The cheapest ones I have seen are 85 C. I have never seen 50 C. I would use 105 C in a tube amp.
## IF the ambient air temp surrounding the cap is say 20-25 deg C...AND ripple current is low.... the cap will last a LOOOONG time. Ripple current is just dc plate current x 2.56 per all the cap maker's tech notes. they also have detailed maths on their tech notes for commercial applications. Like motor drives, power supplies etc. Some of the example used are... caps must last at least 14 years... running full bore 24/7. Ripple must be no more than XXX %. Expected ambient temp is XXX degs C. they take it through the various choices... and a lot of times.... to get long life... it's wiser to use a cap with a higher ripple current rating... cause THAT'S what's gonna heat em up inside.
## beware, the 105 deg C rated ones also have a higher internal temp..and they allow for higher spot temps inside the 105 deg C caps. Those caps used inside the compact lights sold these days ? They go dead cuz the 105 deg C cap is cooked in em. The latest version uses a 125 deg C rated cap. The heat inside the base can't get out..esp when plugged into the ceiling..all the heat rises..and is trapped inside the base. They are not vented either.
### Just a bit of air from a small ,low cfm fan in any amp power supply does wonders. So does NOT running them at their max V ratings. For commercial applications, caps are run at 75% of their ratings. For aerospace, it's 50%. Get a cap with a BIG ripple current rating. They will last the longest. A lot of linear amps had caps go bad. They lacked enough bypass caps..and RF will kill em very fast. A 1000 pf bypass cap at base of a plate choke is not good enough. You need 2-3 4700pf caps. u need 2-3 more of em inside the HV supply as well. RF can easily jump past the plate choke..via the tank coils..and end up on the HV lead that feeds the base of the plate choke.
Later... Jim VE7RF