Capacitors' Voltage Rating

Nov 28, 2007 at 3:29 PM Thread Starter Post #1 of 19

DolbyR

100+ Head-Fier
Joined
Sep 13, 2005
Posts
220
Likes
10
Is there any issue/advantage in applying higher rated capacitors in a PSU circuit?
I'm just choosing the capacitors for my SOHA, I will be using Fujicon (chinese crap) caps as I cant find anything higher quality locally, although im planning on upgrading them later to BG's or something.
So the indicated rating in the BOM are 10v, 16v, 35 and 100v. I just like the look of big caps, so I was thinking of putting 100v caps all around, maybe even 200v or 400v for the 100v positions as these ones would be 16 and 18mm like the places in Digi's board.
Could someone tell me if this change in voltage rating will make any positive or negative difference?
 
Nov 28, 2007 at 6:40 PM Post #2 of 19
Higher voltage caps sometimes have better leakage characteristics.

If you're using the digi01 board, you should be most concerned about whether the caps will physically fit on the board.
 
Nov 28, 2007 at 8:04 PM Post #3 of 19
Thanks eric.
They will fit perfectly as the ones i'm looking at are exactly 18 and 16mm., while the ones with the recommended ratings would be 8-12mm. That's the main reason I even though about these, I oriented myself according to the physical size.
 
Nov 28, 2007 at 8:28 PM Post #4 of 19
The other thing to watch out for is many high voltage caps are snap-in rather than standard radial leads... snap-ins won't fit without drilling out the board holes. At which point you lose your thru-hole plating.
 
Nov 29, 2007 at 3:54 AM Post #5 of 19
Don't cheap on caps. Put decent caps in the first time.

Look for some rubycon, nichicon, chemicon, matsh****a(panasonic) or samxon.

Crappy caps have been well known to leak or explode, taking stuff out with them. Some within hours if they are pushed too hard.
 
Nov 29, 2007 at 4:24 AM Post #6 of 19
Well, I could get atleast Panasonic FM's ordered from Farnell, but what about for the places of "Stdandard Blackgate (boutique) Alt. Blackgate NX" and"Nichicon PW", would the FMs be OK?
As for the film caps, i can get wima MKPs locally, are they fine?
 
Nov 29, 2007 at 4:30 AM Post #7 of 19
Panasonic FMs are really good caps... probably better than BG stds.; certainly lower ESR. You won't find any 100V however in FM.

Wimas are fine, good quality.
 
Nov 29, 2007 at 6:01 AM Post #8 of 19
Quote:

Originally Posted by DolbyR /img/forum/go_quote.gif
Well, I could get atleast Panasonic FM's ordered from Farnell, but what about for the places of "Stdandard Blackgate (boutique) Alt. Blackgate NX" and "Nichicon PW", would the FMs be OK?
As for the film caps, i can get wima MKPs locally, are they fine?



Panasonic FM's will be good anywhere except for DC Blocking. a film should be used for that for best results.

If you actually need 100v caps (not sure why you would) the FC range has those values.
 
Nov 29, 2007 at 6:04 AM Post #9 of 19
Quote:

Originally Posted by LawnGnome /img/forum/go_quote.gif
Panasonic FM's will be good anywhere except for DC Blocking. a film should be used for that for best results.

If you actually need 100v caps (not sure why you would) the FC range has those values.



The SOHA actually needs 100v caps in the high-voltage supply, which should run about 81v.

Aside from the HV supply, the SOHA features a film dc-blocking cap between the tube gain stage and the solid-state output buffer.

fwiw it was prototyped with Xicons in all electrolytic positions, but it may be worthwhile to use a good quality cap in the cathode bypass.
 
Nov 29, 2007 at 7:35 AM Post #10 of 19
Quote:

Originally Posted by ericj /img/forum/go_quote.gif
The SOHA actually needs 100v caps in the high-voltage supply, which should run about 81v.

Aside from the HV supply, the SOHA features a film dc-blocking cap between the tube gain stage and the solid-state output buffer.

fwiw it was prototyped with Xicons in all electrolytic positions, but it may be worthwhile to use a good quality cap in the cathode bypass.



Agreed. You'll have big problems if those B+ caps are not 100V.

