I am attaching an image of thhe cap that came with my M250R2.

I am attaching an image of thhe cap that came with my M250R2.

If anyone can see the pic I posted, is that a good cap to be running on a 250W HPS luminaire at 110-120V?  The luminaire I ordered and have is capable of multi-volt (has a multi-volt ballast) operations (120x208x240x277), but I have it wired for, and use it for 120 volts.  The cap is mounted directly behind the heat reflector.

Is the ballast cwa or HX HPF? It should be fine, normally HX hpf ballasts have the cap hooked to the highest voltage tap (277 here). It should be fine where it is.
And no matter what, the output to the lamp will always be 170-200 volts no matter what tap you supply to the ballast, so the cap wouldn't have to worry about that sort of thing.

It's CWA

The cap would be seeing that 170-200 volts, so I think it would be ok.

Only problem with Any CWA ballast is when they start loosing capacitance the lamp gets dimmer is indication
Cap is failing, problem is on CWA ballast that a shorted
Cap is a quickest way to kill  a CWA ballast cause it used to help regulate the current to the lamp.

Now GE mexican made caps are not that great.
As i replace many in A/C units motor run caps
Also it just as bad lighting.

Caps from china is a Hit and miss

I find Amrad caps are best becase they are 370-440 rated and perfect for lighting.  They make a metal can cap called Turbo mini make up any cap value from 2.5 all the way to 12.5  that i use in my thorm popular pack.
So these amrad motor run cap are USA made and can be ordered from home depot web site, just make sure u use the correct capatance like Medved said.

Obviously be careful to use correct capacitance - if you do not compensate the power factor correctly, the ballast primary will be overheating, but I guess that you already checked...

Sorry to resurrect an old thread, but I tried operating a M98 70W MH HX-HPF ballast without a capacitor and it smelled of overheated components in the ballast box. I measured the temperature of the plates with an infrared thermometer and got a reading of 95 ºC, which is too much, I believe. Should the installation of a power factor correction capacitor reduce the operating temperature of the ballast ? I might just try it to see...

Sorry to resurrect an old thread, but I tried operating a M98 70W MH HX-HPF ballast without a capacitor and it smelled of overheated components in the ballast box. I measured the temperature of the plates with an infrared thermometer and got a reading of 95 ºC, which is too much, I believe. Should the installation of a power factor correction capacitor reduce the operating temperature of the ballast ? I might just try it to see...

This is valid for HX ballasts featuring multiple voltage taps, with the capacitor to be connected across the all primary (240V or so)
The reason for the higher losses (so heat) comes from the fact the reactive power load (about 240VA for a 70W ballast) without the capacitor has to be handled by just the 120V section of the primary, which causes rather high current there (2A). When the capacitor is connected across the 24V winding (of the ballast is fed from the 240V), the reactive load is distributed over the whole 240V primary winding, so leads to just 1A. So even when it flows through double length of wire (so double resistance), the fact the power is proportional to the square of the current means the consequent losses are about half when capacitor (or a feed) connected across the 240V terminals, compare to the setup where all the reactive power has to come just from the 120V.

If a ballast has its capacitor connected directly across the same terminals where the power feed is connected, the capacitor does not influence the ballast losses, so may be omitted when you can afford the reactive mains loading.

That’s what I thought although I couldn’t explain it as elegantly as you. Thanks.

I guess that also explains why the capacitor was loose in the box, as it is not necessary with a 240V use.

I have installed the capacitor and the temperature seems to have dropped to about 77 degrees Celcius. The ballast is remote mounted and has no heat from the lamp.

Most of the M-1000, OV-50, Unistyle 1000, & other 700-1,000 watt fixtures had CWA ballasts with capacitors that fail to the “short” condition causing the ballast to draw double/ triple the normal operating current. When this failure happened, the ballast would self destruct from the heat  generated by the immense current draw which would destroy the windings. The heat from 700-1,000 watt bulbs rapidly contributed to the capacitors failing rather early on in the lamps service life. Because of these factors; original ballasts for 700-1,000W luminaires are extremely rare to find in already hard to find luminaires.

Most of the M-1000, OV-50, Unistyle 1000, & other 700-1,000 watt fixtures had CWA ballasts with capacitors that fail to the “short” condition causing the ballast to draw double/ triple the normal operating current. When this failure happened, the ballast would self destruct from the heat  generated by the immense current draw which would destroy the windings. The heat from 700-1,000 watt bulbs rapidly contributed to the capacitors failing rather early on in the lamps service life. Because of these factors; original ballasts for 700-1,000W luminaires are extremely rare to find in already hard to find luminaires.

The question was, where the capacitors were located in the fixture, mainly with respect of how the heat is spreading from the ballast or lamp. It can make hige difference in component lifetime.
But it could be there was not much freedom (e.g. if the lamp is supposed to be below the ballast in a high bay,...)

The "selfdestruction" could be caused by the magnetic part being the first thing to fail: If it is let to corrode, the isolation between the sheets in the core fails, increasing the core losses. These higher losses heat up the ballast box more, exposing the capacitor to the higher temperature. Then the capacitor fails, then just "finishing off" the coils.
To me it looks more like the fixture was already at the end of life: When the seal was not able to protect the internals anymore and when the corrosion had progressed, the fixture had to be replaced completely instead of just the lamp. Or be content with it failing completely with the new bulb (because the failed bulb may have failed not only from its age, but got overstressed by the ballast already failing)