Fluorescent light

Yes, I've done several. Plenty of people selling HF ballasts on ebay. You can get the pukka 'fluorescent fittings wire' from RS (85 deg rated, for use in the push-fit connectors). Blank off the starter hole if there's any risk of wandering fingers getting zapped.

But with new HF luminaires falling in price it's hardly worth the bother now...

Reply to
Andy Wade
Loading thread data ...

Brian-Gaff a écrit :

If there is a large DC component in the output of the electronic ballast, then you can get some blackening around one end. Reverse the tube end for end and it will disappear.

In a normal ballasted fitting the blackening is due to heat from the heaters, made rapidly worse as the tube ages and fails to start quickly.

Reply to
Harry Bloomfield

The storage is in electrolytic capacitor(s). The usual input circuit is a bridge rectifier, followed by the PFC circuit (essentially a boost converter with the raw rectified AC providing its input), with a capacitor on its output. This gives a steady 300+ V DC output to drive the main HF inverter. Any mains-frequency ripple present will be very small and at 100 Hz.

These ballasts will usually run quite happily on DC, over quite a wide voltage range (check the label or data sheet).

Reply to
Andy Wade

Reply to
Andrew Gabriel

Yes there is - first thing they do is rectify mains to 340VDC, and use a storage capacitor.

Secondly, the high frequency driving circuit does not care about having some ripple on the DC supply, since it's driving the tubes at constant current. If the capacitor significantly fails, you can get 100Hz ripple creeping through - you can't see that, but the light output will be noticably reduced. It's rare though.

Thirdly, you don't get 50Hz flicker which you can with magnetic ballasts (due to the tube partially rectifying near end of life, and from tube ends sometimes too). The tube runs at 20-50kHz instead, which apart from not being visible, actually means the tube won't extinguish as the current passes through the zero point, because that's a shorter duration than the mercury ion decay time (which is of the order 1/10000th of a second).

Reply to
Andrew Gabriel

The clue is in the name - HF ballast. Stands for high frequency. Transformers are smaller and cheaper to make for HF.

Reply to
Dave Plowman (News)

Just so, how many years have you been able to put your fingers into a live lamp holder (some recent ones are switched off when lamp removed )But by no means all.

Reply to
F Murtz

Has anyone got figures for the relative efficiencies of LED and florry, like for like?

And some real world guaranteed figures for life?

Reply to
Dave Plowman (News)

I just replaced the under-cupboard fluorescent lights in our kitchen with Ikea LED lights (OMLOPP). These are 24 volt, and come with a transformer. Surely that is reasonably safe?

My only complaint is that the transformer, which had 10 sockets, only supported 5 units (2 80cm and 3 40cm)

Reply to
Timothy Murphy

Yes, no problem there - Ikea stuff is usually well-engineered (apart from their flexi sink wastes).

The ESC article I cited was specifically concerned with LED replacements for T8 linear flu' tubes, that are not compatible plug-in replacements for standard fittings - clarified by ARW's link.

Reply to
Andy Wade

The only time you need high temp wire is if its passing over a hot iron ballast. As you're removing that for the refit, it's not needed.

NT

Reply to
tabbypurr

ES, BC & related lampholders have always been like that. For whatever reason it does not result in deaths in practice.

NT

Reply to
tabbypurr

Interesting that for what I'd call a normal florry with say a 5ft or so tube, neither LedHut or TLC sell replacement LED tubes for. Wonder why that is?

Reply to
Dave Plowman (News)

HF is the drive to the tube, where as you say inductors are smaller and cheaper.

On the other hand capacitors are expensive. If someone asked me to make a HF ballast, the first thing I would do is eliminate such a component that is both costly and unreliable; and allow the HF to be modulated at

100Hz. Hell, I could even design a unit to have near perfect Power Factor.

After looking at circuits, I do see that most have a ~33uF 400V cap to provide a "smoothed" nominal 325V DC supply. I'm just surprised.

Reply to
Fredxxx

In message , "Dave Plowman (News)" writes

LEDhut do:

Reply to
Chris French

Yep, almost as good as the half century old technology of "Quickstart"(tm) magnetic ballasted lamps in regard to flicker free 'instant start'[1] and much less startup wear and tear on the cathodes compared to the cheap and nasty switch start ballasted fittings that have been the only point of reference for over 90% of the population these past 60 odd years.

I bought a 4 foot slimline electronic fitting from B&Q last year (about

14 quid complete with 36 watt slimline T8 tube) to replace the "40 watt" T12 tubed not quite so slimline fitting in our kitchen which I had upgraded to quickstart over quarter of a century ago[2] by removing the starter switch and wiring in a quickstart transformer that I'd had to strip the casing from in order to squeeze it into the fitting.

The reason for 'splashing the cash' was one of aesthetics due to being forced to have the ceiling repaired due to water ingress from a leaking asphalted flat roof which had required removal of the fitting to facilitate the repair work. This gave SWMBI an excuse to insist on a nice new fitting (and in such insistance, she was elevated to the status of SWMBO).

Before I fitted the new fitting, I took the opportunity to compare the performance of the old and new in regard to light output and energy consumption.

The light output levels were, as far as I could discern (it hadn't occurred to me to use my camera as a light meter on this occasion) were on a par with each other. The power consumption, otoh, was a different matter. The new fitting showed a consumption of 36 watts whilst the old one showed something like 54 to 56 watts.

Since, given enough run time, the ballast inductor got very hot it was obvious were most of the vampire watts (some 14 to 16 of them) were being turned into heat. Mind you, the electronic ballast in the new lamp wasn't above turning some of the power into waste heat, it just wasn't as much.

