[ QUOTE ]
jtr1962 said:
I was laughing hysterically at parts 1 to 3 of your post, udaman, although I think the inside jokes will be lost on just about everyone here. I'm still trying to figure out the significance of those two links in the grand scheme of things. /ubbthreads/images/graemlins/thinking.gif I can only imagine what went through your mind when you saw my name and the thread title after reading my post in the other place. /ubbthreads/images/graemlins/faint.gif /ubbthreads/images/graemlins/ooo.gif
FS and SC with a male problem? I think you're on to something here, especially with SC. Seems like he's always going out of his way to brag about something about himself to perhaps cover up some other embarrassing deficiency. I just hope he's not interested in lights or he'll be pissed at both of us. /ubbthreads/images/graemlins/happy23.gif
Yes, small things are what I like. Small, delicate, and refined if you know what I mean. /ubbthreads/images/graemlins/icon11.gif None of that course in your face stuff with bulges everywhere. /ubbthreads/images/graemlins/thumbsdown.gif Anyway, back to the main topic since I think you're on to something here.
I opened it up today to take a look. Here is a composite of the inside and the cover:
It has that "burned out electronics" smell but none of the semiconductors on the board appear to be fried, unlike the Feit CFL ballast which blew both MOSFETs a few months ago. The black on the cap is just electrical tape of some sort. Most likely one of the transformers overheated because that is what the smell mostly reminds me of. Electrically, it's an open circuit. A day or two before it went, it had trouble starting the tubes. About half of the length of the tube would light up, and on the next try it would start. On the day it failed nothing happened.
It seems surprisingly simple inside-just a few passives, four diodes, 3 MOSFETS, and the transformers. No ICs so it's not a "smart" ballast like some of the better designs. Having the transformers sitting on the PC board with only still air to conduct away the heat is asking for trouble unless the power dissipation is very low. I've seen some ballasts which use either pitch or potting epoxy to conduct the transformer heat to the case more efficiently.
Evidently heat was the culprit here as udaman said. The fixture is in a dropped ceiling with about 6 inches of free air over it. The top of the fixture is only slightly warm except above where the ballast is. This part might be maybe 50°C. Given that these electronic ballasts are supposed to be more efficient I'm surprised at the temperature most of them operate at. Ballast factor for this ballast was .89, which means that the tubes were driven at roughly 26 watts each. Power input was 114 watts. This means about 10 watts was being dissipated in the rather small ballast case. Given the poor thermal path, no wonder it failed. I'm amazed it lasted as long as it did. I also wonder why these things don't do better in the power dissipation department. My cheap $8 shoplight from Home Depot dissipates maybe 3 watts in the ballast based on the slight temperature rise. Why can't a more expensive ballast like the one above do the same? I don't get it why most of the 120VAC electronic ballasts seem to be mediocre in the efficiency department. I slapped together a ballast a few months ago to drive a 40W tube from 12V and didn't even need to heat sink the MOSFETs. The hottest thing was the transformer which maybe dissipated 1.5W of heat. No reason you can't design a 4-tube ballast which dissipates 3 or 4 watts. This would solve the premature failure problem without potting. Except for the ballast, nothing in that fixture gets much above 35°C.
Yes, heat is the enemy of electronics.
/ / /
I hate cutting corners.
[/ QUOTE ]
hehe, Jtr '6-degrees of separation', you need very sharp wit and outstanding memory to keep up with all of my vague OT references. Cameron Richardson, that commercial...um, imagine reading FS (stud muffin of 'hours, days, even a week' LMAO), and SC's initial thread post; then watch the video on the link I provided and remember the
lyrics of the Foghat classic rock song 'Slow ride'....it might come to you, the humor/irony of how both might have an embarrassment in that situation /ubbthreads/images/graemlins/grin.gif. Also related to Prof. Wiz (do search on his usrnm, exactly spelling, with search term 'decadence' or 'classic' to see his response to my other reference to another American TV humorous commercial he can't get in the EU , then it should all make sense.
