CFL filament burnt - jump it?

[ben7]

So I recently had a CFL go dead on me. Well, I took it apart, and the electronics look good. Was in an upright position, so the PCB hasn't turned black xD

I wrote the date on it when I put it in ... it's almost 5 years old.

But, one of the filaments has burnt out. I bypassed the open filament and the bulb works again.
Now, for any lamp experts here, Is it ok to have it run like this? I know it won't heat the gas/mercury up as fast, so could this damage the ballast in the long term?

I measured the current draw and it was about 160mA ... the bulb is rated for 200mA ... go figure. Interestingly, as the bulb warmed up, the current increased! Just by a few tens of mA.

-Ben

[kaboom]

Re: CFL filament burnt - jump it?

If you were measuring AC line current, that increase is normal.

As the tube heats up, more voltage appears across it. Since lamp current is more or less regulated, a voltage increase corresponds to an overall power increase.

This is reflected by the input current going up- the "ballast" is a form of buck converter.

What you really did was connect both ends of the (open) filament together, obviously connecting an area with more cathode material "closer" to the circuit.

You could often do something similar with linear fluro tubes, to get a few weeks/months out of them when they got cranky, by rotating 180 degrees (not end-for-end) in the sockets.

[ben7]

Re: CFL filament burnt - jump it?

As the tube heats up, more voltage appears across it. Since lamp current is more or less regulated, a voltage increase corresponds to an overall power increase.

Yep, that is what I thought.

What you really did was connect both ends of the (open) filament together, obviously connecting an area with more cathode material "closer" to the circuit.

lol, not sure what you meant by making the material "closer" to the circuit.

-Ben

[SteveNielsen]

Re: CFL filament burnt - jump it?

When the cathode material fails the cathode separates into two parts opening the circuit. One part will contain more filament material than the other. Shorting the cathode leads completes the circuit again and puts a high potential on the remaining filament material.

[kc8adu]

Re: CFL filament burnt - jump it?

my educated guess is that there is no emitter left and it will finish burning up what little is left.then the ballast will blow up.wont be as spectacular as the f96t12's we have in the shop though.

[SteveNielsen]

Re: CFL filament burnt - jump it?

I think the other cathode is probably going to fail before the ballast fails, but it depends on how tolerant the ballast circuit is. If it can tolerate the increased power dissipation then when the other cathode fails the circuit becomes open again. Remember, the "good" cathode has a higher voltage across it now. That is going to shorten its life further.

[stj]

Re: CFL filament burnt - jump it?

you just doubled the current to the remaining heater, it's not gonna last.

more interesting, how is it controlled?
i have seen 3 types.
a mylar cap between the heaters << this is really shitty.
a PTC - this is the correct way to do it.
and some cheap lamps that use continuous drive (the location for a ptc is linked) < very bad!!

[SteveNielsen]

Re: CFL filament burnt - jump it?

The cathode current remains the same (Kirchoff's Law) but the voltage has increased which increases the power dissipation (Ohm's Law). The end result is the same, the cathode runs hotter and the emitting material will become depleted faster and fail.

[stj]

Re: CFL filament burnt - jump it?

The cathode current remains the same (Kirchoff's Law) but the voltage has increased which increases the power dissipation (Ohm's Law). The end result is the same, the cathode runs hotter and the emitting material will become depleted faster and fail.

your missing the point,
if the heaters fail then you cant vaporise the mercury and the lamp wont start.

you can start a lamp with a big spike, but cfl's dont use that method because it's dangerous and the polyester cap is too big.

[TechGeek]

Re: CFL filament burnt - jump it?

So I recently had a CFL go dead on me. Well, I took it apart, and the electronics look good. Was in an upright position, so the PCB hasn't turned black xD

I wrote the date on it when I put it in ... it's almost 5 years old.

But, one of the filaments has burnt out. I bypassed the open filament and the bulb works again.
Now, for any lamp experts here, Is it ok to have it run like this? I know it won't heat the gas/mercury up as fast, so could this damage the ballast in the long term?

