Ban on Incandescent Bulbs

that is rather counter-intuitive. given skin effect in conductors, and eddy current & hysteresis losses in the iron, I would expect the exact opposite. Of course I have never run 50/60Hz magnetics at 500-600Hz, but I have stuck 2kHz current (and not much of it, e.g. 50A 5% line choke with 500mA @ 2kHz) thru 50Hz chokes, and boy do they smoke!

still, if you keep the volts the same and crank up F, B goes down....so I guess what you see is the reduction in losses that result from NOT slamming the core well into saturation every half-cycle, which is what most 50/60Hz magnetics do - microwave transformers being the best example of that.

have you taken skin effect into consideration? if you have say 50% 5th harmonic, then the skin depth is about 2x lower, so (assuming skin-depth limited) the losses are about half that of the fundamental, rather than

1/4th

then there is crest factor etc. so without too detailed an analysis, it looks like the net result is roughly status quo.

but with PFC, the total lighting load on the national grid could go down by a factor of three, providing much-needed margin - dunno about USA, but over here we run at (IIRC) about 98% of capacity.

sure it costs more, but not that much more. e.g. one does not have to duplicate all the mechanics. there are also series/parallel tricks (e.g. valley-fill circuits) to improve PF that are VERY cheap. and in silly volumes, you might be surprised how cheap a boost PFC can become - heck, the CFL electronics costs about $1, and a boost PFC is no more complex.

I think if required PFC could be added to CFLs and the resultant price increase would be perhaps $2 - $3 for a $5 CFL (which is what the cheap ones go for over here).

interesting discussion though.

Cheers Terry

Meanwhile it seems that everyone is busy ignoring the poisonous mercury in the CFLs as well as all of the energy and widespread technology (translation: energy) necessary to make the damn things, which do *not* last the original "guarantee" of seven (then revised to

5) years. No more life guarantees, or warrantees on CFLs, but incandescents continue to have a life rating that most meet.

Not significant in even 14 AWG wire at hundreds of Hz.

That transformer core loss is mostly proportional to square of volts-per-turn and fairly independent of frequency.

To first order, proportional to frequency and square of magnetic field - which is proportional to 1/f times square of volts-per-turn.

I have!

Is that a common mode one with an imbalance?

At 2 KHz, I expect current required to achieve same volts-per-turn as 50 amps at 50 Hz to be 1.25 amps, for a first order extrapolation.

Skin depth at 300 Hz is a goodly few millimeters. Resistance of even 2 AWG copper wire (diameter about 6.4 mm) is only about 1% higher at 300 Hz than at DC if I understand correctly the relevant stuff in the "CRC handbook" (with asssumption of linear measurements in cm), page 3324 and thereabouts in the 43rd edition.

If this is relevant to core loss, I don't see conductive core laminations that thick. For that matter, I see for eddy current losses the EMF being the independent variable that current magnitude is dependent on, and losses benefit (decrease) from a resistance increase. Further for that matter, I see that iron transformer cores tend to be made of higher resistivity iron alloys, such as "silicon steel" in order to minimize eddy current losses. (Of course the "skin depth" at 300 Hz is more than a few mm for such high resistivity alloys!)

The crest factor in line current is the same issue as harmonics. Spikiness of the current waveform is from harmonic content in the current waveform.

Can you cite examples? Can you cite a circuit? Can you cite a cost of implementing it less than maybe 30 cents apiece at annual volume of a million units of each model that this gets implemented in? Can you cite alternatively a higher cost that would "get down to earth" from a government mandate?

I think that could be quite interesting! As in, who loses and how if watts consumed decrease by a factor of 3-4 while VA decrease only a little? I only see downturn to those who make money from billable watt-hours and little negative impact to anyone else from VA decreasing less than watts do as long as both decrease.

- Don Klipstein ( snipped-for-privacy@misty.com)

politicians while favoring CFLs)

Replacinmg incandescents with CFLs does mostly reduce mercury pollution by reducing mercury emissions from coal-fueled power plants.

The energy rerquired to make a CFL is definitely generally less than the energy saved by using a CFL in place of incandescents, even if it only lasts 2,000 hours.

