HP54520A SMPSU problem

[sv3ora]

Hi,
after a few seconds or tens of seconds the scope blows the mains fuse (inside the RFI can filter).
I have located the problem in it's switching mode PSU module, because when I disconnect it from the rest of the scope, it still blows the fuse.
Here is a schematic of the very similar model which I have checked it to be very similar inside to mine.

This PSU has a 5A fuse as shown in the schematic, but externally to it, there is another RFI filter can which plugs directly to the mains socket. It is the fuse inside this RFI filter that is blown (4A), not the internal fuse on the PSU module shown in the schematic. I have tested this external RFI filter without any load (disconnected from the PSU module) and it does not blown it's fuse. I have also connected 220v directly to the PSU, bypassing the RFI filter and then the internal PSU fuse is blown, so I do not think it is this RFI filter's problem.
Then I have tried to locate the problem in the PSU module by removing components. The mains side of the PSU (page5) is almost identical to mine. The only differences are that the input mains capacitor in mine is a varistor (275v/20A), and instead of the varistor shown in the schematic, mine has a thermistor. Also R13 and R14 are not present in mine. All the other things are the same.

Now, I have tried to isolate the HV mains side by disconnecting R12 and q3 completely. The external mains fuse is still blown. So I thought it would be this part of the PSU that has the problem. I changed C22 and C23, fuse still blown. I changed C1 and C51 near the RFI filter, fuse still blown. I changed CR1 with an identical one, fuse still blown. I temporarily removed C3, fuse still blown. I have replaced the thermistor with an identical type, fuse still blown. I have completely removed the varistor, fuse still blown.
Sometimes the thermistor cracks and blows away completely!
I have also bypassed the internal RFI transformer and fuse is still blown.
Note the fuse shown in the schematic is not blown, it is the fuse inside the external RFI filter can (not shown) that blows away, and the thermistor of the PSU module cracks. Probably because the external fuse is 1A less than the internal.

I am starting to get pissed of with this thing, I have almost changed every component in the HV side apart from the spark tubes and still nothing.
The scope powers up on but after a few seconds/tens of seconds, blows the fuse (and the thermistor sometimes).
Maybe an ingenious man here can give me some useful advice of how to save this scope!

[shakalnokturn]

Assuming you have done the other tests with Q3 removed that only leaves you with C4, R1, R2, VR1, VR2.
I wouldn't suspect R1 or R2, could be the others. If PCB is double-sided I would keep an eye open for components mounted too close to the copper (here you could suspect R1, R2) but that should leave some marks, it does sound like there is some current involved.

Consider letting it sit powered, Q3 removed, with a 100W bulb in series, that will allow you to spare fuses and PTC's, it may even give you some time to make measurements.

[sv3ora]

Assuming you have done the other tests with Q3 removed that only leaves you with C4, R1, R2, VR1, VR2.

I know but I thought these are reliable components. C4 is a ceramic disc, the resistors see no visual damage and VR1, VR2 shouldn't be reliable? These are neon trigger tubes, can they fail?

[shakalnokturn]

I don't have any experience with gas (not sure it's neon) TVS's going defective and triggering too low, have not encountered that many though.
If your short circuit culprit is not leaving any burn marks it must be contained in a component, the gas tubes are designed to do that.
Test without or throw a couple of appropriate MOV's in place.

[sv3ora]

I tested it without these gas tubes. The fuse is not burnt! It must be these tubes then. I search on the net and they have a finite life.

[shakalnokturn]

I search on the net and they have a finite life.

Zener diodes owe them a lot...