I noticed that from one of the LT494 there was a steady freq of 34k on 2 pins while on the other lt494 there was nothing steady. Is it that one of them is mulfunctioning?
Parasitic oscillations might account for the 34kHz.
Unless that is supposed to be the operating frequency of the switching IC. Then the problem might be the other LT494 which lacks the 34kHz. (I have no direct experience with this IC.)
Parasitic oscillations tend to result in components heating up. Suppose a safeguard in your circuit kicks in, and turns something off for a split second? That would create the glitch. This would happen regularly, same as your glitch happening regularly.
A strategically placed capacitor is often a solution to stop parasitic oscillations. This means a lot of experimenting. It carries risks of making something worse.
Also, with regards to the capacitors, i think you are refering to those at the high voltage side.. (coin shaped). Shal i take them out and test them?
i think that a capacitor wizard is very usefull for these type of repairs. what do you think. I read that you can test them in circuit.
Capacitors are frequent weak points. An ESR meter is handy to test them in circuit.
It's a question whether a deteriorating cap is likely to allow normal running except for a glitch every several seconds?
Also if the arcing is in the transformers there is not much i can do i think.
Maybe not. It is not likely to happen anyway. If arcing occurs at all, it would tend to be constant.
Nevertheless just in case, turn off all lights and look for any arcs in the dark.
------------------------------------
Your photographs show a lot of components affected by the burned mosfets. If possible you should test them. This won't be easy since they're surface mount.