Need help fixing a halogen transformer

[BigBreaker]

I have some Lightolier mini spotlights on a track in my place.  The fixtures have 12V halogen bulbs in them around 15 watts or so and the track is 120v AC so a transformer is crammed into the fixture post.

I had two fixtures go out, and not quite at the same time.  One had a sooty bulb - no mystery there, but the other one looked fine.  I swapped the good looking bulb into the other fixture and it worked.  One down, one to go.  That also gave me some good evidence that the transformer on the remaining fixture was blown.

I checked online and the fixture costs $140+ new.  That seems like a big margin over the cost of a couple of 2 amp MOSFETs.  I'd like to fix the transformer instead.

So I popped off the cover on the transformer stem and pulled out the guts.  There were three main components: two TO-220 packaged MOSFETs and a 12:1 toroidal transformer.  the rest of the board has several diodes, resistors and caps.  No component shows obvious damage.  My guess is that one or both of the FETs blew.  Sadly the FETs are marked with an M that appears to identify it as a Motorola and "NL805 BUH".  Motorola seems to have transferred their discrete semiconductor biz to "ON-semiconductors" but I can't track down those markings with "ON" or anywhere else.  How to replace FETs when you don't know what they are?

Anyone have any advice?

[BigBreaker]

Sorry - the markings on the FET are NL807 and then BUH underneath.

[BigBreaker]

Further research turned up something at digikey with BUH in the part code.  It is a line of TO-220 transistors from ON semiconductors.  Seems I was wrong to assume FET.

[wooferhound]

I don't know about the transformer . . . but you can easy make the lightbulbs for 12vac track lighting last much longer. Just put 2 diodes back-to-back and put this set of diodes in series with the lightbulb itself. This will drop about a volt and it's easier on the filament which will last longer. Be sure that the diodes don't get heated up by the Bulb.

You can dim any AC lightbulb to 66% brightness by putting a single Diode in series with the bulb. The resulting half-wave power makes the light dimmer and the bulb will last 20 times longer.

[jimjjnn]

12 volt halogens with regulator circuits? Only circuits like that that I know of are for dimmers.

If they don't have dimming circuits, I would just get a 12 volt output transformer tp power them.

As someone else stated for longevity, use diodes in the line to lower the voltage.

[BigBreaker]

Yes, this is a on a dimmer, so it is safe to assume this is an ac to ac transformer.

The toroid core is small so I think the two transistors are set up to chop?

I could use my multimeter to measure frequency in a working fixture.  If it locks on to 60hz then there is no chopping, but if it sees a big number and/or keep hunting around bigger number than it think it's a chopper.

Perhaps some wisdom from someone who has worked with there circuits would help?

Just to be clear... I want this to would with my dimmable track.  Hooking it up to a separate 12V DC supply is not a solution.

[oztules]

If the trannies blew without some sort of current limiter, then you would see a sign.... so hunch.. check to see if there is a ac .1uf or similar in series with the line... it may be open circuit.

If the trannies form a simple multivibrator to drive the toroidal, and if they failed without current limit of sorts, then bang.... didn't happen, so check for voltage at the trannies.... I'm guessing there is none... open cap.

No, I have never seen them before, but thats my best guess with the info supplied.

If no-one with experience of these things replies, maybe a pic would help to unravel it.

.........oztules

[wooferhound]

You should NOT Dim the power going into a Transformer. The power coming out of a Dimmer resembles Pulse Width Modulated power and it will burn up a transformer, smoke and all.

 

[jimjjnn]

My oops !!

I meant to put diodes between 12 volt transformer and light NOT line.

[BigBreaker]

The lights and dimmer circuit were installed by an electrician and the other lights on the same strip work just fine.  We can operate under the assumption that the 120v to 12v electronics in the fixture are at fault.

I noted that it was on a dimmer because I believe that implies an ac mode step down circuit rather than an ac to dc step down circuit with voltage regulation.  If the circuit tried to maintain 12vdc at all times, the lights wouldn't dim but the current would get out of hand.

I'll grab a picture of it when I can.

[oztules]

Woof,

I find this difficult to comprehend. The basis of most pwm circuits, is to feed a pulse train (ideally square) to a transformer to work.

Eg. in a computer type supply, you drive the driver transformer with square waves, which drives the power switchers with a square pulse train, that drives the power transformer with an even larger square wave.

There are ranges of commercial traction power chargers that use triac switching (simple dimmer mode) to drive the 1kw transformers as normal for variable output.

Can you be more explicit (or why am I missing something) please.

Also my 2.5kw multivibrator inverter works well with square wave input. (48vdc-240ac 50hz) 40 lbs transformer used here... homebrew.

............oztules

[wooferhound]

I have worked with dimmers by doing Lighting for Live events and doing Disco lighting in nightclubs. I have seen many fried transformers over the years caused by dimming a transformer. Some of the funny waveforms that a dimmer puts out cause high Back Voltages in the transformer as the magnetic fields collapse, this is what makes the windings turn black and crispy.

I have heard that if you have a 5 amp load, you can use a 10 amp transformer with a dimmer without a problem, but I have always considered it to be Bad Form to dim the power going into a transformer.

[BigBreaker]

I can often hear the circuit buzzing at 50% dim, so that supports your argument.  The transformer looks fine to me though.  I can check it.

[oztules]

That makes sense Woof.

I suspect that cheap lighting dimmers (for incandescents) can have a asymmetric output. We need the current after the dimmer output to be symmetric, to stop a DC component forming in the transformer that may cause saturation in the transformer core... leading to failure. I doubt the flyback voltage would bother it though, and that the spurious voltages would be more a problem for the triac. Thats why I think you have been advised to use a beefier transformer in these instances (bigger core less..... likely to saturate)

In this case, he seems to have a professional install, and so I figure a good symmetrical dimmer is used.... probably a transistor rather triac based design.

I am guessing a dumb multivibrator driving the toroidal core chops the mains at say 40khz to drive the toroid.... being dumb.... less  power in,less power out.

........oztules

[BigBreaker]

That was my guess too Oz.  There are no ICs on the board, so it can't be that smart.