Tektronix DPO3000 Power Supply Repair Help

[analogRF]

I have this DPO3054 which just stopped working the other day and does not turn on at all. Power button has no effect whatsoever, went completely quiet.
There is no output on the PSU connector. No voltages anywhere. The primary side is 160.00V (yes, really to two digits )
but nothing else. There is a blown resistor (open) on the back of the PSU board which a little search told me it is 10 ohm 1206 resistor
so I replaced it and as soon as I plugged in the scope to my isolation transformer. that resistor got blown violently again
there are no voltages going on in the circuit, there are two main converter chips: one is a TOP244GN and the other is a LM5033SD
but which one does what? I dont have a clue, right now they are not receiving any voltage...

anybody has any experience with these power supplies? They are so hard to work on....the heatsinks soldered to the board and bent over the hole
board so there is no access to anything...I use long needle shaped multimeter probes. Scope probe, forget it 

if you have any experience with these, please share...thanks

[Jetecnet]

We had a few of these to repair, the first had the 10 ohm resistor burn and the Top244 destroyed. There's a bank of semis on the heatsink, the one to the right side (dead) was a FQPF9N50C that had been replaced but subsequent boards had a P12NM50, probably a reasonable sub. Replacing the resistor and Top244 enabled the main board to show a green power led but no action on operate. We pretty much shotgunned the thing with no success. The last 2 had working TOP244 but again no operation, green led on main board, one was just tripping. We came to the conclusion that the 5033 chip was dead and that was beyond our ability to replace. There is also a PFC circuit in there which was suspect. We stopped doing repair on these as the power supply sucks, time constraints, cost etc. I hope you have more luck.

[analogRF]

yes the damn thing is impossible to work on. replacing chips blindly and at random seems the only way    something that I have never done before...even doing that is really difficult on this thing. This combination of SMD and TH is horrible.....there is this huge ceramic cap tightly next to the LM5033 (plus other stuff...) and not to mention the badly designed heatsink cage....it's a nightmare....

so far my first suspicion is the LM5033. but I cannot even measure any input voltage on it, so what gives?
still trying to find out where the 160.00V goes...

[analogRF]

I cannot find any obvious shorts...none of the MOSFETs are short but they could be open? right? can that happen?
I have checked most of the caps including ceramics just for short circuit and there is none...

I think somehow that 160V suddenly drops on the 10 ohm resistor because something is not working but not necessarily a short circuit....
the blown 10 ohm seems to me like a current sense...it is connected to the low side of a transformer (among other things) but the other side
does not go to ground...it disappears among bunch of other stuff that is hard to even see...

[Jetecnet]

I'm working from memory here so check it. The TOP244 is supplied via the 10 ohm resistor and drives the primary side of the small transformer. It should be switching at 130KHz and you should have output on the secondary side for the low voltage supply to the circuits nearby. If its working you should have a led lit on the main board. Pressing the standby switch to on, via a convoluted pathway of logic seems to drive the 2 optocouplers near the transformer to activate the 5033 to start the main switcher outputs. TOP 244 in my experience do fail quite often without any damage or they blow their lids off which is more obvious.
To work on this board I just carefully bent all the heatsink fingers up, they survived being bent back later 
The 5033 apart from being tiny is surrounded by stuff and has a pad underneath that needs to be soldered. I just wasn't going to try replacing it.

[analogRF]

yes your description more or less matches what I have been able to extract so far.
i just took that heatsink with the components out to also check the mosfets with proper tester and they are all ok
I just placed order for TOP244 and LM5033D. I think I will replace both before putting that heatsink back in.
I have to take out a couple of big ceramic caps and perhaps even the transformer before attempting to replace the LM5033
but I think I'll do it. Basically you need to disassemble this crap to be able to do anything...still there are tiny components
very very close to that chip but I'll try my best...

another option is to just replace the TOP244 and see if it works. if it doesnt then I will have to disassemble the whole shit again
i dont know...still undecided...looks very much the TOP244 might be the only thing that is dead because the supply is just completely dead silent

can it be something is defective on the main board? or one of those opamps/comparators on the PSU that I think sense some signals from the main board? i dont know what exactly they do..

[Jetecnet]

We had some working scopes at the time and swapped power supplies around, the faults seemed to be only with the power supply, not the main board. I would just replace the Top244 to start with you might get lucky. Its a lot easier than the 5033.

[analogRF]

ok, that's a relief!

yes, I am planning to only change the TOP244 for now and put it back together without actually soldering the heatsink legs
and see what happens.
I just took out the TOP244 and it is clearly dead (31 ohm between S and D which directly puts 160V on the 10 ohm resistor through the transformer winding) Thank God the transformer seems ok
I have  now checked every other component in that area (except LM5033) and everything checks out. I think the TOP244 should resolve the problem

I am waiting for it to arrive, perhaps tomorrow or on Wednesday. Digikey didnt have it so I ordered from Mouser and in Canada they usually dont get to me in one day as they claim . Sometimes they do but not always...Digikey is always spot on.

[analogRF]

I misspoke  , Digikey do have TOP244 in the right package but they don't have the LM5033D in the right package.

Since I wanted to order both, I went for Mouser.

[analogRF]

Replaced the TOP244GN chip and put everything back together and the scope works like a charm again 

here are some information for those who might need it in the future:
as soon as you plug in the scope, before turning it on, a Green LED lights up near the CMOS battery (above the ADCs)
on J0401 connector (7 pins coming from the PSU) we have these voltages:
pin 3 is 3.5000V and pin 4 is 5.18V.
after powering the scope on, the 3.5000V disappears but pin 6 now has 3.22V. Pin 4 still has 5.17V
also after power on 2 more green LEDs light up on the motherboard

the main voltages on the big PSU connector are 6.03V (red) and 12.6V (orange)

on the PSU board itself, the TIP244 chip is ON right after plugging the scope into outlet. It generates 3 voltages.
1) 13.16V on C90 which supplies the LM5033 (the main DC-DC converter) and other stuff on the primary side of the PSU (the PFC circuit, etc...)
but the LM5033 still remains shut down through an optocoupler which senses something from the main board (the power button)
The GND for this 13V is on the primary side and is isolated from chassis
2) 16.75V on C85. it supplies some opamps and LM339 comparators on the secondary side. This supply is chassis referenced
3) 9.2V on the capacitor next to C85. This feeds an LM317 (SOIC package) that generates 5.18V

when the scope is powered on, the LM5033 comes on and generates the 6V and 12.6V.

[analogRF]

more pictures

[analogRF]

and 2 more...

[analogRF]

I also replaced 3 electrolytics including the input filter (100uF, 450V) but all of them were perfectly fine and absolutely no need to change.
I only did than because I had taken them out already and had replacement on hand.
The original caps on this thing are Rubycon and United Chemi-con and all are 105C. Pretty high quality stuff.

[Jetecnet]

Nice job, glad you got it to work. We were not so lucky, a lot of stuff comes via other repair shops and is messed up before we get it.