Fluoro ballasts as audio inductors?

So I was rewiring a stylish old fluoro fitting today and while staring vaguely at the insides for a bit the thought crossed my mind that the old-style 'magnetic' ballasts are just big inductors. I pulled one out and it's measuring at ~40R DC and 1H. I hooked it into a loopback on my audio testing system and it acted like an inductor (surprise!) with no significant hum pickup.

  A quick search in the internets showed a few discussions on using these as chokes in HT rails, with lots of debate on the implications of using them with DC. I couldn't find any references to using them in an EQ type situation as an audio inductor.

  Has anybody tried this? Any issues anyone can think of? Good cheap source of big inductors for 'passive' EQs? I'll get around to hooking it into a proper circuit sometime in the next couple of weeks.

(This is an Australian ballast, made for 240V. I think I've read that 110V ballasts are different, requiring a bit of a step up in voltage to spark the arc, so maybe they're a transformer?)

That is very interesting. They will be made in huge numbers so the cost will be about as low as it gets. 1H is a very handy value for use in a passive EQ. Definitely worth further investigation.

Cheers

Ian

It seems Tridonic is still making a wide range of new magnetic ballasts, and the 'high efficiency' Class B1 and B2 should have significantly lower DCR, although there's no inductance or resistance specs I can find on their site. http://www.tridonic.com/fr/download/data_sheets/Magnetic_Chokes_2012_EN.pdf
  It does go on about their tight manufacturing tolerances and very low magnetic field leakage. And the chokes for different wattages will have different inductances.

Although it would be a shame to buy new ones when there's so many millions of them out there already, with many being replaced by new LED fittings. Shouldn't be difficult to scrounge a few.

I just remembered I had a couple of the 5003 EQs with 1H inductors here, based on an API circuit. I hooked one of the ballasts in place of the 1H inductor and it looked the same curves on the PC analyser. I'll try having a listen tomorrow if I get time.

Tridonic seem to be available here in the UK:

http://www.bltdirect.com/tridonic-switch-start-choke-ec-2x18-oc502k?adcid=pla&gclid=CJPHk6nxpdECFcOVGwod620JuQ#tab-specifications-block

I think I will buy one to play with.

I also see lots of electronic ballasts (whatever they are ) from China for next to nothing. Don't know if they would contain any useful inductors.

Cheers

Ian

I doubt whether the electronic ballasts would have anything useful, they're essentially little high freq switchmode supplies.

If you're going to buy a magnetic ballast new make sure it's a 'dumb' one. Most are, but some have 'features' built in these days.

I measured the parameters of a few available in Australia, for power supply choke applications (for which they are excellent for valve amp use). 

I only measured the first resonance peak of one of them, but the winding and construction method of all of them was similar, so I would expect a not too dissimilar response for 240VAC nominal application parts.  Note that the inductance of lower tube wattage rated chokes, such as for 9W tubes, would be higher.  They are certainly compact, and I would suggest very well designed, given their mass manufacture.

ATCO EC40 (40W lamp ballast): 1.2H @ 104mAdc: 1.1H @ 300mAdc; 0.96H @ 500mAdc;  DCR=40
ATCO EC18/20: 1.88H @ 147mAdc: 1.69H @ 260mAdc: 1.45H @ 420mAdc; 1.26H @ 517mAdc; DCR=54 ; Fres=10kHz
VS LN 18.162: 1.66H @ 165mAdc: 1.57H @ 288mAdc: 1.22H @ 458mAdc; 0.49H @ 740mAdc;  DCR=47
VS LN 18/20.129: 1.69H @ 145mAdc: 1.61H @ 254mAdc: 1.42H @ 407mAdc; 1.2H @ 509mAdc;  DCR=47
VS LN 36.505: 1.08H @ 295mAdc: 0.73H @ 517mAdc; 0.3H @ 825mAdc;  DCR=21

Most passive EQs have no dc flowing so it would be usefull to know the inductance with no dc.

A self resonance of 10KHz is rather low and may preclude them from audio work.

I have a couple on order so I will test them and post the results here.

Cheers

Ian

NOON said:

So I was rewiring a stylish old fluoro fitting today and while staring vaguely at the insides for a bit

Click to expand...

.. and a light came on in your head?

Nick Froome

The inductance can be somewhat extrapolated to 0mA - as I recall, they did not exhibit any swinging choke characteristic, and a cross check using just an AC-only signal generator and sense resistor did not show an unexpected inductance measurement compared to an extrapolation of DC biased values.

The single winding appears to be a very tightly wound simple bunched format, but I have only done a simple tear-down, and not carefully checked the winding.  The core is made from two C section lamination rows - nominally arranged face to face - but with a square lamination central core acting as psuedo former and main flux path.  It is certainly uniformly gapped from the two butt faces where each C lamination fits to the square central leg.

So I was rewiring a stylish old fluoro fitting today and while staring vaguely at the insides for a bit

.. and a light came on in your head?

Nick Froome

Click to expand...

Your wit is brilliant and Illuminating. Fluoressed with aplomb.

I've had a bit more of a chance to play around with them between other jobs.. I have an old (possibly 60's) Ferguson (defunct Australian transformer manufacturer) and a much newer (80s-90s?) Atco (now Tridonic) and both were measuring very similarly for DCR, inductance and resonance, as you'd expect for 40W ballasts. Compared to some other misc inductors I tested they seemed to roll off very gradually from around 10k to the self resonance point somewhere around 16-17k.
  One of the other inductors I tested was an unusual beastie from the oscillators in an old valve organ (no idea what it was, I just got given a few parts) it's a 3 wire device, with one side around 150R and 0.5H and the other around 450k and 3H. Self resonance was very sharp and defined at around 7kHz. (I've got a few of these if anyone has good ideas on what to do with them. Might start another thread.)

Will hopefully put some pictures of the victims up at some stage later today.

You will need to measure inductance change with DC current to determine usefulness as a power supply choke.  Link shows a method for that test, plus info on probably same/similar organ tuning inductors (which I have used a few for screen/preamp filtering in valve amps).

http://dalmura.com.au/projects/Choke%20measurement.pdf

I personally have no interest in them as power supply chokes, my interest is in using them in EQ circuits, but the link posted is interesting anyway. The organ inductors I have are very similar to the ones shown, with adjustable gap.

Close up of ATCO ballast

Close up of end of Ferguson ballast, showing core inside shield

I was tidying up the other day when I came across an unfinished project I started a few years ago - I was building a set of KinoFlo style fluorescent video lights. It'd make more sense to build LED panels now, and I have a drawer with a bunch of ballasts I bought that were never used, and I was thinking of getting rid of them... now I'm having second thoughts