no protection.
The other caps in the vicinity were warm (likely from heat absorption), but nothing like the failed cap which was scorching hot.
What happens if the cap fails? Any damage to other components and should I try replacing it?
The other caps in the vicinity were warm (likely from heat absorption), but nothing like the failed cap which was scorching hot.
What happens if the cap fails? Any damage to other components and should I try replacing it?
Were you running it hard just prior to the failure of the cap?
Caps generally overheat and fail when excessive voltage is applied or when there is excessive ripple current.
Measure the DC voltage from the center leg of the output transistors to the negative speaker terminal. It should not be more than the rated voltage of the capacitors. Measure the voltage on both banks of outputs.
Caps generally overheat and fail when excessive voltage is applied or when there is excessive ripple current.
Measure the DC voltage from the center leg of the output transistors to the negative speaker terminal. It should not be more than the rated voltage of the capacitors. Measure the voltage on both banks of outputs.
The amp was just idling. I wanted to let it idle for a few minutes to make sure current draw would not creep up at idle. It was was probably idling for 5 minutes or so before the cap went.
As suggested, I measured voltage from the (-) speaker terminal to the center terminal of all the outputs. Measured from the NPN's, I recorded 78.2V. From the PNP's it was 79.2V. Measurements at each output transistor fell within a percent or two of these values. The rail caps are 2200uf rated for 63V.
Interesting.. What would cause the rail voltage to go so high?
As suggested, I measured voltage from the (-) speaker terminal to the center terminal of all the outputs. Measured from the NPN's, I recorded 78.2V. From the PNP's it was 79.2V. Measurements at each output transistor fell within a percent or two of these values. The rail caps are 2200uf rated for 63V.
Interesting.. What would cause the rail voltage to go so high?
I don't know if this amp has a regulated supply. If does have a regulated supply, the regulator circuit may be faulty.
What was the DC voltage across the B+ and ground terminals?
What was the DC voltage across the B+ and ground terminals?
Perry, I'm nearly certain this amp has an unregulated supply.
I didn't realize this until now, but my "12V" supply is putting out 17V at B+ and ground.
At another setting on this supply (this auto charger has several current and voltage settings), I was able to measure 11V at B+ and ground. With 11V input, I only measured around 50V between center terminal of all outputs and the (-) speaker terminal.
My mistake for assuming this power supply was really 12V. I really appreciate your patience in helping with this project!
Where do I go from here? I can replace the defective cap no problem.
I didn't realize this until now, but my "12V" supply is putting out 17V at B+ and ground.
At another setting on this supply (this auto charger has several current and voltage settings), I was able to measure 11V at B+ and ground. With 11V input, I only measured around 50V between center terminal of all outputs and the (-) speaker terminal.
My mistake for assuming this power supply was really 12V. I really appreciate your patience in helping with this project!
Where do I go from here? I can replace the defective cap no problem.
I think you're correct. I would be surprised to see a regulated power supply in this amp.
You can run the 12v supply/charger at the higher setting when running the amp hard. That will drag the voltage down to a safe level. But you should switch back the the lower output setting when the amp is drawing very little current.
You can run the 12v supply/charger at the higher setting when running the amp hard. That will drag the voltage down to a safe level. But you should switch back the the lower output setting when the amp is drawing very little current.
Replace the cap and test as you were going to do before the cap failed.
If the capacitor is in parallel with the adjacent capacitor, you may be able to continue testing by simply removing the defective cap. If they are not in parallel, you'll have to replace the cap before you do any more testing. If you continue with only one capacitor, monitor the remaining capacitor closely. Excess ripple could cause it to overheat and fail.
If you're running directly off of a battery charger (no battery, no external capacitor), you should also monitor the temperature of the capacitors near the B+ input.
If the capacitor is in parallel with the adjacent capacitor, you may be able to continue testing by simply removing the defective cap. If they are not in parallel, you'll have to replace the cap before you do any more testing. If you continue with only one capacitor, monitor the remaining capacitor closely. Excess ripple could cause it to overheat and fail.
If you're running directly off of a battery charger (no battery, no external capacitor), you should also monitor the temperature of the capacitors near the B+ input.
Would a large cap (100,000uf or so) in between the charger output and amp B+ & ground eliminate any ripple current to the caps nearest B+?
The rail caps are indeed wired parallel to the adjacent cap. I'll remove the defective one until I can replace it tomorrow.
The rail caps are indeed wired parallel to the adjacent cap. I'll remove the defective one until I can replace it tomorrow.
Great news! I have music coming out of speakers.
It's a crude test bench, but I have a cheap DVD player plugged into the RCA inputs, and a full range speaker hooked up to the audio outputs, run by the 12v battery charger.
It's drawing upwards of 5A currently, the 4" driver can only take so much. That being said everything works great except that the speaker output has significant distortion when the input voltage (B+ & ground) is less than ~11V.
The amp really hasn't heated up under this kind of load, but so far I've got 0.000v across all emitter resistors, using a 2.1k resistor for the bias circuit.
I'll do some better testing tomorrow when I can drag all this stuff outside.
I'll be ordering new emitter resistors to be certain they're matching before I put this amp to use. Is it necessary to replace all the base resistors (for outputs) with identicals? These were the ones that were 2.2ohm (IIRC), I only replaced a few that were bad. If there's anything else (other than capacitor and reverse-polarity diodes) you can think of that need replacing, let me know so I can order everything at once.
It's a crude test bench, but I have a cheap DVD player plugged into the RCA inputs, and a full range speaker hooked up to the audio outputs, run by the 12v battery charger.
It's drawing upwards of 5A currently, the 4" driver can only take so much. That being said everything works great except that the speaker output has significant distortion when the input voltage (B+ & ground) is less than ~11V.
The amp really hasn't heated up under this kind of load, but so far I've got 0.000v across all emitter resistors, using a 2.1k resistor for the bias circuit.
I'll do some better testing tomorrow when I can drag all this stuff outside.
I'll be ordering new emitter resistors to be certain they're matching before I put this amp to use. Is it necessary to replace all the base resistors (for outputs) with identicals? These were the ones that were 2.2ohm (IIRC), I only replaced a few that were bad. If there's anything else (other than capacitor and reverse-polarity diodes) you can think of that need replacing, let me know so I can order everything at once.
- Status
- Not open for further replies.