Gallien Krueger 250ML (Unit 2)

[g1]

Just to sidetrack a little bit, it's been a long time since I worked on one of these, so I gave myself a refresher course on the grounding nightmare, and why you can't rely on the rear panel as a ground.

This is the input jack. It is the only connection from circuit ground to "chassis", actually the front panel. If the plastic thread is stripped, you do not get a solid connection to the metal front panel and lose chassis ground..


Here is a side view of the chassis that someone posted on the web. You can see that the front panel and rear panel are separate pieces. I believe there is foam along the bottom of the front panel but I may be mistaken on that.
The top cover connects the front to the rear panel, and if the screws get loose, or if you have the top cover removed, you have grounding problems. So you can not rely on the rear panel to give you a reliable ground, especially when testing.

[sweatyk]

Just to sidetrack a little bit, it's been a long time since I worked on one of these, so I gave myself a refresher course on the grounding nightmare, and why you can't rely on the rear panel as a ground.

This is the input jack. It is the only connection from circuit ground to "chassis", actually the front panel. If the plastic thread is stripped, you do not get a solid connection to the metal front panel and lose chassis ground..


Here is a side view of the chassis that someone posted on the web. You can see that the front panel and rear panel are separate pieces. I believe there is foam along the bottom of the front panel but I may be mistaken on that.
The top cover connects the front to the rear panel, and if the screws get loose, or if you have the top cover removed, you have grounding problems. So you can not rely on the rear panel to give you a reliable ground, especially when testing.

Congrats g1!!!!

You sir are correct. The back panel was not making the proper ground connection being it is separated from the speaker cab. The front panel was making a good connection with the ground on the input jack. Having said that here are the voltage reading with the proper ground. The + side is reading 47 VDC, the - side is reading -34 VDC but that leg is lifted out of the circuit. I will reread it after reconnecting it and see if that changes.

Again thanks g1!

[sweatyk]

Both +/- supply rails are 47 vdc. That's good news for that. I still have a 20 degree difference in temperature between components in either side of the amp. Later today I will note exactly which ones in hopes of finding something out of specs causing this.

[sweatyk]

Now that I have established the 45 vdc +/- rails there is still a heat issue to deal with. When I let it sit idle with no speaker load or input signal a few transistors reach temps higher than the others. One is in the power section and the other is on one of the channel outs. The ones that I have highlighted in yellow will reach 100 degree F while the rest mounted on the back plate marked with blue will operate around 75 to 78 degree F. I'm using one of those laser infrared thermometers to measure the temps. May not be accurate but it can show the differences. The output and driver transistors have been replaced on both sides of the amps....they were shorted. None have been replaced in the power section. So Q31 along with R157 on the amp side and Q35 of the power section is getting hot. Where do I go from here? recommendations?

[Enzo]

Set the bias. Especially after changing the output devices.

[g1]

Bias spec is about 5mV DC across each of the .33R 5W emitter resistors.

[sweatyk]

Set the bias. Especially after changing the output devices.

I had set the bias after installing all of the new parts. As a precaution I did it just now again. I measured one side and then the other. They were pretty much right on the mark. One side - R157 = 5 mv, R139 4.1 mv. The other side - R123 = 5 mv, R106 = 4.5 mv which was where things were when I started this thread. The heat issues still exist on the listed components a few comments from this one. Ids there somewhere else I should be looking at? I have stared at the schematic till my eyes cross. Any suggestions?

[Enzo]

All problems are not necessarily on a schematic. A bead of solder between two unrelates copper traces can cause havoc, but you won't get a clue on the schematic. Transistors not making good thermal contact with a heatsink can cause overheating - won't be on a schematic. Etc. SO don't limit your thinking.

One output transistor is running hot. OK, could be something about that part, or it could be the circuit around it. Was that part changed? Or were others changed but NOT that one? And either way, any chance it needs fresh heat grease or a tighter bolt? Maybe Q31 is just being turned on harder than the others. SO check action at Q30 and R145.

And the one voltage regulator xstr runs hotter? Same deal, heatsink attachment, OK? Is it new or old, and if new is it the identical type?

[The Dude]

Agree with Enzo. Another possibility is a counterfeit parts. Did you get your replacement transistors from a reliable source?

[sweatyk]

Agree with Enzo. Another possibility is a counterfeit parts. Did you get your replacement transistors from a reliable source?

All parts came from Mouser. If not there I would use Digikey or Newark. I never trust parts from eBay stores.

[sweatyk]

All problems are not necessarily on a schematic. A bead of solder between two unrelates copper traces can cause havoc, but you won't get a clue on the schematic. Transistors not making good thermal contact with a heatsink can cause overheating - won't be on a schematic. Etc. SO don't limit your thinking.

One output transistor is running hot. OK, could be something about that part, or it could be the circuit around it. Was that part changed? Or were others changed but NOT that one? And either way, any chance it needs fresh heat grease or a tighter bolt? Maybe Q31 is just being turned on harder than the others. SO check action at Q30 and R145...........................By "check action" you mean?

And the one voltage regulator xstr runs hotter? Same deal, heatsink attachment, OK? Is it new or old, and if new is it the identical type?

^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
Those transistors are the originals. Properly mounted. I have not replaced anything in the power section

Also:
All of the output transistors were changed along with their associated driver transistors.
Q25 was literally cracked in half....it was replaced
Q24 was replaced and I can't quite remember why but it is new also
R98, R131, R145 were burnt to a crisp
R112, R163A had obviously burnt and was replaced prior to me owning it.
R157 was open, I replaced it along with R139, R123, R106

SO check action at Q30 and R145...........................By "check action" you mean?

[sweatyk]

Something that I just noticed. R159 is supposed to be 2.2K and R160 is supposed to be 3.3K per the schematic. However, there are 1K resistors in both positions. The same with R102, 103, 125, 126, 135, 136

I guess there was a value change that the engineers didn't edit on the schematic or parts list. At first I thought I had an AHA moment. But the other amp I have has the same resistors in it

[g1]

Evidently there was another revision either before or after that had different values for some of the components. Your units do not match the available schematics.
The 56K shown in your pic is 22K on the schematic, so that's another difference.

Fortunately, you have 2 units that are the same revision. Have you verified that every part you replaced matches your other unit rather than the schematic value?
For example, a burnt resistor replaced with a value from the schematic may be wrong value and could cause overheating.

Also, check your IR thermometer readings against the good working unit. The hotter transistor in the power supply area could be normal, or could be another problem.

[Enzo]

Companies rarely back edit schematics, because that would confuse people with then-stock amps. They issue revised schematics. With luck, a serial number note. I have a boatload or engineering change orders for the ML series

Please do not try to change a bunch of stuff to match the schematic. I have a few schematics where those resistors are 1 and 470 ohm. Look in the circuit, they are just the current limiting circuits. Assume the installed values are correct.

[sweatyk]

Evidently there was another revision either before or after that had different values for some of the components. Your units do not match the available schematics.
The 56K shown in your pic is 22K on the schematic, so that's another difference.

All parts were matched based off of what was physically in the amp not the schematic. I've learned over the years that the schematics aren't always accurate.