GM36 issue (seemingly) in preamp section. Please help

Hey guys, Sorry to say this is my first post here but after reading many threads hoping to find some help, I didn't seem to find any advice on the same issue that I seem to be having. I appreciate any help you can provide and thank you in advance.

I'm a long time H&K fan and have owned several triamps over the years. I recently picked up the GM36 used and it is one hell of an amp. Happy to say it does everything I need and sounds exactly how I want. But who wouldn't say that, right? So not too long after getting the amp this issue started poking it's head out. The problem displays itself through a very weak tone and very low gain. It affects the crunch, lead and ultra channels the most and I almost want to say isn't affecting the clean channel at all? At first it was intermittent and seemed to be when I started the amp cold. I could fix it for a short time by tapping the V1 tube (could have just been lucky, not convinced the problem is isolated there) or by banging the amp. It was usually fixable by some means pretty easily. Now the problem seems to be stuck and I can't get the amp to sound normal at all. I've spoken to the original owner and unfortunately, confirmed that the warranty has expired on the head. He did offer to split the cost of repairs with me which not only speaks for the character of the guy, but also lays to rest any suspicion that he may have knowingly sold me an amp with some gremlins.

Like I said, I did do a little research and have tried the following things already with no luck:

Swapped all the tubes.
Jumped the effects loop
Tested the preamp out through the power am section of another head
Obviously checked and tested all the cables

So this is why I'm fairly certain it's definitely an issue in the preamp section because I'm getting the same tones regardless. Thanks for any help that you guys can provide. I'm about to blast all the inputs etc. with some canned air just in case something is just dirty. I'd love to get that lucky.

What is the layout of the preamp section? I'm almost certain the clean channel is unaffected so I'm guessing that would narrow the probably down to either V1 or V2 depending on which gain stage the clean channel does not use?

Hi there John, welcome to the forum.  I have to hand it to you, you've gone about this the right way, those would have been the exact things I would have recommended to you to try.  If tapping the valves corrects things it pretty much shows a problem around that particular valve.  I would strongly suspect the valve in the V1 slot, (that's the one covered with the metal shroud), has a problem in itself or immediately around it.  You say you swapped the valves and the problem persisted.  Do you mean you swapped the valves around in the amp or you replaced them with a known good set?

The preamp is laid out so that V1a/V2a/V2b are in circuit for all channels.  V3 is the Phase Inverter of course so is part of the power amp anyway.  V1b is added as an additional gain stage for Crunch/Lead/Ultra.  If the Clean channel is unaffected then V1b seems the logical culprit.  The first thing to do is substitute that valve for a known good one.  With the known good valve in there try tapping it or wobbling it gently by the tip.  Does it come back to life then?  Are you running anything "exotic" as your valve choice?  Not some wheezing clapped out old NOS thing?  Or worse, one of the infamous "not really substitutes", the 12AU/Y/V/Z7?  Or are you still using the original Chinese valves supplied when new?

It is possible that you have a problem with the soldering of the valve base to the PCB but H&K's build quality is good in those areas so it is rare.  It is also possible that you could have a problem with the relay which switches that valve in and out but I would really doubt that one.

How far can you go SAFELY yourself in working on this problem?  Remember valve voltages REALLY BITE!  Sometimes to death.

Thanks for the reply, BB. After the research I did initially, I understandably hoped and expected you'd comment soon. I had to respect the time difference between continents but thank you for the prompt response!

First I'll answer your questions and then I'll get to the confusing stuff. I did swap new and known good tubes in every position more than once. This did nothing. I did open the chassis and inspect for any obvious issue or bad soldering around the tube sockets on the pcb. I also removed the sockets screws from the back panel allowing me to life the smaller pcb and inspect a little further. I did all of this without dying so that was the only silver lining because even after a thorough air blasting of all components I had the same issue upon firing it back up.

Here's the confusing stuff. AS OF RIGHT NOW it's working as normal. And here's my only relative theory that could turn out to be nothing more than coincidence and I'm not sure there is any logic to back this up. I have been on the hunt for a footswitch (or two) to use in the control inputs of the fsm-432 to switch off reverb, etc as it's designed to accommodate. I tried a few footswitches hoping to get lucky but they didn't work correctly due to having stereo 1/4" jackets rather than mono. Today the two single button footswitches (mono jacks) I ordered on amazon came and even though the amp wasn't sounding right, I was anxious to see if they worked so I could stop the hunt. I plugged one in to control jack 1 which I had already previously programmed to switch on/off the fx loop during the failed trials. To my chagrin, even this footswitch did not work switching on the loop. Just to make sure, I programmed the midi pedal one more time to control the loop (55) just in case. Still nothing via the new footswitch which should work? However I could get the loop to switch on and off momentarily just plugging/unplugging the 1/4" cable in the control 1 jack. So my ONLY theory here is that somehow during my footswitch trials something got switched wrong or halfway internally causing this preamp output issue. Of course I will have to give it a few days to prove that this is not simply a coincidence but, my fingers remain crossed.

I will keep this thread updated.

If you have no patches which you can't do without try a factory reset of the amp.  If you need to store your patches you could try setting up control from Fredo's tablet/laptop app and pull them off the amp and into storage to upload again later.  It should be impossible to get switching items in the amp circuitry out of phase.  What I mean by that is that for each channel say, there are a number of switched changes made at different points in the circuit, each one controlled by its own switching circuit element.  Some are via miniature relays and some by FETs.  Each of these switching blocks has its own overall control circuit which coordinates all of its elements.  Even if two elements did get out of phase with each other they should all drop back into line once the channel is deselected then re-selected again.  The circuitry can't set up a switch to go out of phase and then keep on switching out of phase.  It's kind of like having a straight line of toggle switches which you push each way at the same time with a piece of wood against all of their toggles.  If you knock one switch out of phase with the others by accident you only have to push all of the other switches over to match it then all of them back again and they are in phase once more.

The only one I can see being immediately relevant to this situation is RL1 which connects the input of V2a to either the output of V1a or V1b to go from Clean to Crunch/Led/Ultra.  This adds in the fourth triode stage for the higher gain channels.  If that relay has failed or stuck then you could be left with a selection for Clean at that point when everything else is being set up for one of the other channels.  In its off position it is selecting Clean so if the coil has burned out it would default to that option all of the time.  I have to say it is very unlikely but it would be logical, (pointed ears on at that point).

What type of footswitches did you order from Amazon?  They should be simple types with any LED in series with the switch not in parallel.  If yours are in parallel, (across the two switch terminals), that could lock the Fx Loop as the FSM432 could see the LED as a closed switch.  Wiring in the switch should be:  [Tip -> LEDanode : LEDcathode -> Switch : Switch -> Sleeve] and not [Tip -> LEDanode and Switch : LEDcathode and Switch -> Sleeve].  I have seen switches made to work in both ways and they aren't compatible.