Actually you can't...since you have now blown up TWO of em. The oem setup used 5 x 75 pf caps = 375 pf.
I would remove ALL the caps, and replace em with 4 x 100 pf @ 30 kv types. Dirt cheap, and you won't blow em up anytime soon.
One might survive with one less if not driving the amp so hard? I understand that the failure of these caps is not uncommon in some Alpha amps - I wonder if the failure is due to exceeding the voltage rating or excessive current? Since my last post, my new ceramic disc caps arrived in the mail and I installed them. They were 6KV, 75 pF. During testing into a dummy load at full power, one of them failed. So, your suggestion is probably a good one but are you suggesting I use ceramic disc caps or doorknobs? Doorknobs certainly aren't cheap - $20 each and up. If the problem is that the current capacity is being exceeded then the 30 KV rating won't be of much value. Can you suggest a specific capacitor to try? How can I determine the current flow through the capacitors as padders to the pi net input capacitor? I think I can find room for four small doorknob caps. The best recommendation I could find for voltage rating of the pi net input capacitor(s) was twice the plate voltage. That matches the 6 KV rating that Alpha uses in the amps.
73, Floyd - K8AC
[/quote]
Disc ceramic caps don't come with any RF current ratings. BUT, typ the higher V rated disc ceramic also handle more current. That's why Alpha now uses the 18 kv rated discs. The RMS RF voltage across the tune cap (and any padders) is aprx .6 X loaded B+. Dunno what the loaded B+ is on that amp, but lets assume 2200 vdc loaded. RMS RF voltage would be 2200 X .6 = 1320 vac. (1866 v peak). That 1320 vac will appear across the main air variable tune cap..and also all the padders..since they are all in parallel. A 75 pf cap at 1.8 mhz has a XC of 1179 ohms.
1320 Vac / 1179 ohms = 1.12 amps rms. Any padder has to be able to handle the 1866 v peak, and also handle the 1.12 amps rms.....AND also NOT drift too much.
The issue with your oem padders is current handling capability, it's not a V issue.
Your oem setup used 5 x 75 pf caps = 375 pf. The HEC (high energy corp) HT-50, in 100 pf (now 7.5 kv rated), will handle 3.5 amps CCS @ 2.0 mhz. it has a TC of N750...(which is gonna be a lot better than any disc cap).
At least with HEC (or centralab) caps, I know exactly how much current they will handle at any freq...and also their TC.
Ok, let's say you used 4 x 100 pf HEC doorknobs for padders. Each 100 pf cap has an XC = 884 ohms. 1320 vac rms / 884 ohms = 1.49 amps. 1.49 amps / 3.5 amp rating means the cap is operating at 42% of it's rating.
To mount doorknobs correctly, machine screw one end to the chassis. Use wide copper strap to connect the hot ends...then off to the bandswitch. I use tin snips to taper the wide copper strap down to a narrow point, to insert and solder into the bandswitch assy. Ok, done that way, you have heat sinked BOTH sides of each doorknob cap.
Myself, I avoid padding a TUNE cap like the plague on my hb amps. Padding a LOAD cap is a similar, but different issue. On my hb amps, any 160m LOAD padder is at a 50 ohm point, and the RMS voltages are a LOT lower. 1.5 kw = 275 vrms across 50 ohms. Padder current with the same 100 pf cap would be just 275 / 884 ohms of xc = .31 amps rms. .31 / 3.5 amp rating = 8.8% of it's max rating.
However, the Alpha uses a PI-L..so the load cap is probably at a 200-350 ohm point. 1.5 kw = 2.45 amps into 250 ohms....= (2.45 amps X 250 ohms = 612 vrms). 612 vrms / 884 ohms of xc = .7 amp rms. .7 / 3.5 amp max rating = 19.8% of it's max rating. This is for a 100 pf load cap padder. Typ a higher value is used in some cases.
Those 20+30 kv rated disc caps are pretty cheap, including shipping from China.
You can do similar calcs if loaded B+ is different from what I assumed.
What you ultimately use will depend on how much room in there you have to play with.
