by bordonbert Wed Nov 07, 2018 9:52 pm
You have certainly done some really useful diagnostic work there and there is not much more I can suggest I am afraid. Forget the Standby switch, it isn't the problem here.
The fact that the valves can be switched and the problem remains with the same LED is not a good sign as you are aware. It really sounds to me as though you have a TSC/bias problem rather than just a valve one but it would pay you to be sure before taking it to be looked at. Can you not borrow a pair of known good EL84s to swap out to check that for definite? I wouldn't buy a pair just for this but borrowing them just for a test would really help you. Any friends have an EL84 amp you could raid or do you have a friendly local guitar shop who would help you out? This would be the logical next step to take.
Your slip with not connecting the speakers is not a good one to make. The GM36 would have detected it and quickly switched to Quiet mode with the internal power soak acting as the load but I don't believe your TM amp has this built in. Running an amp without any load on it is very stressful and can cause all sorts of problems even to the extent of burning out the output transformer. That is a distinct possibility unfortunately.
Other things which could be at fault here are in the TSC circuitry itself and that is processor controlled I am afraid. Though it is rare the control MOSFETs which sit in the cathode circuitry can be damaged and it may even be possible that the problem is being caused by a fault in the PI circuit which feeds the output valves. By the way, the Standby switch simply works by turning off these MOSFETs and shutting off current through the valves. That's a step up from the old method of removing the output valves' anode voltage which actually kills the valves by cathode poisoning rather than increases their lifespan as is widely touted. Old classic amp? Don't use Standby, just turn down volume!
So each EL84 has an independent bias setup with the TSC board controlling the current via that power MOSFET in each valves' cathode circuit. The PI stage is a single triode type so one output valve is fed from the anode and the other from the cathode. If one of the DC blocking caps feeding the output valve grids become leaky then they will allow the PI stage to influence the bias on that valve and it may go beyond the range which the TSC can adjust. These are labelled C18/C19 and are 22nF 400V. If it is in the TSC control circuitry it would almost certainly mean a PCB swap and H&K won't supply that to you direct, they only work through accepted maintenance sites. A decent guitar workshop could do that for you.
I know the idea of fixing it yourself seems a good one but if you haven't got much experience in working inside units like the TM18 it is a tall order to work at the level needed for such an intricate amp. Most guitar techs prefer to steer clear of them and curse and spit at the design approach. It's a state of the art piece of modern electronics not one of the old classic 'wet string and chewing gum' designs that they love because you can fix them with a hammer, a screwdriver and the internet for instructions. And of course, as always, remember that the amp works at voltages which will kill you. Really. I mean stone dead! Painfully!!! A mains shock at around 220V (only 110V for woosy foreigners
) is a nasty thing to experience. That's AC and, despite what Edison and his elephant said, AC is much safer. When you get an AC shock it kicks you away because the voltage is turning on and off and that allows you to release your grip. A DC shock spasms the muscles into permanent contraction and can make you grip tightly the thing which is killing you so you can't let go.