As for the others, I've used the ES caps in the cathode bypass positions with good results. I'm not sure that Black Gates are warranted, however. Strictly speaking, the cathode bypass caps are not really signal coupling caps - more of an "open door" for the signal wave if it needs to expand. You can do a lot to hurt the sound with the wrong cap in that position (too small or a blooming mids film cap), but not a whole lot to make it better. Just put a decent quality cap at 1000uf there and forget about it - that's pretty much quoting Runeight about that position, too.
wink.gif


The only true coupling caps are the ones that Eric mentions - the film couplers between the tube and opamp stage. Those will benefit from some very nice caps such as Vitamin Q's, etc. I've used Wima's on all of mine, though, and they sound great.
 
Nov 29, 2007 at 8:53 AM Post #11 of 19
IF all else is equal, higher voltage rated caps are superior. They allow (by virtue of their larger can size, within same family of caps) lower ESR while not causing increased charge time from the PSU.

The key is to determine the sweet spot for capacitance, sometimes the larger size of a larger voltage rated cap is real-estate better spent on more capacitance. When possible, a little higher capacitance is better.
 
Nov 29, 2007 at 1:28 PM Post #12 of 19
how does the FC series compare with the FM? As the 100V caps i'll have to take from the FC series, would there be advantage (in the means of quality) to take all thre caps from this series, atleast in Farnell, the price is higher that the FM series, but in terms of characteristics, is there a difference. Sorry, I just dont understand all these parameters in order to compare them by myself.
 
Nov 29, 2007 at 1:55 PM Post #13 of 19
Quote:

Originally Posted by DolbyR /img/forum/go_quote.gif
how does the FC series compare with the FM? As the 100V caps i'll have to take from the FC series, would there be advantage (in the means of quality) to take all thre caps from this series, atleast in Farnell, the price is higher that the FM series, but in terms of characteristics, is there a difference. Sorry, I just dont understand all these parameters in order to compare them by myself.


The FM's have better specs than almost all readily-available electrolytics. Simply, they are a sterling power cap. As you note, however, they are not available in as many sizes as other types. There is nothing wrong with mixing in the FC's in this case. The 100V caps apply to a specific part of the SOHA circuit - the B+ voltage on the tube. C3, C4, C5, and C6 are these caps. FC's or Nichicon UPW's are great choices. I prefer FC's, but that's probably just from some favorable hype from the speaker/guitar guys over on DIYaudio. Plus, I delude myself into thinking some of the FM magic came from the FC's.

The other caps of note are the ones behind the twin voltage regulators of the opamp rails - C9 and C10. Those may benefit from upsizing to 1000uf or more. C7 and C8 are the fast caps on the front side of the regulator and need to respond quickly to the music signal. FM's are a great choice for all four positions. but UPW's, UHE's or FC's would be fine, too.

C12 is the cap reserve for the heater regulation. It's not critical IMHO and may be any old 470uf-35V cap. C11 is the same - not that critical (comparatively speaking), since it only supplies the heater.

C1 and C13 are the cathode bypass caps for the tube. These need to be 1000uf for the best bass. It's possible that 470uf's will have as much bass, but it hasn't been tested and documented by anyone AFAIK. If you run the corner frequency equation, it will probably tell you that 220uf's are sufficient, but we're finding from the experience with the MAX and from Runeight's years of first-hand experience that maybe this isn't so. As Runeight has said, the 1000uf is not much more expensive and it will remove all doubt.

Traditionally, this has been a boutique signal cap position, but I'm not so sure anymore that's necessary. Nichicon Muse ES, KZ, FG, or Elna Silmic RFS, Cerafine are traditional choices. IMHO, Black Gates are not justified and would be extremely expensive. As mentioned earlier, that's because only the strongest amplitude signals hit this cap, I think - when an expansion of the wave is necessary (room to grow). Not very scientific, but as mentioned, the rote equation doesn't seem to work anyway.

C2 and C14 are the film coupling caps that connect the tube stage to the opamp output stage. They are definitely full-fledged signal caps - every bit of the signal passes through them at all times. The possibilities are almost endless. You might refer to Dsavitsk's excellent reviews at:
Notes on Output Coupling Capacitors for ideas on your selection.
 
Nov 30, 2007 at 5:50 AM Post #14 of 19
Thanks for the great write-up Tomb.
Well, more problems
smily_headphones1.gif
Farnell doesn't seem to have any of those caps you recommended me for C1, C13, neither FM's in 47uF for C7, C8. They do have Rubycon (ML, MS5, RX30, YXF, YXM, ZA, ZL, ZLG, ZG, ZLH, ZT) and Nichicon (LGU, TVX, UHD, UPM, UPS, UUD, UWX).
Does someone have an idea if those are any good?
 

Users who are viewing this thread

Back
Top