It is easy enough for me to buy CFLs for $2 USD apiece. That is now worth about 2.1% of a barrel of petroleum, or about .88 gallon of petroleum. At "Texas crude" ("my words", from "Heat of Combustion of Liquid Fuels", page 1936 in the 43rd edition of the "CRC Handbook") I find

7.286 pounds per gallon, and 19,460 BTU per pound. So far I find about 125,000 BTU per $. Keep in mind that a KWH is about 3414 BTU. This means about 36.6 KWH per $. Also keep in kind that combined generation and transmission efficiency of electricity is mostly close to 1/3! And the loss is mostly in inefficiency of fuel-burning engines, as opposed to transformers and wires. So far this means 12.2 KWH per dollar to generate electricity by burning oil, or 8.2 cents per KWH.

A 15 watt CFL replacing a 60 watt incandescent for even only 3,000 operating hours saves 135 KWH, or close to USD $11 at 8.2 cents per KWH.

A 13-15 watt CFL usually costs a lot less $11 USD even for purchase quantity of 1 from a retail shelf - I can find these easily enough for $4! I do not for even a second give any chance that energy requirement to manufacture and deliver it is going to cost that much!

- Don Klipstein ( snipped-for-privacy@misty.com)

nope. using 50Hz chokes to filter output of VSD.

I never investigated it fully, rather I gave up & designed low-loss inductors. I suspect the square-wave excitation caused the problem, as the harmonics would be thru the roof!

we might be talking at cross-purposes here. I am thinking of power distribution. I havent measured the conductor diameter of overhead lines, but I can see its fairly thick, so 10mm isnt far off. odds on the wire size *is* fairly close to the limit imposed by skin depth.

well a simple boost PFC comprises a choke, a diode, a FET and a controller. I'm pretty sure, in million-off quantities, I could make a

18W PFC for around a buck. I could certainly do a controller for about $0.15....

again, I'm thinking about the distribution authorities. in NZ these are state-owned, so the object of the exercise is (allegedly) NOT to suck as much cash as possible from we the victims. distribution is all about the heat dissipation ability of the transmission lines, which is why a line can carry more power in winter than summer. In which case they care a lot about VA - it uses up a portion of the current-carrying-capacity of the line, making it unavailable for real (billable) power.

and if you are like NZ, running at close to 100% of capacity, with about (ISTR) 30% of ampacity taken up with VA, getting rid of even 10% of the VA would be hugely beneficial - it would DOUBLE the system margin.

When I worked in MA we talked seriously with PG&E about fitting VAR gobblers on the output of disty xfmrs - again, it was around 30% VARs using up xfmr capacity (also thermally limited), and it looked like power electronics was cheaper than up-sizing 30,000 xfmrs.

Cheers Terry

They could be recycled. Just like batteries are *supposed* to be.

Tim

-- Deep Fryer: A very philosophical monk. Website @

In article , Terry Given wrote in part:

(On making electronic ballasted CFLs with better power factor)

You have a point there!

- Don Klipstein ( snipped-for-privacy@misty.com)

Yeah but... The PF of CFLs is leading, while at the moment, most utilities are still trying to compensate lagging power factors (due to transformers, motor loads and the generally inductive nature of the power system itself). So for the near future, the bad power factor is a good thing.

"Ralph Barren of any Idea "

** Utter rubbish !!!!

The PF of electronic devices is due to a having bad ( pulse like ) current waveforms.

CFLs are among the worst.

Scroll down and see figure 11 for a typical example.

FYI: PF = Watts / VA

No phase angle need be involved.

....... Phil

I don't see electronic-ballasted CFLs with low PF having low PF by "leading", but by drawing a spiky current waveform that is rich in harmonic content.

The "non-real amps" are mainly at harmonic frequencies rather than from

90-degrees-out-of-phase-from-voltage component of fundamental frequency portion of the drawn current. That one is indeed leading, but small.

The usual integral-electronic-ballast CFLs with low power factor have fundamental frequency component of the current waveform leading voltage by something like 15-20 degrees or so. Believe me, most of the "non real VA" here is from harmonic content in the current waveform rather than leading/lagging.

- Don Klipstein ( snipped-for-privacy@misty.com)

back when I first started designing motor controllers, our R&D mgr had written an internal white paper, telling the technical literature & sales guys to use Displacement Factor (DF - phase angle of fundamental). Because, of course, with a rectifier-capacitor filter, DF ~ 1. When he left & started up a consulting company, he wrote a white paper pointing out that leading australasian drive manufacturers all cheated by speccing DF rather than PF, in order to sell harmonic filtering solutions. Ballsy, but brilliant.

Cheers Terry