i gust got an old fixture that takes 2 f40t12 bulbs. it has an older universal therm-o-matic ballast with green label, it works and runs tubes fine at full brightness and does not hum or run hot. the only issue is it wont start tubes on its own, when first turned on the ends of the tubes light orange for a second or two then nothing, it i touch one of the tubes they will come on fine, after running after starting it manually for about 30 minutes it will restart until it cools down. any ideas?

Sounds like a grounding issue. I would check to see that the ballast itself is properly grounded, that paint hasn't gotten in the way of the mounting screws, the 3-prong plug is properly grounded (if used) and that a good electrical connection is made from the reflector to the ground wire. If all this is OK, then wash the lamps and the reflector. Dirt can inhibit starting of RS lamps and ballasts.

If that doesn't work, another solution is running a thin wire around the lamp, in a spiral fashion (like the red stripe on a candy cane) and fastening one or both ends of said wire to the reflector mounting screw(s). Be careful that the wire does not slip between the lamp holders and hit the pins as you might damage your ballast.

i already checked, it is an older fixture which originally had preheat, I ran a ground wire to the case if the fixture, and that did not work so i connected one from the wire connected to the case if the fixture and ran it to the bulb, did not seem to help.

Then the question is, how far is the reflector from the tubes. You said, it was originally a preheat fixture, so it is likely its design did not observe the reflector proximity requirement of RS ballasts (it needs to be within 1/2 inch from the lamp, if I remember well).

Otherwise a question: What is the lamp configuration? If they are in series, the culprit could be the faulty starting bleeder (either low value resistor or high value capacitor parallel to one of the lamps). Its function is to bring the complete ballast OCV across single tube to let it start, then the low voltage drop of an ignited tube forces the full OCV across the other tube, so both have nearly the full OCV for ignition. Once the bleeder is faulty, the OCV may distribute more evenly across both lamps, so becoming insufficient for both of them.

I think I fixed it, but not sure yet. i ran a ground wire to the metal inside the fixture on the mounting screw, then to the actual ballast case and then ran one wire to each of the tubes. it started after having it off for about an hour, but i dont know if it will start after being off all day, i will see if it will work when i get home, which will be over 7 hours, so if it is going to have an issue starting i think it would after that long.

also, i believe the tubes are in series, if i remove one the other will glow but very dim, and if i remove the other one, the other will not light at all.

also, i believe the tubes are in series, if i remove one the other will glow but very dim, and if i remove the other one, the other will not light at all.

That may suggest the bleeder working. It is the low current for one of the lamp glowing dimly, when the one parallel to the bleeder is removed...

not 100% sure yet, but i think that fixed it, it started okay after about 6-7 hours of being off. hopefully it will keep working

I don’t see why grounding the ballast itself would make a difference in terms of operation.
Do all the ends of the tubes glow?  Make sure all the connections are clean and tight.
I’ve also seen lamps were the pins weren’t crimped properly.

Some of the older ballasts say ground ballast for proper operation.

Grounding the ballast is required for safety, not for the operation.
The only exception could be in theory isolated secondary ballasts, but Ive heard only about HID ballast made in that form (CWI or Mag-Reg). And HIDs don't use any external electrode (external to the whole lamp assembly, wires along or around arctubes are common with HPS, but these are part of the lamp assembly, being inside of the outer bulb).

Some ballasts may have used the ballast cases as a means to connect the reflector. Such ballast has a dedicated PE wire internally connected to its case, assuming the case mount will ground the reflector via this connection. But dont think this is acceptable now, as the PE connections should be made inspectable (you can not inspect a connection inside the tar filled pot)

hi, i am pretty sure it is working now, it started over night and it was about 50 degrees in my room, so it was pretty cold and it started no problem, so i think it is okay now. i don't know if it was grounding the ballast case that did it, maybe it just made better connection with the fixture there, it does not have a reflector per say, it has a louver and the top is mostly opened, so it could have been to far from the tubes, but it does not seem so, i only ran a wire to one tube. but as long as it is working i am not going to investigate it too much, i am hoping to convert it back to preheat some day anyway.