I got some more questions... I am going to preface this by listing facts that I am fairly sure are true before listing my questions:

---

- HPS lamps, when used on MH-style CWA gear, can runaway and limit their lifespan.
- HPS ballasts exist that are CWA, and they generally use physically larger magnetic shunts and larger value capacitors than similar MV/MH ballasts. This could be because of the higher currents used in HPS lamps, or it could mean something else.
- MH/MV CWA ballasts almost certainly operate in the lower left corner of this grossly idealized graph

---

1) Accuracy:
Am I right with these statements?

2) HPS CWA Characteristics:
I don't have an HPS CWA ballast so I can't test it. If someone could run the ballast on low input voltage (12-24 VAC) and compare the short circuit current of the ballast with the capacitor connected and the capacitor bypassed, we could figure out generally what quadrant of the idealized graph they operate in and see if this differs any from MV/MH ballasts.

3) Actual Severity:
We say that running HPS lamps on CWA is real bad, but what exactly happens? Does it warm up and then runaway and start cycling? Or will it just make the lamp not last as long as normal?

---

Thanks!
 

So, I don’t know enough about the lamp part of your question, but for the Ballast part, maybe I could help soon. I have a 250 watt Hps CWA ballast with a cap. Maybe I could measure the voltages from the ballast.

@stillaintjeff24
Thanks for offering! It seems to be very difficult to get people to do this kind of thing.

Anyway, if you have a low voltage transformer like a thermostat or doorbell transformer, you can use that (or if you happen to have a variac). Just use the same input voltage for both tests, preferably 24V or below. Using any cheap multimeter should work just fine since there won't be a lamp connected.

Test 1:
Connect the ballast as usual, but where there would normally be a lamp connected, just short the two wires. Measuring the current with a probed ammeter or a clamp meter, record the value when it is running.

Test 2:
Same setup as before, except this time the capacitor will also be bypassed. Record this current as well.

The difference between these two currents will hopefully be able to answer our questions. With MH ballasts, the capacitorless current is much much higher than the current with the capacitor connected, we shall see how this compares to HPS ballasts. Let me know if you need a drawing of how to set this up, you know how I love schematics lol.

Hey, I haven’t forgotten about this, just couldn’t get my setup going. I will have the results for you in a bit. Sorry for the wait!

@stillaintjeff24
No problem, I am certainly in no rush lol

Yeah, sorry for the panicked message, the transformer scared me, I thought I blew my light up lol

You scared me there for a bit, I thought I was the cause of a destroyed fixture lol! Though even at full voltage the ballast should be able to handle a short for several minutes, never mind at 16V.

Technically speaking the ignitor shouldn't have to be removed for the test, but you can remove it if you want. The voltage shouldn't be high enough to trigger it, but who knows.

Sounds good, I don’t know to much, I didn’t remove the ignitor, I had just thought the fixture broke because that transformer failed. All good, I might have the results in by the end of today.

Alright! Results are in and interesting. So this is on a 250 watt Hps CWA ballast. Everything connected normally, except shorted lamp socket. With a 16v Ac input, the amps are:

With Cap: 0.55

Without Cap 0.65

Interesting, not sure if this is helpful, but that’s the result I got. When I bypassed the capacitor, I’d say the result was shocking! Shouldn’t of held onto the wires to short it… Hope this helps!

@stillaintjeff24
Thanks so much for doing this! Those results are indeed interesting. So the capacitor's impedance is less than the leakage inductance. Very interesting to see.

Hope this helps! Thanks for letting me do this experiment!

So what this means is that HPS CWAs and MH/MV CWAs both operate in the same quadrant of my oversimplified graph. This does make sense, but if they both operate the same way, then how are they different? Because they are almost certainly different. Maybe they are only slightly different and we would need to take more complex measurements? Idk, maybe I will figure it out when I get my hands on one of them.

Hope my readings were accurate. These were readings were from a relatively accurate clamp meter. If you ever get one I wonder if the wattage affects the readings.