A tale of two Gallien Krueger 400RB heads

[xtian]

It's always nice when you have to units to compare, though I'm still too new to solid state amps to know what I'm seeing.

Unit A passes signal, but has a moderately bad hum even with MV down. +/-60v rails measure OK, and show ~75mVac ripple.

Unit B also passes signal, but has moderate white noise even with MV down. Rails measure OK, and show 50mV and 80mV ripple. So I guess the ripple on unit A is not causing the hum...maybe.

According to the service manual, one should set the bias to 33mVdc on R26 and R27. Here's where things get weird. Unit A (the one with the hum) has bias set to 10mv and 5.2mV. Bias pot works OK, but 2:1 bias relationship remains.

Unit B has bias set to 15mV and ZERO, yet still sounds good passing signal. How? If I turn up the bias so the zero side starts to show voltage, the other side's bias voltage shoots up, to 40 or 50mV.

Does this info provide any clues about the hum in unit A, or the wacky bias in unit B?

[J M Fahey]

1) I guess that you are setting bias with speakers connected.
Try again without them-

2) even if you have a difference transistor to transistor ,this shows simply that they are not matched, no big deal.

If both sound good, there is no problem to solve; as of slight hiss or hum, it might be heard in a silent rom, but being 200W PA power amps, doubt that is a problem in any conceivable live sound situation.

[xtian]

Thanks, JM. Bias readings don't vary when load disconnected. Hum on unit A is bad enough to constitute an issue that needs to be solved.

[g1]

According to the service manual, one should set the bias to 33mVdc on R26 and R27.

I don't know where you get the 33mV from, my understanding is there should be 10mV between the 2 emitters (sum of R26 + R27).
GK sometimes did it that way.
(idle current would be approx. 15mA)

[J M Fahey]

Ok, I donīt think that there is something *broken* and GK certainly wouldnīt have left a defective unit leave the Factory (lots of assumptions here, but we have what we have) so I *guess* somebody messed with it once, and didnīt reassemble it the exact same way, or something like that.

To start with something I guess the rectangle containing " +15b -15c In " is a connector.
Try to unplug it to see what happens.

As I understand it, R1 - R4 - R8 will still be directly grounded, there shoud be a track joining them to main filter caps centerpoint, no matter how tortuous it is.
If so, check wheter hum lowers or disappears when disconnecting that connector.

[xtian]

I don't know where you get the 33mV from, my understanding is there should be 10mV between the 2 emitters (sum of R26 + R27).
GK sometimes did it that way.

I made a mistake in my OP. 10mV is the target bias reading, and I was measuring across each resistor, instead of from emitter to emitter. Thank you.

[xtian]

Ok, I donīt think that there is something *broken* and GK certainly wouldnīt have left a defective unit leave the Factory (lots of assumptions here, but we have what we have) so I *guess* somebody messed with it once, and didnīt reassemble it the exact same way, or something like that.

Unit does not show any signs of previous repair or mods, but I take your point.

To start with something I guess the rectangle containing " +15b -15c In " is a connector.
Try to unplug it to see what happens.

As I understand it, R1 - R4 - R8 will still be directly grounded, there shoud be a track joining them to main filter caps centerpoint, no matter how tortuous it is.
If so, check wheter hum lowers or disappears when disconnecting that connector.

Yes, R1,4,8 are directly grounded to caps and to chassis. When I unplug the jumper, hum vanishes (still some hiss coming from speaker, so output section is still working). That's a great way to divide the circuit, JM.

One weird thing--the power transformer buzzes like a fish tank pump sometimes. Worse at power up, then settles down. No loose connections; prodding and pushing doesn't seem to affect the buzz.

Anyway, maybe this is a grounding issue. The preamp board is grounded to the front chassis subassy, and the power section is grounded to the rear chassis subassy, and there is a 4ohm resistance between the two. I'll try to improve this.

[xtian]

Well, I tightened up all the (loose!) screws and nuts on the chassis, and that reduced the resistance of the ground plane to >1ohm. Didn't improve the hum much, though.

[g1]

This sounds like it may be a real old unit that does not have the ground wire connecting the front and rear panels?
Try adding one. They did that on the later units as relying on the screws and top & bottom panels for grounding was not reliable.
Make sure all jacks are tight too, maybe remove them and check for corrosion or oxidization where the star washers meet chassis.

[xtian]

I thank you. As I mentioned above, after tightening all fasteners, the front-to-back chassis resistance dropped to under one ohm. Didn't improve the hum, though.

I agree that in a band setting, user will not notice the hum. When I'm testing, I have a speaker right at head level, pointing at my face! I'll ask the user to evaluate it and decide if he wants to do anything. GK offers a replacement toroidal transformer for $118.

Thanks, guys!

[g1]

Those screws may loosen off again and the front to rear panel grounding will be compromised. We were told to always install the wire as a warranty item. Even if you just connect the wire to the ground pin of the input and speaker jacks.