Note: I am in the USA, so 120V 60Hz. Also this post is really long so, sorry in advance, but if y'all can help that would be super great!

---

PREFACE:

Normally if you are using an inductive choke ballast for any type of electric discharge lighting, the supply voltage has to be (approximately) at least twice the running voltage of the lamp. But, if you use the right capacitor in series with the right choke, this requirement is somehow bypassed(?). I hear this is called semi-resonant. I am sure many of you have heard of this, like in those cheap Lights Of America shop lights.

---

THE EXPERIMENT:

I recently powered an F40T12 rapid start lamp with a .2H choke and 2.5uF capacitor in a standard preheat configuration with a manual starter switch. The lamp lit beautifully (with some initial striking problems), even though the arc voltage of one of these is supposed to be 100V (way more than half of 120V). The circuit drew .455A, which would suggest that the lamp was overdriven, but the lamp was still quite dim (we'll get to that later). I posted a picture of this lamp lit connected to this circuit in my gallery:
https://www.lighting-gallery.net/gallery/displayimage.php?album=8574&pos=0&pid=255553

The following circuit was made (excluding the starting preheat circuit):

Hot--(A)--{.2H CHOKE}--(B)--{2.5uF CAP}--(C)--{F40T12 TUBE}--(D)--Neutral

I also made a to-scale paper schematic drawing with physical distances representing voltage differences (also with voltage labels):

---

QUESTIONS:

My questions are as follows:

1: Cap and choke voltage
How is the voltage across the inductor and capacitor so gosh darn high if the supply voltage isn't? I know the phase angle of them makes it all add up to 120V in the end but what is making it do that in the first place? Nothing about this makes sense.

2: Tube voltage:
How is the voltage across the tube only 37 volts? This calculates to around 17W, which explains the dimness of the tube. How is the tube being overdriven by a few milliamps but still managing to output less wattage?

3: HIDs?
Could this concept be used with HID lighting? I am thinking of trying it out but I am not sure if it is a good idea.

---

Thanks for reading this whole darn thing, I know this is a lot to ask, but maybe someone can help me out. I tried as best as I could to make this easy to understand, but let me know if you need any clarifications. I think this is an interesting thing for us to investigate, especially for the people like me who live in 120V countries. Thanks!

I bet that rough explanation of how these circuits work is while this kind of ballast has low impedance at fundamental 60Hz, inductance value is high enough to store energy available to fast transients like discharge re-ignition. In harmonic way of thinking, ballast impedance is low and capacitive at 1st harmonic, but high and inductive at higher harmonics.

Next, you are doing a few things wrong -

You are probably getting strongly distorted voltage and current lamp waveforms in your circuit. So you can not measure them properly with a regular multimeter, you need true RMS one. More, the product of multiplication of even true RMS voltage to true RMS current will not give you a proper power value for signals of non-sinewave waveforms and also unknown phase relationship.

That's the area where oscilloscopes belong to.

As for why you are getting high voltage across the capacitor and inductance, that is the way series resonant circuit works. You still have (not accounting for actual distorted non-sinewave current!!) U=I*impedance (of an inductor or capacitor) but the voltage across the whole circuit is low as these voltages are about 180 degrees out of phase and largely cancel each other.

Have a good read here: https://en.wikipedia.org/wiki/LC_tank

Darn, I was thinking maybe it was a meter issue. I technically have a true RMS meter, but I recently blew it up trying to measure the output of a vacuum tube transformer lol.

But could this kind of ballast work for HID lighting?

If you look at CWA ballasts, they have a capacitor wired in series with the lamp and they seem to function in a similar manner from what I understand.

Yes, used with some modern SOX circuitry.

Can you explain me how you get to your pointer image? I am not sure if fully understand it, have a thought error ore it being not correct...

As far as I understand, many collectors often say that low pressure sodium is NOT HID.

Actually I have to think about some SOX circuitry being classified as SRS i'll comment again soon.

Regarding SOX jot being a high intensity discharge lamps, i challenge everybody who disagree with that statement to explain their point of view. I agree that by some classification they are a grey area, however I think they should be classified as HID lamps as these lamps have a significant light output, use a positive discharge of metal vapour and are traditionally used for outdoor lighting.

EDIT I dit some thinking and the SOX circuits I thought of could be better qualified as CWA Circuit then SRS SRS and CWA are however different. With CWA the capacitor is part of the current limiting circuit of the lamp. With SRS At lest the one i am familiar with mainly beeing used in the UK, it is used to ignite the lamp.

Now i am seeing what you described. Your circuit is a series compensated choke ballast (Lead ballast) which functions more like a CWA ballast then semi resonant start SRS.

Here in the US sox lamp use high voltage anput 900 OCV
So there instant start ballast.   Cap  just for oower factor correction

  U can take any two lamp electronic F32 Instant start with both output combine from electronic to run a 35/ 55 watt sox lamps no problem  it has high ocv  too.

   By the I go look at my 60 watt GEC SLI/H a pair of them runing on  magnetic F96T12 HO rapid start ballast on LG under funkybulb.   

 I can understand why u guys like to complicate things when there already a suitable replacement ballast for 35 / 55 watt sox lamos,  our advance F32T8 instant start gear will work on 100 to 277 volt 50 Hz or 60 Hz problem solved

Can you explain me how you get to your pointer image? I am not sure if fully understand it, have a thought error ore it being not correct...

I just measured the voltages from various points on the circuit and used a compass and a ruler to draw it out, but now that I know that these are not true RMS readings, the drawing I posted is almost certainly incorrect, so essentially irrelevant.

For the drawing to be correct you need to read out the 1'st harmonic, not the rms.
To be exact you would need to do the drawing for all harmonics and then "assemble" back the various true rms values.
If the meter uses "average of absolute" type of rectifier (practically all non-true-rms meters), the reading is ironically closer to the 1'st harmonic than the true-rms reading.

So I don't think it is possible, without special oscilloscope setups, to get anything better.