Fluorescent desk lamp repair

[redwire]

Re: Fluorescent desk lamp repair

This is cold-cathode? I see only two red wires going to the tube, but the fixture's schematic shows four wires.
Philips Circline fluorescent tubes are all 4-pin hot-cathode (for starting) and they mention run voltage 61-84V for the 20-30W class T9 (12") size. You don't need really high strike voltage when the heaters are on, I've manually lit tubes then with even 24VAC.

I think it's not a good idea to measure output voltage because it's high frequency AC which many multimeters can't read, chinese multimeters rolloff above 2kHz for example. Also, it should be high voltage with no load (tube not lit) which can kinda damage a multimeter.

[caphair]

Re: Fluorescent desk lamp repair

@redwire Yes that’s the one you’ve linked. There are 4 wires that connect to the lamp itself further up and two of them have a capacitor across them.

What’s a good way to test if the circuit is functioning properly? Can I just ohm check the lamp to determine if it’s not open inside?

Do you know what voltage I should expect at both transistor pins?

[eccerr0r]

Re: Fluorescent desk lamp repair

The two "ends" of the tube should appear open circuit at the voltages of a DMM, and the two filaments should be like incandescent filaments.

Oddly enough I had a iinear F15T8 fixture that had a magnetic ballast that I replaced with a CFL circuit. The lamp turns on right away -- however, the CFL circuit does not heat the ends hot enough so basically it's running the lamp CC and though some tubes light, there are some F15T8 bulbs that do not light up in the fixture anymore.

[redwire]

Re: Fluorescent desk lamp repair

At the lamp, you should get low continuity at each end (between the two filament pins) and of course nothing end to end. Compare it to the old tube.

It seems like the oscillator is working because you measure some ACV on the output.
I would test the 47uF electrolytic capacitor that had blown up before and the film cap at the tube.
How it seems to work is the filaments at each end draw power through the film capacitor that is across the tube ends. It's just a dumb oscillator going for max all the time. No starter circuit.

[RJARRRPCGP]

Re: Fluorescent desk lamp repair

The two "ends" of the tube should appear open circuit at the voltages of a DMM, and the two filaments should be like incandescent filaments.

Oddly enough I had a iinear F15T8 fixture that had a magnetic ballast that I replaced with a CFL circuit. The lamp turns on right away -- however, the CFL circuit does not heat the ends hot enough so basically it's running the lamp CC and though some tubes light, there are some F15T8 bulbs that do not light up in the fixture anymore.

Do the Phillips Alto F15T8s light up in that condition?

[eccerr0r]

Re: Fluorescent desk lamp repair

Do the Phillips Alto F15T8s light up in that condition?

I don't have an Alto F15T8 lamp, I don't remember what brands of LFL bulbs I had tested with. Chances are one is a GE and other is a Sylvania, both old and only one lights with the CFL ballast while both work with the magnetic ballast. The CFL bulb that I stole the guts from of course would be something like a spiral 13T4 or something like that, but that also did not noticeably heat the filaments on the spiral, unlike with the original automatic thermal starter on the F15T8.

[RJARRRPCGP]

Re: Fluorescent desk lamp repair

I don't have an Alto F15T8 lamp, I don't remember what brands of LFL bulbs I had tested with. Chances are one is a GE and other is a Sylvania, both old and only one lights with the CFL ballast while both work with the magnetic ballast. The CFL bulb that I stole the guts from of course would be something like a spiral 13T4 or something like that, but that also did not noticeably heat the filaments on the spiral, unlike with the original automatic thermal starter on the F15T8.

The ballasts I have for F15T8, are the usual new solid-state stuff, TMK.
(most likely a higher-than-AC-input-frequency switcher) (And it has to use a cap, TMK)

I think all the T12 ballasts I have are low-frequency, (60 Hz and the like) they sound like the classic ballasts.