Little refined sexy/thin things...yep that's how I like them, do search on my one and only post to js's rechargeable SF M6-R (very long thread on CPF) and you'll get the idea and inspiration for the 'JA Special' /ubbthreads/images/graemlins/smile.gif . Or better yet do search on Ginseng's Ultimate EDC thread in Incandescent forum? with search term 'university', lovely tiny little things /ubbthreads/images/graemlins/smile.gif. More fully explained on the 'side-effects' of 'Da Weather' thread shortly on the other forum.
Back OT, I forgot mention another of the less than 1yr old CE 23w CFL was in base down, open fixture indoors and it too failed in the same point at the junction were the tube is affixed to the ballast base. If you look at the link you refer to regularly for full spectrum fluro's you will see in nice small print what you are not supposed to put the CFL's in. Yet CE has no such warnings on the packaging for the one's you buy at Home Depot. With the higher wattage CFL, you'll notice they all have the ballast with slots in the plastic enclosure for necessary cooling.
http://www.fullspectrumsolutions.com/compact_fluorescent.shtml
"
These lamps can not be used with dimmers, in totally enclosed or recessed fixtures.
Because these lamps operate 60% cooler you can use them in any open fixture that the lamp will fit." Base down operation should be an added exclusion to all of these poorly designed, hot running CFL, IMHO.
Since the electronics of dimmer switches in various inexpensive desktop lamps will fail after a certain number of hours of being exposed to the high heat of 100w incandescent bulbs (rated for 150w, hah!), I am wondering which would fail 1st. A dimmer switch itself, or one of these new CFL that are dimmable? On this site they even have a dimmable 5w CFL that they claim will get 25,000hr LOL. At only 5w, that maybe possible, but also look at just how minute the ballast/base of the bulb is!
https://secure.tcinternet.net/buylighting/shoppingcart/Microbrite.html
Still, I think I will try the 100w equivalent full spectrum, dimmable CFL listed on that site. Although the cost is the highest(so the components are failing at greatly reduced lifespans and you take into account energy expended to produce and dispose of the waste, how much are we really being 'green' anyway?, just because of bean counting, cost competitive influences?), the best energy saving to be had would be with a IR sensor, and the dimmable CFL's; so that the light is only on when people enter a room, assuming you could put up with the longish start-up to full-output of these CFL's.
Yet, I think it matters not about which manufacturer or even what type of ballast. For the neighbor down the street; since the time I installed that American made metal enclosure, 100w HPS, with larger, less super inexpensive, older style 3-pin locking photo-eye; had an all plastic LOA 70w HPS flicker and then die, to be replaced by LOA 67w 'Fluorex' tech, work for just 3yrs?, then it died recently too, and they put up another LOA? plastic 70w HPS to replace that, LMAO! Some people have lots of money to spend, and not get anywhere near the useful lifespan of these energy efficient bulbs, HID or fluoros. Its simply a matter of
false/deceitful advertsing (I'll bet the actual engineers are aware of the limitations), and using components that can only stand the heat if put in an open fixture burning with base up...and even at that, if you put your finger on the ballast by the tube, they are hot. Either they need more ventilation, or the parts need to be able to withstand higher heat levels.
From the looks of that ballast (I should mention the 100w HPS outdoor area light I installed was not expensive, but not super inexpensive; and certainly not a costly 'commercial' grade fixture/light) in your picture, I see the one transformer, that is probably supplying voltage in the 4 tube arrangement with the label on it, the square 'frame' rails look like there is a burning mark at the 1 'o'clock to 5 'o'clock position??? I'm wondering (not knowing how ballast designs work) if the other 2 transformers are only used when you wire the ballast for the other tube configurations?
Enough play time on this thread, the natives are getting ancy for my thread on cleaning contacts in the M*glites (not too many CPF members has the photographic accumen or equipment to get nice 1/2-way decent/sharp macro close-up shots), with some even doubting the effect of small amounts of resistence(in their minds) to maximum current flows with smaller, less than ideal wiring/contact points...hehe, uninformed non-electrical engineering types /ubbthreads/images/graemlins/smile.gif
BTW jtr, I'm surprised you have two 3-way CFL's in the main room. Those dinosaur/ badly designed, crappy 3-way sockets are a pain to maintain clean contacts for incandescent bulbs, and make CLF's almost useless for 3 levels of output, I hate them.