I measured the current draw and it was about 160mA ... the bulb is rated for 200mA ... go figure. Interestingly, as the bulb warmed up, the current increased! Just by a few tens of mA.

-Ben

Dude, solder a 10K ohm resistor across the leads. That outta reduce the voltage.

[SteveNielsen]

Re: CFL filament burnt - jump it?

your missing the point,
if the heaters fail then you cant vaporise the mercury and the lamp wont start.

you can start a lamp with a big spike, but cfl's dont use that method because it's dangerous and the polyester cap is too big.

What point? The one you just brought up that was not being specifically discussed? That's absurd!

[kc8adu]

Re: CFL filament burnt - jump it?

post pics of the outcome.
my guess is the excessive cathode fall will overheat and melt the glass.

[ben7]

Re: CFL filament burnt - jump it?

more interesting, how is it controlled?
i have seen 3 types.
a mylar cap between the heaters << this is really shitty.
a PTC - this is the correct way to do it.
and some cheap lamps that use continuous drive (the location for a ptc is linked) < very bad!!

It's got the cap between the heaters. It's a GE lamp. I've taken apart a number of different CFLs and never seen a PTC. I guess it must be rare.
They always have the cap-between-the-filaments. Actually, one bulb (a n:vision) had blown this cap. It had a split in the side of it. I replaced it and the bulb worked again. I think it is still in use as well

Dude, solder a 10K ohm resistor across the leads. That outta reduce the voltage.

I think the 10K resistor might burn out, with the kind of voltages that are used here. Also, the "normal" filament resistance is around 2 to 3 ohms (from what I have measured in the past).

if the heaters fail then you cant vaporise the mercury and the lamp wont start.

Yes, I know, but, what if only one heater has failed? That is what I am asking about.

post pics of the outcome.
my guess is the excessive cathode fall will overheat and melt the glass.

I haven't put it back into use yet.
By the way, what exactly is cathode fall? I looked it up, but, why would only one filament being used cause "excessive cathode fall"?

I think the other cathode is probably going to fail before the ballast fails, but it depends on how tolerant the ballast circuit is. If it can tolerate the increased power dissipation then when the other cathode fails the circuit becomes open again. Remember, the "good" cathode has a higher voltage across it now. That is going to shorten its life further.

Yeah I thought it would shorten the life. The HV cap in there between the heaters is 4.7nF. I wonder if I halve that, maybe the current will be reduced to a more normal level on the remaining filament? Either that, or I can put a capacitor in parallel with the filament.

my educated guess is that there is no emitter left and it will finish burning up what little is left.then the ballast will blow up.wont be as spectacular as the f96t12's we have in the shop though.

Well, the bulb was run while not working for a while. It didn't blow up, interestingly.
The circuit closely resembles this:
http://cdn.makezine.com/make/blogs/b...kea7wschem.png
Except there is no R5 (PTC). I think there is also no C4.

[stj]

Re: CFL filament burnt - jump it?

btw, a tube will last forever other than heaters and phosphor.

i have a 30year old atleast UV tube in an eprom eraser with broken heater.
i just warm the tube with the hot-air station before i plug it in and it strikes and runs every time!!!
(yes, i know i'm cheap!)

[ben7]

Re: CFL filament burnt - jump it?

i have a 30year old atleast UV tube in an eprom eraser with broken heater.
i just warm the tube with the hot-air station before i plug it in and it strikes and runs every time!!!
(yes, i know i'm cheap!)

L - O - L !!!

You sir, get the cheapskate of the year award!!!

[TechGeek]

Re: CFL filament burnt - jump it?

btw, a tube will last forever other than heaters and phosphor.

i have a 30year old atleast UV tube in an eprom eraser with broken heater.
i just warm the tube with the hot-air station before i plug it in and it strikes and runs every time!!!
(yes, i know i'm cheap!)

Nice. Another great way to save electronics. You might wanna consider attempting to rejuvenate the heaters or replace them.

[stj]

Re: CFL filament burnt - jump it?

i dont have a vac chamber or glass blowing skills.
besides, a new tube is only 5-10€
(but that € is better spent on beer. )