It's been known for some 40 years or more that using a high frequency current (circa 20KHz) in place of mains frequency current (50 or 60 Hz) improves fluorescent lamp efficacy by some 10% or so. This is the main reason why electronic ballasting is usurping the old fashioned inductive ballast in fluorescent lamp fittings. The "Instant Start" just happens to be a useful by-product which, for most home users, is "The Icing On The Cake". For me, the only icing on the cake was the 16 watts or so power saving - I've been taking "Instant Start" for granted these past 3 decades or so.

Such electronically ballasted fluorescent tube fittings have an efficiency rating somewhere between 90 and 100 lumens per watt (the best CFLs, BTW, are only around the 70L/W mark).

For about a year now, it's been possible to buy LED based linear lamps with double that efficiency and better (ultimately reaching 400L/W efficiency for all LED based lamps over the next year or so's time). Unfortunately, you have to pay a hefty premium right now for the 200L/W lamps. The ones still being cleared out of the supply chain are only now approaching the 100L/W mark (typically 81L/W for most of the not quite so outrageously priced samples on the shelves of Aldididle and Asda).

The point of mentioning LED lamps is to warn you that you might be swapping your nice new fitting for a LED based unit in just two or three years time when the pricing on the 200L/W lamps finally drops out of the stratosphere.

[1] The big difference in startup characteristics between the modern lamp and the older one using half a century old technology was that the new lamp wasn't quite as instant in that it lit up immediately from stone cold but with about a half second's worth of half brightness flicker. The old lamp was fully lit up in something like a quarter of a second or less from switch on.

Since I was already used to the idea of 'instant start fluorescent lamps', this seemed a less than stellar performance. Not a major deficiency to be sure but a deficiency nevertheless in my mind. For anyone brought up with switch start fluorescent lamps, this would be considered a marvel of technical development.

Otoh, it could just be that I happened buy the one luminary in the stack that wasn't quite up to snuff (either the tube or the ballast - I don't have a spare T8 tube to test with). I wonder if I'm the only one to witness this startup behaviour?

[2] I can't be absolutely certain but I believe I'd only had to re-tube the fitting twice in all of those 25 years plus.
Reply to
Johnny B Good

Yes, Brian but it's very rarely economic to do so since the electronic ballast kits are typically more expensive than a whole new electronically ballasted fitting complete with tube. :-(

Reply to
Johnny B Good

If there's room in the fitting (non- slimline fittings have plenty of room) you can simply wire a quickstart transformer into the circuit (removing the starter switch) but you'll be limited to the older type of T12 tube which, I understand, are no longer readily available - ok if you happen to have a stock of spare tubes to use up over the next half century or so. :-)

Quite frankly, if I was in the market now to replace a linear fluorescent lamp fitting, I'd, at the very least, be considering LED versions before splashing the cash on a "Modern" electronically ballasted fluorescent luminary.

To my mind, the kitchen is the one place in the home where a fluorescent fitting is the ideal luminary to use as a primary light source. If such linear luminaries didn't already exist, they'd have to invent a LED based equivalent.

Reply to
Johnny B Good

The delay improves tube life.

Maybe you missed out on thermal starters, that came before glowstarters. They behaved impeccably, and started every time with no flicker. Glowstarters took over because thermal starters wasted 1 watt, and watts aren't free.

Manual switching also doesn't flicker, but isn't very other-people compatible. (FWIW I also remember using a cast iron cased mercury lamp ballast with a manual switch to strike it.)

Modern electronics makes all that stuff obsolete now.

NT

Reply to
tabbypurr

You can always try a google or bing or duckduckgo search for such info. A key phrase worth trying is "Cree announces 301 lumen per watt lamp!", an announcement that's now some 16 months old. You should also come across a similar announcement by Philips Lighting, made around the same time.

Damn! I couldn't resist following my own advice! :-)

Here's a link to that Cree announcement... enjoy!

Apologies for the line wrap.

I've posted such links a few times over the past year. This time ICBA to use up my google fu yet again. In essence, we aught to be seeing 200 lumens per watt lamps appear on the supplier's shelves sometime over the next six months or so.

Both Cree and Philips reckon it takes about 18 months or so to go from 'Lab Demonstration Sample' to actual product on the shelves. Philips were certainly offering 200L/W luminaries in the fluorescent linear lamp format about 18 months back which is a doubling up of the best efficiency of a modern fluorescent luminary (circa 90 to 100L/W).

For reference, the best CFL efficiencies are only around the 70L/W mark and the current LED plug/screw in GLS lamps are around the 81L/W mark (hopefully improving to double or better than that over the next six months or so, depending on how swiftly they can clear the old stock out of the retail distribution channel).

The longer you can postpone the upgrade to LED lighting, the better the ROI is likely to be. If you've already 'upgraded' to CFL, I'd be inclined to only replace them as and when they fail or your stock of spares runs out, being very choosy as to which lamps you upgrade to maximise the benefit of the premium price you'll undoubtedly find yourself paying.

Disappointingly, a search restricted to just the past month's worth of articles/results doesn't show LED lamp efficiencies higher than 95L/W for a 4000deg K lamp. Indeed, at least two different suppliers maxed out at a rather curious 88.88888L/W figure, including Philips LED tube replacements for fluorescent fittings which were claimed to be double the efficiency of fluorescent lamps which we *do* know are in the 90 to 100L/W ballpark. The figures (45W 4000 lumens tube) and the claim just seem irreconcilable to me.

The fact that I can only see 2700K/3000K lamps with efficiencies barely better than the 81L/W mark of a year back suggests there must be a huge log jam of old product clogging up the distribution channel. Never mind, I've got plenty of CFLs to ride out this hiatus. We'll see who survives this stand off, whose nerve fails first, eh? :-)

Reply to
Johnny B Good

HomeOwnersHub website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.