I have a halogen desk lamp with an internal transformer that recently stopped working (no light when switch goes on; fairly simple failure mode). The bulb still lights upon applying 12VDC directly, and there's no internal fuse I can see. Inspecting the transformer's internals, 120VAC shows fine at the input, and the dimmer pot appears to range reasonably from the range of megohms to zero.

The transformer is a fairly standard half-bridge design (in fact, it's almost textbook out of this application note down to the transistors used, which in this case are BUL39D's (and the cap is either C2/C3 in the schematic on p1).

The Cap In Question

Suspect cap after desoldering

I strongly suspect issues with either the cap, its corresponding transistor, or both, but the discoloration (more prominent when flash doesn't wash it out) makes me suspect the cap first. However, I'd like input on the following things before buying and trying a replacement:

  1. Am I reading the values correctly as 0.22 uF +/- 10%; 250V max?
  2. What type of cap is this in terms of construction?
  3. Other than needing to fit into a narrow space, is substituting a different type of non-polar cap with equivalent values likely to cause any problems in this application?
  • \$\begingroup\$ Discoloration is a bad indicator of a components health. Measuring its value is quite an ok indicator of its health. \$\endgroup\$ – PlasmaHH May 15 '17 at 9:12
  • \$\begingroup\$ I would hope there is a fuse somewhere even if it looks like something else to you. Check the transistors' junctions with a multimeter (diode check range), out of circuit. \$\endgroup\$ – Spehro Pefhany May 15 '17 at 10:40
  • \$\begingroup\$ As others say - do transistor diode check. For NPN BE and BC present as diodes (anode at base, conduct with diode test +ve on base). A slightly trickier test (4 arms helps) is CE as diode with high R bias resistor connected BC. Said resistor can be a tongue licked finger. Doing this with good transistor produces far more Ice than wet finger alone does. Then consider caps. \$\endgroup\$ – Russell McMahon May 15 '17 at 13:56
  • \$\begingroup\$ +1. Nicely put question. People tend to vote to close repair questions even though they are allowed when presented as this one is. It is an excellent example of trying to "design" a repair, It's relatively simple circuit wise (always nice), has circuit diagram provided (rare for repairs) and an intelligent asker. Well done :-) \$\endgroup\$ – Russell McMahon May 15 '17 at 14:00
  • 1
    \$\begingroup\$ @RussellMcMahon You also forgot well defined question scope. Was there something specific that I needed to see? \$\endgroup\$ – W5VO May 15 '17 at 14:23

I'll ignore the question of whether it's the capacitor that has failed and talk about the design:

  1. Yes. MD will be the manufacturer's product series.

  2. This film capacitor looks like a metallized polyester.

  3. The replacement could also well be metallized polypropylene. I've seen these used in the same role, snubbing the transformer in a halogen lamp power supply. The replacement capacitor has to be suitable not only for the voltage, but for the typical dv/dt seen as a result of the switching frequency used. Self-healing is also a feature of these caps.

Any old replacement with the same capacitance and voltage rating won't do. Don't put in a ceramic. You could put in a film capacitor with the same capacitance and voltage rating from a name brand and it would very probably work, but you won't get certainty without a lot more information about the design.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.