> How do I trouble-shoot it?
Similar to a tube amp.
Observe which knobs affect hiss. Treble and Reverb are particular clues. If a knob has no effect on hiss, the problem is likely *after* that knob.
Take it out in the sunshine and look for trouble. Burned resistors. Bad solder joints.
The main burned-resistor item here is the 330 ohm 2W R48 R49 jobs in the power supply; fairly notorious on other Fender sand-amps.
Remember that soldering is a minimum-wage occupation; solder inspector gets a dime bonus. Assume there are solder joints that just-barely worked only for so long.
Is the reverb pan working, is the reverb pan hooked-up? If it's disconnected, bad stuff may happen.
Check voltages. Your + and - 12V rails should be 10V-14V each, not zero, not 20V. All opamp pins (except U2B 567 which follow footswitch state) (except power pins which must be near power voltages) should be near-zero.(DC biasing of bipolar opamps is much simpler than tubes with their +2V and +200V cheack-points.)
> silent in the #2 and #4 position
I don't know the amp. What is this "#2 and #4 position"?
> I think it is a front end transisitor
You say *both* channels; but there are separate opamps for each. While it is possible one contaminates the other, that would not be my first guess.
However there is a major oversight in this amp, and if it has been abused it *might* be part of the problem. U2A and U1A connect *direct* to the input jack. If more than 12V input is applied (say, a loudspeaker OUTput), the chips may breakdown. There's some internal protection, but not much. A partial breakdown could hiss madly. Is there *any* way to insert a 33K resistor between jack and opamps? Maybe cut a PCB trace, scrap the lacquer off the trace, and bridge the gap with a 33K resistor? That makes it much more robust against stupidity.