[caphair]

Re: Fluorescent desk lamp repair

The lamp tests good. The capacitor not sure if it's good or bad but could only measure in ohms and gives a reading of 1.2 ohms across it when disconnected from the lamp

I may have previously measured voltages incorrectly. Those reading were DC when I try measuring AC it only reads 1.2vac at the wires going to the lamp

[eccerr0r]

Re: Fluorescent desk lamp repair

Well that seems suspicious?
Capacitors should show up as open or eventually open once it's charged, unless you're measuring its ESR then you want it as low as possible?

[caphair]

Re: Fluorescent desk lamp repair

Well that seems suspicious?
Capacitors should show up as open or eventually open once it's charged, unless you're measuring its ESR then you want it as low as possible?

Can I test if it'll light without that capacitor? Don't have another to replace

[eccerr0r]

Re: Fluorescent desk lamp repair

This is the cap that bridges two pins of the tube (that connects the two heaters together)? Might be able to test it, but without it, it's really running cold cathode now. However with it shorted it would cause other issues...

[stj]

Re: Fluorescent desk lamp repair

something i'v wondered about these circuits,
how do you set the current/brightness?

[caphair]

Re: Fluorescent desk lamp repair

This is the cap that bridges two pins of the tube (that connects the two heaters together)? Might be able to test it, but without it, it's really running cold cathode now. However with it shorted it would cause other issues...

Yes it bridges one of the two filament pins. I tried without the cap but still nothing.

Im confused on how you would test this circuit if the output voltage is high frequency. Is an oscilloscope required?

And is my understanding correct that mains ac is converted to dc and then back to higher ac voltage to get the lamp running?

[eccerr0r]

Re: Fluorescent desk lamp repair

yeah without the cap the heaters wont heat and need a higher voltage to cold cathode start the lamp.

You need a faster high bandwidth RMS multimeter if you want faster results, or you could rectify the voltage with a diode and filter with a capacitor, but be careful of loading down the circuit.

Yes pretty much these things do a switcheroo from ac to dc and back to ac. The higher frequency is the main reason for doing this, not necessarily the higher voltage.

[caphair]

Re: Fluorescent desk lamp repair

yeah without the cap the heaters wont heat and need a higher voltage to cold cathode start the lamp.

You need a faster high bandwidth RMS multimeter if you want faster results, or you could rectify the voltage with a diode and filter with a capacitor, but be careful of loading down the circuit.

Yes pretty much these things do a switcheroo from ac to dc and back to ac. The higher frequency is the main reason for doing this, not necessarily the higher voltage.

So at the output of the tea sudores it would be ac voltage again? Just seeing what I could measure before that point to help see if the transistors are doing anything.

Curious how does higher frequency but not higher voltage help with turning the lamp on?

[eccerr0r]

Re: Fluorescent desk lamp repair

it's AC again after the transistors but once again it's high frequency.

The higher frequency doesn't help turning the lamp on ... it reduces the cost of the lamp (and has a side effect of faster, smoother startup, usually. Ever used the old mag ballast lamps? the ones that flicker as it turns on? Gone.)

[caphair]

Re: Fluorescent desk lamp repair

Turns out it was the green cap across the lamp terminals. Didn’t have exact one so momentarily substituted with a 483pf one. Lamp lit then went off and puff of smoke came from the circuit board lol. Guess using the wrong cap for testing wasn’t a good idea? Lol

[eccerr0r]

Re: Fluorescent desk lamp repair

Yay magic smoke

483pF? not a standard value, weird.

Original appears to be 2200pF?

[caphair]

Re: Fluorescent desk lamp repair

Yay magic smoke

483pF? not a standard value, weird.

Original appears to be 2200pF?

At least that's what my capacitor tester said. Yeah original is 2200pf

Would the smaller cap cause the light to go out and cause secondary issues?

[eccerr0r]

Re: Fluorescent desk lamp repair

Would have imagined no cap could possibly do damage, then too small cap would be the same, and shorted thus be the 'safest'... but who knows how the circuit was made.