Component idenification

That's an NTC inrush limiter. It has roughly constant power dissipation, when in use and is expected to run warm.

What is the actual current draw? Is your bulb dim?

You should be able to identify parts from board designators and the reference manual.

RL

That's what I thought too. But there's no such component visible on the schematic:

Cursitor Doom wrote: ________________________

Resistor Thermal 7.5 ohms.

..... Phil

Then it's time to find the short cct.

RL

Sure is. I came across a section of the service manual I hadn't found before (I find service manuals in PDF format *so* difficult to navigate) and it's provided a list of parts to check in the event of this fault arising. The list is quite lenghty, but is at least limited by virtue of there being a multi-pronged jumper which interconnects certain sections and by removing that you can drastically reduce the number of parts to take a closer look at if the fault then clears, which it has. Fingers crossed, eh?

The list is quite lenghty, but is at least limited

OK - if the fault clears when a jumper is removed, then you know that the fault is in the removed section, correct? Not the other way around?

Too many individuals think it works the other way, sadly.

Peter Wieck Melrose Park, PA

Yes, you're effectively removing the connection to the damaged section; the section which has some component in it which has gone low-resistance.

This kind of fault would be so much easier to detect if we had IR glasses to view the suspect areas with. Someone really needs to invent those.

Peter Wieck Melrose Park, PA

You should be able to track a short this close to the AC input with a multimeter on diode test mode.

brodge rec, inverter fets bulk caps

to start.

RL

yes, and actually, tracking down the short on the downstream sections (modules) should not be that difficult either, using the same approach. After all the faulty component is SHORTED which should be easy to find. Yea, it is a PITA to check every single component (sort of the brute force way to check), but judicious studying of the schematic and identifying appropriate points to isolate and subsequently test can prune away a number of branches in the tree fairly quickly so as to speak.

Maybe bounce over to the tek google groups and ask if there are any common component failures that should be targeted first? Otherwise, enjoy getting intimate with the guts of your scope....

Turns out those early items in the power chain are fine, so I'm moving on down the line to the less obvious items. The service manual has a list of chief suspects to look at first, so hopefully that should prove invaluable.

Yup. Totally agree. I'm doing just that now.....

Got 'er!

I'm not at all familiar with the topology they've used in this PSU, but I believe it's the chopper transistor (I should have guessed) that's gone low-resistance: Q1050. It was releasing quite a bit of heat even at only 18VAC supply with the current limited via 4x40 bulbs to boot. I removed it from circuit to check and it was dead. So I'm hoping that's the only part that's been damaged. It's an IRF820 and I have a bunch of IRF840s so will press one of those into service in place of the burnt-out one. Pin-outs just the same, just a lower Rdson figure, which is no bad thing. One question, though: they've mounted these transistors without using a screw through the mounting tab of this TO-220 package. The devices are just 'leaning' against some white pad mounted on the heat sink (see photo). Is there some special grease required to mate the transistor to the white pad or can I just use the regular silicon grease for this? It doesn't seem like a very good way of ensuring thermal conductivity. :-/

That'd be 'silicone' grease; the picture showing studs over the power package indicates that an L-shaped or Z-shaped pressure arm (maybe three separate ones) was designed into the system. Why else would there be a threaded rod there?

If that's the case, maybe it was removed at some time, shortening the life of the chopper. I'll do something about that regardless. BTW, the threaded rods are there to retain plastic covers which encase the 3 devices (don't ask me why as I haven't a clue what purpose they serve but certainly not to press bear the transistors to the heatsink, that much is certain).

The 'plastic' covers may serve the purpose of forced device retension.

Depends on material composition and design.

'just leaning' is not what is intended here. ...

Diode test for other shorts.

Check resistance gate to source pads of empty fet location.

Should read diode drop minimum, or gate drive cct is compromised.

RL

"cct"? Is that short for circuit?

Thank you, Legg, I shall duly do so. I'm just hoping the transformer windings are still okay and that the shorted chopper hasn't sucked too much current through them, 'cos that would probably be terminal for the scope.

I would guess insulation - is that part connected to mains?