Elevated tube temperature and power draw for CFL

[CFL2009]

I've been studying the effects of enclosures on CFL performance and would like to confirm something with regards to power draw. I tested the same CFL bulb in a tube-up orientation both with and without an airtight enclosure around the ballast and tube, and measured the ballast temperature and the power draw. The light was left on for 1 hour to reach steady state, and ambient temp. was 69 deg F. I bored a tiny hole through the case of the ballast to poke the thermocouple in, and it was fastened to not move between the two tests.
Granted, my sample size is one and I'm not applying PF correction to get true power, but I was more interested in the difference between the two cases Here are the results:

CFL tested: GE 71804 15W spiral

Uncovered:
Ballast temp: 112 deg F
Center of Tube temp: 105.6 deg F
Power draw: 97.7 mA (AC) (Correction: This is on a lower lamp setting)

Covered:
Ballast temp: 114 deg F
Center of Tube temp: 180 deg F
Power draw: 79.6 mA (AC) (Correction: This is on a lower lamp setting)

This suggests an 18% reduced power draw when the bulb is enclosed. For those CFL gurus here, does this make sense or is there something else going on? Does the vapor consume less power when it's hot? And will the phosphors give out at these higher temperatures, eliminating any gain from the power improvement?

 

[JohnR66]

When I did my "true" power tests as posted here, I noticed the lamps drawing more current until they warmed and stabilized (then I took my reading).

I believe these lamps use a mercury amalgam and most operate at a certain temperature range for optimal performance. The phosphors output is affected by temp as is their useful life. Hopefully someone can give concrete answers as I am not an expert here.

It would interesting if you could perform your ballast temperature test base down vs base up use. My parents have 40 watt CFLs in the base up fixtures mounted to the rafters in the ceiling of the basement. These are not recessed so the only issue is the heat from the bulbs rising to the ballast. One has died and one is dim and strobing at 60 Hz as if the filter cap dried up and half the fullwave bridge is open. Less than 1000 hours life!

 

[jtr1962]

It would interesting if you could perform your ballast temperature test base down vs base up use. My parents have 40 watt CFLs in the base up fixtures mounted to the rafters in the ceiling of the basement. These are not recessed so the only issue is the heat from the bulbs rising to the ballast. One has died and one is dim and strobing at 60 Hz as if the filter cap dried up and half the fullwave bridge is open. Less than 1000 hours life!

I actually did perform similar tests with some LED drivers I made. Once you hit about 75°C, the inductance of the inductor starts falling rapidly. At around 80°C the driver failed entirely when the inductance fell to more or less zero. I imagine it's the same with CFL ballasts which are a similar type of circuit dependent upon inductors made of ferrite material. In the situation you're describing the ballast easily gets too warm to last. If a CFL is going to be used in a base up fixture, especially a higher wattage CFL, then you need to have some sort of ventilation in the fixture (preferably fan-forced). It doesn't need to be a hurricane. Even slow air movement can be 10 times better than still air.

 

[CFL2009]

It would interesting if you could perform your ballast temperature test base down vs base up use.

I've got some time today, so I'll take your suggestion to measure current draw ballast-up to see if I get the same results.

I've done some additional research since I posted this, and I think I've found that the largest degree of degradation with respect to the tube is at the ends. If higher temperature causes the tube to fail, it will be due to the phosphor and mercury compounds at the ends of the tube, which is why dead tubes turn black at the ends when they fail (?? Anyone can correct this statement if not true).

So the question is, if we add a heat shield to the base of the spiral such that the ends of the tube poke through (and are kept thermally seperate), can we get the same lifespan with an 18% reduction in power consumed? There is a company called PowerLux that makes a light fixture that does just that, but you have to retrofit the entire fixture. They have shown extended bulb life, but make no claims about reduced power consumption.
It appears as if GE has some IP around putting a heat shield between the ballast and tube:
http://www.google.com/patents/about?id=wgMjAAAAEBAJ&dq=5691598

So this may be one more reason why the cheap twistie bulbs don't last, if they don't pay up to GE they can't put a heat shield there. These are all finer points the common consumer doesn't get, but they continue to buy millions of these things and then chuck them in the trash when they burn out upside down. Does anyone else see a huge mounting problem with this?

 

[CFL2009]

I've got some time today, so I'll take your suggestion to measure current draw ballast-up to see if I get the same results.

OK, here are the results of the inverted test. Note that I corrected the original current draw above after I realized the lamp was on a lower power setting, but the setting was the same between the tests. The results below are at the full power setting:

CFL tested: GE 71804 15W spiral
Configuration: Inverted (ballast up)

Uncovered:
Ballast temp: 144 deg F
Center of Tube temp: 96 deg F
Power draw: 157 mA (AC)

Covered:
Ballast temp: 149 deg F
Center of Tube temp: 175 deg F
Power draw: 130 mA (AC)

This once again shows that there is a significant decrease in power consumed when covered (17% in this case). I'd have to assume that it's due to the tube temperature. It's interesting to note the ballast temperature increases a huge amount between tube-up and ballast-up, but the tube temperature difference is small.

Comments welcome!

 

[JohnR66]

Thanks for running the test. Although expected, the ballast temp was much higher, but 150F is nothing critical. I'd bet those 40 watters that are dying must be getting much hotter.

I don't know the nature of your covering, but ballast temp is not affected that much.

I'd say base up position is not ideal for CFLs as the higher wattage ones literally cook themselves to death.