Fluorescent Lamp Inverter

[JDW]

The following URL hosts some photos of a Fluorescent lamp I've been using for about 13 years.  It's been working fine until recently.  It now won't power on.  I recently changed the tube so I know that is good.  I confirmed the push switch is good.  The fuse on the inverter circuit board checked out good.  Yet it won't power on.  I know nothing about how these inverters work, so I'm not sure how best to test it further, hence my post here.

https://www.icloud.com/sharedalbum/#B0w52plgjNZIfl

I thought the small electrolytic capacitors (C5 & C16) might be leaking, so I replaced them but the lamp still won't power on.  And I know the power cord is good because I can measure the correct wall voltage coming through it to the circuit board.

Any testing tips or suggestions you could give me would be appreciated.

Thank you.

[goaty]

If you have a multimeter you could check the diodes and transistor for shorts and alike. I had one repaired where the power transistor (like your 2SC4546 NPN) was shorted.

[JDW]

If you have a multimeter you could check the diodes and transistor for shorts and alike. I had one repaired where the power transistor (like your 2SC4546 NPN) was shorted.

Thanks.

Diodes D1, D2, D3, D4 and D5 all show their forward voltage just fine with my diode checker, tested in-circuit.  I desoldered the 2SC4546 (NPN BJT) and put it on a breadboard and wired it up to switch GND to a test LED and it worked fine.

Any further thoughts?

[JDW]

I guess I've stumped the experts here!   

Perhaps some electronics aren't meant to be fixed.

Thanks.

[fzabkar]

I would check the capacitance of the film caps. I also make it a practice to reflow the solder joints at all coils and transformers.

[JDW]

I would check the capacitance of the film caps. I also make it a practice to reflow the solder joints at all coils and transformers.

Thank you for the advice.

I removed all 6 film caps and measured the correct capacitance on all of them, as per their markings.  Not sure what ESR should be, but measuring with my DE-5000 at 100kHz, ESR on most of them were under 1-ohm, although C15 was about 2.5-ohms at that frequency.  Not sure if that is high or normal.

I then reflowed the solder joints at all coils and transformers and then tested.  Still a NO-GO.  My lamp won't light.

Any further ideas would be appreciated.

[soldar]

Ifr you really want to spend time on this project I would start by reverse engineering the schematic.

But IMHO, if you want to be practical, I would go for an LED replacement.

[james_s]

Those are usually based on a Royer oscillator. Look for dipped film capacitors, especially if there is one across the fluorescent tube to allow current to flow to heat the cathodes before the tube strikes. I remember years ago I repaired quite a few of those curly CFL bulbs just because I could and often I'd find one of those film capacitors had shorted.

[QQE06-40]

Just a thought, because i had this happen several times before with SMPS's: check the higher ohmic resistores (>100k). They sometimes go o.c. with no obvious signs and for no obvious reasons. I see some of them on your pictures (R2/R22/R3/R4 e.g.). Have You checked R5? Seems to be a low ohmic fuse-resistor from the emitter of the switching transistor to ground.
Greeting,
Wolf

[Kartman]

I’d second checking the resistors. The smaller resistors suffer from electromigration at higher voltages and go open circuit. I spied a couple of 330k resistors near the transistor - i’d suspect them.

[JDW]

Gentlemen, thank you for your suggestions.

R5 is marked "1R0J 1W" and measures 1.3ohms in-circuit.  It is light in color and easy to see there are no visible burn marks.  So this part would appear fine.

R24 measures 98.3k in-circuit.  Bands indicate it is a 100k.  So this too is OK.

R22 measures 231k in-circuit.  Bands indicate it is a 240k.  OK.

R2 measures 332k in-circuit.  Bands indicate it is a 330k.  OK.

R3 measures 325k in-circuit.  Bands indicate it is a 330k.  OK.

R4 measures 322k in-circuit.  Bands indicate it is a 330k.  OK.

R7 measures 33.5 in-circuit.  Bands indicate it is a 33ohms.  OK.

That seems to be all the resistors.  No problem here.

The 3 blue color caps are not "dipped film" that I can see insofar as they feel as hard as a rock and I must assume them to be ceramics.  I didn't test these yet but there is not visible cracking on any of them.

Hmmm...

[james_s]

Film caps are hard as a rock too, and when any of these fail you can't tell just by looking at them.

[JDW]

Film caps are hard as a rock too, and when any of these fail you can't tell just by looking at them.

I desoldered all the blue caps and tested outside the circuit:

C8: Marked "104U" and measured 102nF, ESR=7.8-ohm @120Hz
C9: Marked "474" and measured 473nF, ESR=2.9-ohm @120Hz
C25: Marked "184" and measured 183nF, ESR=8-ohm @120Hz

No problem here.

[xmetal]

I vaguely remember having a similar problem with a circuit like that some years ago. It turned out to be the input voltage to the device was too low.

[JDW]

I vaguely remember having a similar problem with a circuit like that some years ago. It turned out to be the input voltage to the device was too low.

If by "device" you mean the entire PCB I've been talking about, let it be known that the voltage is a wall socket.  It's designed for Japan to run on 100VAC at either 50Hz or 60Hz (since Japan uses both frequencies, depending on area).

[xmetal]

I vaguely remember having a similar problem with a circuit like that some years ago. It turned out to be the input voltage to the device was too low.

If by "device" you mean the entire PCB I've been talking about, let it be known that the voltage is a wall socket.  It's designed for Japan to run on 100VAC at either 50Hz or 60Hz (since Japan uses both frequencies, depending on area).

I can't remember what the source voltage was on the device I was working on but it was lower than it should have been and was stopping the fluorescent from working. It created the same symptoms as yours so I thought it was worth pointing out.