New control chip for compact fluorescent lamps

John Popelish wrote in news:go- dncxIqadHusPbnZ2dnUVZ_s snipped-for-privacy@comcast.com:

Awful busy for a compact fluorescent lamp; looks more like a conventional long-tube lamp. The CFLs I've taken apart usually have a smaller parts count, except for the largest models.

--Damon

"John Popelish" schreef in bericht news:go-dncxIqadHusPbnZ2dnUVZ_s snipped-for-privacy@comcast.com...

Nice chip indeed, though it still requires a lot of external components compared to the current discrate designs. The latter lack PFC but nevertheless.

petrus bitbyter

Way too complicated with a way too high parts count including three FETS and transformer/inductor for a single 40 or 80 Watt tube. An integrated solution has to reduce the parts count plus eliminate expensive parts to be viable. Secondly it has to operate two or more tubes in a typical lighting assembly.

I don't understand why more parts and more complexity can make a solution ugly. My designs tend to grow in size to increase performance and reliability. D from BC

So do mine but then it's all jelly-bean parts. Often to avoid an expensive chip or a single source situation for my clients. Transistors for 1.5c, BAV99 diodes for 1c, resistors for 0.5c, opamps for 10c a four-pack and so on. This ballast chip is over a buck! Not a chance IMHO unless they sharpen their pencils again.

Also, what rode them to call the chip IRS? :-)))

--
Regards, Joerg

http://www.analogconsultants.com

IRF has unloaded their MOSFET line. This chip sure isn't going to be the new direction of the company. There is no good reason the chip needs so many added parts to do its job. A couple of logic lines could tell it what tube is connected.

I'm guessing that I.R. have an inside line on a future government regulation that all CFLs will have to have a power factor higher than 0.9 that will make all the simple designs obsolete.

Your and MooseFET's idea of "simple" may not quite fit with us pros' definition where "working" means "properly, as specified, for a period longer than the warranty".

And parts count and simplicity do not correlate.

...Jim Thompson

--
|  James E.Thompson, P.E.                           |    mens     |
|  Analog Innovations, Inc.                         |     et      |
|  Analog/Mixed-Signal ASIC\'s and Discrete Systems  |    manus    |
|  Phoenix, Arizona            Voice:(480)460-2350  |             |
|  E-mail Address at Website     Fax:(480)460-2142  |  Brass Rat  |
|       http://www.analog-innovations.com           |    1962     |
             
         America: Land of the Free, Because of the Brave

All I'm saying is that if the specified operation changes, the designs in production, now might not meet those specifications. This controller is pretty obviously not intended to compete with the discrete designs now in production, but to replace them when the requirements have been redefined. I think this controller anticipates law.

"John Popelish"

** BOLLOCKS.

Nothing in the data says the IC is particularly intended for CFLs at all.

Cleary it is meant for commercial lighting, where long tube fluoros are used.

In that application, PFC has long been one of the essential requirements.

You are barking up the wrong tree with no paddle here.

....... Phil

I see.

Oh well, it isn't the first time I have been out on the shitty end of a limb without a clue.

of

The transformer is almost always needed, and something to switch the current (typ 2 for the tube, one for the PFC), so that leaves a bunch of cheap passives (at a cost as low as 0.1 cent each + mounting) and a controller IC (which is necessary if doing anything other then abusively making the tube light up). This design is cheap in production, and very cheap considering the potential performance. It is not designed as a low cost CFL driver solution - note the wattage rating.

If I made ballasts for larger florescent tubes (eg 4 foot T12's, T8's etc), I would have this design into consideration.

Something to consider - Industrial and commercial power is paid by the kWh used, and by the demand, usually in kVA. The kVA demand meter is reset on a monthly basis (it typically takes 15 minutes to register 90% of a kVA for a

1 kVA load), and each kVA is charged around $5 to $8 (sometimes more) per kVA drawn, per month.

Now say you have 10 kW worth of lighting (a mere 60 to 70 x 4 foot 4 tube ceiling lights). Perfect PF gives a kVA load of 10 kVA (neglecting a small loss from electronic ballasts), which with active PFC is almost achievable. Now say you have an older/cheaper ballast type that draws 12 kVA per 10 kW of light output. that's 2 extra kVA's per month, totaling $120 to $192 per year, not including the likely lower efficiency losses resulting in higher kWh used. Assuming a cheap ballast is $1 or $2 cheaper (which is realistic), I'd make my initial investment back in buying the better ballasts in the first place in under a year. A pretty good return in my books. Another factor is tube life - a well controlled ballast keeps the tube going much longer then a cheap ballast, as it abuses the tube less. Add the cost of labor to change a bunch of overhead tubes with lots of stuff on the floor, and it can be demonstrated that the tube cost is not that significant anymore. Don't forget the disposal fees from the hazardous material. Cheaping out rarely wins.

"petrus bitbyter" wrote in news:465e8e76$0$12148$ snipped-for-privacy@dreader17.news.xs4all.nl:

The bulbs could cost the same as or less than current CF bulbs, given time. They could be priced similar to Philips Iron type bi-pin replaceable mini-flourescents.

Not to mention a socket and pins on the bulb.

But you need to buy the electronics once and throw them out once. with current electronic CF bulbs, you buy new electronics each time, and throw out the electronics when the bulb is gone.