Tone tips for the GM36

Your analysis of the TM18 circuit is spot on.   So a low output HumBucker with no boost might not even hit this that level - pop on a boost pedal or Tube Screamer with just 3 dB of boost and the diodes will start to operate.  Higher output HBs or more boost from a pedal and you are turning on the diodes, not saturating the tubes - this is where my heartburn comes in.

So if I removed the bottom plate as you suggested, I'll see the copper side of the board, not the component?  Well that doesn't help...  I'd have to start disassembling the amp to get to one of those zeners, that could suck.
I've got two gigs in late June, then a few weeks off - I may pop the bottom plate off then just to see what I think...

Note the op-amp used has a temperature range larger than most consumer ICs, it's -40 to 85C.   So H&K are using chips that can take the higher temps generated by the internal power soak.  That's good to know but I still won't use it.  My guitar partner in the band comes from automotive electronics - they design stuff of operate in temperature extremes.  I come from Data Storage, where effort is taken to remove every degree of heat possible to increase MTBF.  Stated another way, the cooler you run your semiconductors, the longer time between failures.

OH CRAP, remove the diodes (just one diode would do it), and an OD pedal will drive the op-amp into saturation (15V rails).  If I recall, op-amps don't recover well from being saturated, possibly resulting in a non-useful tone.  Who the hector decided to put an op-amp on the front end of this thing?  Obviously it made the Boost easier to implement...

So, I guess this amp just doesn't take gain pedals well, unless you plan on using the built-in tube screamer zeners as part of your tone.  I think I'm a little disappointed...

One useful way to get yourself around this issue and keep all of the benefits of that setup is to swap out the existing zeners for new ones.  If you get 4x 4.7V zeners and put them in series reversed pairs, two one way and two the other, you effectively set the clipping levels to around 2x 4.5V + 2x 0.5V = 10V in each direction.  Or just 2x 10V back to back gives you a tad higher.  With a little hand selection of your zeners, they're not really terribly precise devices usually 10% tolerance, you could get this a touch higher though you probably wouldn't notice any real difference in practice.  With boost off and a gain of 3x that gives you input levels of about 3.3V before clipping.  That's a fair bit of headroom.  And you could do that in the existing PCB holes.  It then removes the clipping action as a tonal feature and leaves it as only a protection circuit to stop the Boost amp from clipping internally.  It gives you the maximum undistorted input with the existing setup.

Another way does seem to be easier.  If you can trace out the circuitry around the Boost you could also change out a single resistor and reset the overall gain of the stage.  Remember the TM18 shows a gain of 3x with boost off and about 7x with it on.  That ratio 3:7 is a pretty good balance and with the 10k feedback resistor swapped out for another value you could drop the non-clipping gain of the entire Boost circuit for both conditions keeping their comparative levels the same.  A 3.3k will give gains of 1x and 2.3x and you could easily work out the value for any gain of your choice.  That does mean less drive through to the following stages which may alter their overdriven characteristics, but as so many of us seem to want to drop our gain with valve swaps it may even help that aspect too.  It also keeps the non-symmetrical aspect of the clipping which I like the idea of.  Not straight into the harsh sounding but gradually taking off one side first.

Zener swapping or reducing the 10K will only buy a few dB of headroom, so to paraphrase BB, in practical use not much help.

Back at the 10,000 feet level what this means is - the only thing using overdrive pedals will get you is activating the GM/TM's internal tube screamer circuit, not tube overdrive.  This is where my disappointment is...

So with 128 patch locations, save a patch with the tube saturation you want (the gm channel Gain controls give you plenty) and only use input side overdrive pedals when you want to add a tube screamer sound.

On a side note, a similar op-amp with 15V rails drives the effects loop, so watch your levels.

So if you are using a Strat with vintage wound single coils and then whip out your SG with a Pearly Gates and a JB in the bridge (or your Kramer with active EMGs), you got auto-tube-screamer with the higher output guitar.

So I put Seymour's Alnico II Pros into my SG, to more closely match my P90s in my LP and the output of my Strat, so I changed my #1 to fit my (supposedly) very flexible amp...

And I'll probably end up going with a million almost redundant patches on my HD500X and GM, one set of patches for each guitar.  That's why they give us all those locations...

On another side note, I'm on my way to GC to buy a 7 band EQ to put in the effects loop and dump 7KHz and above on my higher gain patches - as BB mentioned above.

BB, If you ever make to this side of the pond (specifically Denver), I'll buy you a coffee or Newcastle or ...

I know this is an old thread but just getting back into my GM36 settings for a live gig and pleasantly discovered a different side of the amp yesterday.  I was getting my levels set up for an outdoor gig so I moved my rig outside.  I was able to turn the master to between 12-3 and most of my presets set power on 18W or 5W.  Clean presets are at 36W.  Wow...the amp had a different character to my ears.  I keep my EQ close to noon or with slight adjustment and use a 10band EQ with a mid emphasis.  Will see if I can get some video/audio at the show.

It's quite surprising how just a slight change in circumstances can make a big difference in tone. Sound clips would be great. See if you can get some Apenney.