Strange observation

[Sierrasparky]

I was in the local Hardware store over the weekend and saw a man purchasing Mogul to standard base socket adapters and 26 watt CFL's .
I asked him and he told me he was putting them into a MH type fixture and was not bypassing the ballast circuit. He was doing this to save consumption.

I am a bit baffled .
won't he be wasting energy powering the Ballast?

 

[augie47]

heard any report of fires yet ?

 

[Dennis Alwon]

Say goodbye to that CFL

 

[Sierrasparky]

funny thing is he says they work fine

 

[ultramegabob]

funny thing is he says they work fine

dont believe everything you hear....

 

[hillbilly1]

They will last a couple of days,maybe, I goofed once with some fixtures that had edison base lamps that looked exactly like par lamps, the customer complained they were constantly burning out, so I replaced them with CFL par's. lasted a couple days and burned out again. Opened up the fixture to check the voltage and found a metal halide ballast! Very expensive lamps, about $120 bucks each. If it were not for voiding the UL listing (brand new BMW dealership), I would have bypassed the ballast and used the CFL's, they put out as much light as the MH.

 

[electricalperson]

the output of the ballast depends on the wattage of the lamp. it doesnt matter if you feed it with 120, 208 277 or 480 the output stays the same. usually the output of the ballast is higher than the rating of the CFL lamp. those lamps will be toast in no time

 

[physis]

It's all just current and voltage. Too much voltage makes too much current and too much of either one makes toast. I guess you could mix power in there somewhere. if you tried really hard. :smile:

 

[brantmacga]

I was in the local Hardware store over the weekend and saw a man purchasing Mogul to standard base socket adapters and 26 watt CFL's .
I asked him and he told me he was putting them into a MH type fixture and was not bypassing the ballast circuit. He was doing this to save consumption.

I am a bit baffled .
won't he be wasting energy powering the Ballast?

he may have been purchasing them for someone else to install; and that person most likely did disconnect the ballast, if they are working fine.

 

[glene77is]

he may have been purchasing them for someone else to install; and that person most likely did disconnect the ballast, if they are working fine.

Brant,
I agree.
The startup voltage for MH is high, low current being drawn during warmup.
When the lamp ignites, then high current is drawn, and the voltage drops.

IMO, The High startup voltage should overload the CFL power supply.

 

[mxslick]

But wouldn't the "electrical signature" (impedance, resistance, etc) of the cfl ballast affect how the MH ballast sees it? Maybe it just looks like a pure resistive load and thus the MH ballast would just act like a coil of wire in series? (IIRC there are several varieties of MH ballast like series, autotransformer, etc?)

And what about the ignitor?

 

[glene77is]

But wouldn't the "electrical signature" (impedance, resistance, etc) of the cfl ballast affect how the MH ballast sees it? Maybe it just looks like a pure resistive load and thus the MH ballast would just act like a coil of wire in series? (IIRC there are several varieties of MH ballast like series, autotransformer, etc?)
And what about the ignitor?

MX,

I was concerned that the initial start-up voltage (> 200 V),
looking into a Hi Z load,
would be too high for the CFL (expecting 120 V).
The CFL circuit is something like a full-wave rectifier, then a Capacitor,
and then the transistor oscillator circuit driving into a inductance coil.

On the other hand, and IMO,
it is possible for that
the intial CFL load to appear as a small black hole into the CFL capacitor,
in which case the MH ballast voltage would initially be low,
then rise as the CFL capacitor charges up.
This scenario might work, and I won't know until I hook one up
with my oScope to see.

I think the ignitor is a simple thermal bi-metal contact,
to introduce added resistance ,
to keep the current quenched/controlled after ignition.

Comments please.

 

[nakulak]

this whole thing seems extremely nutty to me. I think someone else suggested bypassing the mh ballasts and that's the only thing that makes sense to me, if the intent is to change the fixtures like this (which I guess amounts to building your own/violating the listing (?))

 

[Sierrasparky]

he may have been purchasing them for someone else to install; and that person most likely did disconnect the ballast, if they are working fine.

Nope he was installing them.
He said he did it before with no problem!

 

[LarryFine]

The CFL circuit is something like a full-wave rectifier, then a Capacitor, and then the transistor oscillator circuit driving into a inductance coil.

So, with a full-wave rectifier input, a CFL should operate equally well on DC. Even half-wave rectification would allow DC operation if you found the correct polarity.

 

[LarryFine]

I think someone else suggested bypassing the mh ballasts and that's the only thing that makes sense to me, if the intent is to change the fixtures like this (which I guess amounts to building your own/violating the listing (?))

From what I've read here, the new LED fluorescents require the same thing: 120v directly to the sockets.

I'd hate to think about what happend if an LED tube is installed in a non-modified fixture, and vice versa.

 

[mxslick]

From what I've read here, the new LED fluorescents require the same thing: 120v directly to the sockets.

I'd hate to think about what happend if an LED tube is installed in a non-modified fixture, and vice versa.

My guess is a standard tube in a modified fixture would explode both ends off as the tube filaments aren't designed for 120v. (They use low-voltage AC to heat them and the operating arc is struck from one end to another down the tube.)

An LED tube might not be harmed for the same reason, in that the voltage across the pins on each end is low voltage, and I don't see the led tube having any common connections from one end to the other.

As for the CFLs in MH fixture, here's a thought: since the ballast of a cfl is like a switching power supply, it could probably accomodate a fairly wide range of input voltages as long as it doesn't exceed the ratings of the input diodes or capacitors.

 

[glene77is]

So, with a full-wave rectifier input, a CFL should operate equally well on DC. Even half-wave rectification would allow DC operation if you found the correct polarity.

Larry,

These are two different devices, two different methods.
(1) an LED is a "Diode" device,
emitting light based on Band jumping luminance.
(2) a CFL is a compact "Flourescent" device,
emitting light by Mercury gas vapor conductance.


My CFL did not operate "equally well" on the rectified signal,
it was somewhat dimmer, certainly not brighter.
I did not measure the lumens. I was absent mindedly playing.
My point was that it did not 'appear' to be re-igniting,
at a higher startup voltage,
which I speculate would cause extra stressing
of the insulations on the ic board.
It appeared to last much longer than advertised.

Could it run on filtered DC ? Well, maybe.
I don't have the schematics,
but in concept it looks like it would run from filtered DC.

This is all good speculation, not verified by measured experiment.

Your comments are always welcome.

 

[iwire]

If I recall the open circuit voltage of a 400 MH ballast is around 180 volts.

Of course the pulse from the igniter might be well above that. :smile:

 

[LarryFine]

These are two different devices, two different methods.
(1) an LED is a "Diode" device, emitting light based on Band jumping luminance.
(2) a CFL is a compact "Flourescent" device, emitting light by Mercury gas vapor conductance.

I'm well aware of these facts. I was talking about the full-wave (4-diode bridge, probably) rectifier you mentioned as the first component in the CFL.

Your comments are always welcome.

As if you could stop me!