I would like to know if this is possible.

I know I can get a Payne-Sparkman ultrasonic ignitor to do this same thing, but I don't want to have to do that.

Will temporarily shorting the output of a probe start ballast work in a pinch? Keep in mind probe ballasts have a higher OCV than pulse ballasts.

I had to look up what is this "ultrasonic" ignitor. Read the description on the manufacturer website, and all i can say it describes exactly a standard superimposed ignitor, like what's installed in 90% of ordinary HPS street lighting throughout Europe, but for 3x the price

For a probe start ballast with Voc around 240V (probably will be ok up to 270..300V), Get a Vossloh Schwabe Z400M, Bag Turgi MZN400, or something similar
https://www.ebay.com/itm/316160991645

Momentarily shorting a HID ballast - I expect this be ok for the ballast, as the ballast is made to withstand momentary dips in the arc voltage of a cold lamp. However it may not be effective for starting a lamp. As matter of fact i tried to light an HPS this way, shorting the ballast output 30 times, and have not started the lamp. Maybe i would have better success with a MH lamp, but i have not tested it

Other question - Is the current spec of the ballasts compatible ? I thought your Pulse Start ballasts follow what we call HPS specs (250W lamp 3.0A arc), while Probe Start what we call Mercury specs (250W lamp 2.13A arc) ?

for simple bench testing, get a sun bed/tanning bed starter of the highest wattage rating you can find, wire it in series with 3 120V 100W GLS lamps in series, and stick that whole caboodle in parallel with the main HID lamp your trying to strike

it might not be the most reliable/strongest thing, but its the best yah can do besides getting an actual super-imposed ignitor

you could use an FS4 (or even FS2 starter if you remove it once the lamp has struck) if you need something *now*/cant purchase a Sun bed starter for some reason, but the pulse voltages of an FS4 is not very high at all (I have seen them struggle to even start Krypton T8's) so results may vary




funnily enough I had to do the inverse of this when I had to strike a S55 HPS lamp off 120V mains, 120V Mains ignitors dont exist here, but an S2 starter in series with an incandescent lamp did the trick well enough

https://www.lighting-gallery.net/gallery/displayimage.php?pos=-257030



its an old trick I have been doing for many years, you can see it in the background here as the merc was a bit recalcitrant and so needed a bit of a tickle to get going

https://www.lighting-gallery.net/gallery/displayimage.php?pos=-62353




I first learnt about it through SON-I Glow bottle lamps and this upload also

https://www.lighting-gallery.net/gallery/displayimage.php?pos=-39792

(Ahhh memories)

Other question - Is the current spec of the ballasts compatible ? I thought your Pulse Start ballasts follow what we call HPS specs (250W lamp 3.0A arc), while Probe Start what we call Mercury specs (250W lamp 2.13A arc) ?

US Medium and high wattage ANSI Pulse Start metal halide lamps (175W 250W 400W 1000W) are all Mercury spec (apart from 250W FC2 lamps which are HPS Spec)

Pulse start MH lamps are *generally* rated at 3.5-5kV ignition pulse which is out of reach of glow starters and sparking with a piece of wire. There are some exceptions with lower requirements, like Philips HPI-T at 250-400W and American allstart lamps.

Why not to add a 3-wire superimposed ignitor, I use them all the time...
 

Pulse start MH lamps are *generally* rated at 3.5-5kV ignition pulse which is out of reach of glow starters and sparking with a piece of wire. There are some exceptions with lower requirements, like Philips HPI-T at 250-400W and American allstart lamps.

Why not to add a 3-wire superimposed ignitor, I use them all the time...
 

you would be surprised what you can get away with on a test bench setup, sure as you say officially for totally reliable ignition over the life time of the lamp most Pulse start Metal halide lamps need a 5Kv ignitor

but especially if your dealing with fresh lamps etc, they can sometimes strike in sub-optimal conditions, so if your just a collector messing about on a test bench, nothing to lose by trying




with regards to using a super-imposed ignitor, seeing from Ash's comment that that there is a supply of them in the States, I would normally indeed suggest for the sake a few bucks get that


but I do wonder what the actual voltage rating limit is of a super-imposed ignitor is, because many US Metal halide CWA ballasts will have OCV's in the 350V range (hence why I suggested in my original comment to use 3, 120V GLS lamps in series as a current limiter  for the glow-starter)

If it is for a bench test, the OCV won't be that much of an issue: Once the lamp ignites the voltage becomes its 100..140V regardless of the OCV.
Problem would be only with unsuccessful restrike attempts, where the high voltage for a long time will fry the charging resistor in the ignitor.
But you can modify the ignitor by connecting a resistor in series with the Neutral wire, so reduce the voltage the internal resistor sees during unsuccessful striking attempts, maybe also bypassed that resistor by some 10..100nF capacitor. The question is to determine what resistance to put there, it should be so it keeps the roughly 230V between the Ballast input and Neutral terminals of the ignitor.
Initially you may just measure the Neutral current on the ballast you have (for the few seconds to take the measurement the ignitor will survive the overvoltage) and from that calculate what resistance inside is (usually it is around 5..10kOhm, depends on the output pulse energy) and then calculate what is missing towards the OCV of your bakllast so the ignitor will see the 230V.
The resistor should be of some power type of few W (assume 350V OCV ballast, the ignitor having 10kOhm inside, so you need about 5.6kOhm rated at 5W).

Before looking into this discussion, I have initially had my doubts about using typical European 220-240V superimposed ignitors on North American 175W M57, 250W M58, and 400W M59 probe start metal halide ballasts as I have been thinking that the 300V OCV of those ballasts would have likely been too much for these ignitors to handle. As a result, whenever I have been playing around with European 220-240V superimposed ignitors, I usually used them to run North American pulse start metal halide lamps on mercury vapor ballasts instead as the mercury vapor ballasts tended to have an OCV that closely matched the 220-240V rating of the superimposed ignitors.

In a few of my LG posts, I have in fact shown a few examples of European 220-240V superimposed ignitors being used to run a North American 400W M135/M155 pulse start metal halide lamp on North American 400W H33 mercury vapor ballasts.