Flourescent lamp driver circuit

Papa_Bear

Hi all

I am trying to repair a small circuit board that drives a 24W FL lamp.

When it was given to me it had some dry joints which I fixed thinking that was all that was wrong with it but it still does not power the lamp when connected.

I have tested the lamp itself elsewhere and it does light (looks new as well).

I did some tests on the circuit for voltage at the output and all I am getting is 100+ v DC which is strange - should be AC sq wave of a few kHz or something?

I have also tested all the components insofar as I can (resistors, diodes transistors, coils etc) but can't really test capacitors as I have no ESR meter.

There's a good deal of heat from one of the transistors when I power the board so possibly a capacitor has failed but there is no sign of bulging on the electrolytics.

Anyone know of any known or common faults with these circuits?

Thanks

DublinDilbert

Yea it looks like a type of resonant converter. The Base/Gate of the transistors will be driven by Aux windings on the flyback trafo. This was very common on old TV HV power supplies.

One of the transistors has probably given up and needs to be replaced, measure between the collector/emitter (source/drain) terminals.

It should be fairly simple to draw up the circuit diagram.

Keplar240B

Does the blue electrolytic capacitor on bottom right have a bulge in it
hard to tell from angle
maybe i am seeing things (LOL). That can be a relativity "common" failure for them.
you said you did not check caps .

Papa_Bear

Does the blue electrolytic capacitor on bottom right have a bulge in it
hard to tell from angle
maybe i am seeing things (LOL). That can be a relativity "common" failure for them.
you said you did not check caps .

Actually Keplar, the bottom left cap has a slight bulge it the top of it similar to that shown in your photo - if the bulge is even that slight does that mean it is shot?
I thought it had to be bulging much more like on the sides too, no? They are 10uF 250v anyway.

One of the transistors has probably given up and needs to be replaced, measure between the collector/emitter (source/drain) terminals.

What am I supposed to be measuring exactly DublinDilbert? I tested for ~0.6v between bc & be with diode check on the meter and got as expected.
The transistors have F U S13003 written on them. Does the fact that I'm getting DC instead of AC at the output mean the transistor(s) is faulty?


In the event that the circuit cannot be fixed, could I use the circuit from the back of a CFL bulb to replace it as it does essentially the same job, although at half the wattage at best. Would that likely cause the lamp to fail prematurely or even worse, burn out the CFL circuit board?

Thanks for the help.

Keplar240B

Papa_Bear wrote: »

Actually Keplar, the bottom left cap has a slight bulge it the top of it similar to that shown in your photo - if the bulge is even that slight does that mean it is shot?
I thought it had to be bulging much more like on the sides too, no? They are 10uF 250v anyway.
...........

Thanks for the help.

The bulge is usually on the top in my experience and can be quite small.

see this link.
https://en.wikipedia.org/wiki/Capacitor_plague

whizbang

F U S13003 =MJE13003. Plenty of easily available substitutes.

Check your meter can read high Frequency ac. This converter is probably running at 50Khz, A lot more than most meters can display.

Wiring of the output can be a bit strange, I presume its a 4 pin lamp.

Papa_Bear

Check your meter can read high Frequency ac. This converter is probably running at 50Khz, A lot more than most meters can display.

Wiring of the output can be a bit strange, I presume its a 4 pin lamp.

The meter I was using does not measure frequency whizbang - it was voltage I was measuring - first AC with nothing registering on the meter and then switched to DC where I got +100.

The pins of the lamp are 4 in number, with the two inner ones connecting across a non-polarized capacitor which as you probably know is a common arrangement (you can see it on the circuit of a CFL bulb. The output is taken across the two outer pins.

The TO22 (transistor?) mid left of photo looks like it blew a chunk off.

Papa_Bear

The TO22 (transistor?) mid left of photo looks like it blew a chunk off.

Sir Liamalot, there is a hole in the centre of the transistors for mounting purpose as you know but otherwise they are as they should be. There's no pieces missing of them.

What is your opinion of my suggestion to replace the board with the circuit of a CFL bulb?

Cheers

Papa_Bear

Papa_Bear wrote: »

What is your opinion of my suggestion to replace the board with the circuit of a CFL bulb?

That's all I'd do. Line in, line out, Bob's yer Auntie...not worth your time otherwise for a 24W fixture unless you like troubleshooting.

Papa_Bear

Papa_Bear wrote: »

although at half the wattage at best. Would that likely cause the lamp to fail prematurely or even worse, burn out the CFL circuit board?

If the fixture is half the wattage of the driver board no worries. If the driver is half the wattage of the fixture expect smoke.

Papa_Bear

If the fixture is half the wattage of the driver board no worries. If the driver is half the wattage of the fixture expect smoke.

I kinda thought as much. Unfortunately the fixture is a 24W lamp and the highest wattage CFL bulb I've ever come accross is 11W so the driver circuit out of that wouldn't hold up running a 24W lamp.

Oh well back to trying to fix the original board.

Cheers

Perhaps a tube choke and ballast is a better option?
You can get 28W 2D fluorescents too they might have the gubbins you're after.

What about retrofitting LEDs? Fluorescent light is pretty industrial. Good efficacy but low CRI and high colour temperature.

I'm not convinced these SAD lamps are worth their salt. Amber LEDs put me in a much better mood than bare naked 6500°K. I can hear the high frequency drivers too.
I've boxes of them I've confiscated from MOH.

Papa_Bear

Perhaps a tube choke and ballast is a better option?
You can get 28W 2D fluorescents too they might have the gubbins you're after.

I have been thinking about this but im limited in space requirements. It's actually an old barbers pole im trying to fix so not sure if i can fit the balast inside the enclosure and retro fitting another type of light wont pay most likely.

Cheers for the suggestions Sir Liamalot

You could rebuild the driver on veroboard with new components for not a lot of dosh. I'd probably salvage the transformer, terminal blocks and inductor, replace the rest unless you're 100% sure they're within tolerances and operational.

Papa_Bear

I did replace the electrolythics and am trying to find suitable transistors at the moment but I reckon thats about as far as Im willing to go cause Im not good with colours and I have dranw out the circuit but it does not make a whole lot of sense to me ie. I cant figure out how one of the transistors ever gets biased into conduction, unless ive drawn it incorrectly but ive been over it a few times.

I do have an old 58w ballast knocking around that is still working that might fit, but as I understand things, would that not overpower a 24w tube and cause it to blow up leaving fine glass and mercury vapour everywhere?

A FL tube will just keep drawing any available current to it until destruction; and ballast and driver circuits have to be matched to the tubes, especially those CFL ones.


Thanks again Sir Liamalot

DublinDilbert

There's usually some path to turn on one transistor initially.

Look here,mist probably similar;
http://www.eleccircuit.com/10-30kv-tv-flyback-driver-with-2n3055/