Page 1 of 1

Power Supply (capacitor) Questions

PostPosted: Fri Feb 05, 2010 9:38 pm
by antiquekid3
I have seen some of these cap boards for the ST-70, but some say they are recommended for use with a solid state rectifier only. What's so wrong with using the 5AR4 with these boards? The capacitance is a lot higher on the new boards, which would make me think there would be a larger inrush of current using a solid state rectifier than with a tube.

I intended to make my own cap board. I want at least 30µF at >500V. I found on DigiKey some decent capacitors that won't break the bank for me. They are 68µF caps at 250V and 400V. If I get 4 of each and series one with the other, I get (4) 34µF at 650V.

In some schematics, though, I see some other resistors across each capacitor. First off, what are they for? Do I have to have them?

Are there any other considerations I need to take when making the power supply? I intend to follow the original schematic pretty closely (after all, I'm not increasing the capacitance that much).

Kyle

PostPosted: Fri Feb 05, 2010 10:45 pm
by Geek
This one uses $17 worth of caps from Digikey:
http://geek.scorpiorising.ca/contrib/Ge ... _power.gif

I used it with 5AR4, 5U4 and she's living with SS diodes right now. It's rated for up to 700V, the stock PA-060 maxes at 550V with SS diodes and 125V line.

The resistors are definately needed to equalize the voltage across series caps. They also act as a bleeder. SDS board is simpler (for stock ST-70) and is available on Ebay.

Cheers!

Re: Power Supply (capacitor) Questions

PostPosted: Sat Feb 06, 2010 2:06 pm
by dcriner
antiquekid3 wrote:What's so wrong with using the 5AR4 with these boards? The capacitance is a lot higher on the new boards, which would make me think there would be a larger inrush of current using a solid state rectifier than with a tube.
The difference is that a proper solid-state rectifier can withstand a high inrush - but not so the 5AR4. When rectifier tubes get too high of an inrush, they can light up like the 4th of July.

PostPosted: Sat Feb 06, 2010 2:15 pm
by Spraiski
See the link below for information on calculating the correct resistor values for use across unequal capacitors in series.

http://www.siteswithstyle.com/VoltSecon ... cking.html

PostPosted: Sat Feb 06, 2010 3:43 pm
by dcriner
Spraiski wrote:See the link below for information on calculating the correct resistor values for use across unequal capacitors in series.
Yes, but he is proposing to stack caps of equal valve. The voltage will divide equally across the caps without resistors.

PostPosted: Sat Feb 06, 2010 5:54 pm
by antiquekid3
So with equal values (despite their voltage ratings being different), I can safely NOT use the resistors? If I want, could I stick some 500K 1/2W across them just to be safe? Would 500K 1/2W be safe or should I find another more appropriate value?

Again, I plan on using 68µF caps at 250V and 400V. DigiKey part numbers P13668-ND and P13537-ND.

Kyle

PostPosted: Sat Feb 06, 2010 7:17 pm
by Ty_Bower
antiquekid3 wrote:So with equal values (despite their voltage ratings being different), I can safely NOT use the resistors?


No. If you are stacking capacitors to get a higher voltage rating, you need to have balancing resistors to ensure that one cap doesn't try to "take all the voltage." If the resistors aren't there, you are basically hoping the equivalent resistance of the caps are equal, which may or may not be the case.

PostPosted: Sat Feb 06, 2010 7:19 pm
by dcriner
The voltages will divide in proportion to the cap ratings, i.e., 1:1.

Equal value 250-V caps in series will put half the applied voltage across each. Why are you proposing to to stack together equal value caps that have different voltage ratings?

PostPosted: Sat Feb 06, 2010 10:10 pm
by antiquekid3
So, am I getting conflicting answers? I don't think it would hurt anything to add them. I have some 154K 2W resistors. Can that work?

I'm using different voltage ratings because the 250V are cheaper than the 450, but I want more than 500V for the total voltage. Does that make sense?

Kyle

PostPosted: Sun Feb 07, 2010 6:29 am
by Ty_Bower
Gingertube gives a very thorough explanation of the math behind series capacitors. It would be useful to give it a read:

http://www.diytube.com/phpBB2/viewtopic ... 29&start=6

The math is simpler if your two caps are identical (same uF, same V rating). If you want 700V handling, I'd suggest getting a pair of 350V rated caps. If you really want to use your 250V and 450V rated parts, you will need to size the 250V cap larger (higher uF) and the 450V cap must be smaller (lower uF). Then you will need to size the balancing resistors to maintain the correct voltage split across the caps. Gingertube's writeup describes it well.

PostPosted: Sun Feb 07, 2010 9:50 am
by dcriner
antiquekid3 wrote:I'm using different voltage ratings because the 250V are cheaper than the 450, but I want more than 500V for the total voltage. Does that make sense?
Not really. The voltage will divide equally across both caps in series. If the total applied voltage is, say, 500V, then 250V will be across each cap - and the one with the 250-V rating will be right at it's rating.

With series caps of different voltage ratings, you can't just add the two voltage ratings to get the total.

On the subject of resistors, if you use them: figure power dissipation by V^2/R. Double that for safety to pick the rating.

PostPosted: Sun Feb 07, 2010 1:20 pm
by antiquekid3
Very interesting!! There's a lot more to capacitors than I originally thought!

I think I am going to get some 47µF 450V caps. I already have some 153K 1W resistors, and if I'm not so power conscious, I can just use those. Or if I am, I'll just get some 470K 1/2W.

I know some people tend to go overkill by using very large capacitors. But if the ST-70 is the most popular stereo tube amp of all time, and they used 30-20-20-20µF for it back in the day, I don't think ~25µF per section will be too bad.

Maybe I'll end up splurging and getting 100µF @ 350V...

Kyle