Smaller caps, bigger choke?

[schoolie]

Hello all,

  This is regarding my 5f6a.  If I were to use less capacitance in my filter caps, but increase choke inductance from 4Hy to 7Hy,  the ripple on power tube screens and preamp voltages would actually be much lower, but higher plate ripple would increase.  But I'm wondering what effect this would have on the frequency response of the amp.  Anybody have any thought on this.  I would only be changing the reservoir cap from 44uF to 38uF and C2 and C3  from 25uF to 19uF.    This is all in my quest of replacing all electrolytic caps with film caps, which I already know is foolish:)  Thanks and apologies in advance!

[schoolie]

Thanks, DrGonzo!  Yes, the Duncan PSU II seems to like my idea,  but it can't show me how the amp will react to transient loads and how it will sound.    Those WIMA 800V DC Link caps are smaller than the Solens and I've liked them in other amps.
   The real problem that I'm trying to address is that I placed the electrolytic  filter caps over the power tube sockets and they are cooking.  The WIMAs I can tuck in a different area of the chassis.  I think I'll just have to try this change for myself and see if I like the sound.

Cheers!

 

[FYL]

The Wima DC Link caps are DC-rated. They are OK in power supplies only after the first node, which sees a lot of pulsed AC from the recto and associated ripple.

AC taxes the cap power-wise. It's dissipation factor is much higher - up to 20 times for some film dielectrics, and the cap can overheat real fast, melting it's dielectric or provoking internal corona discharges.

You should use use suitable AC-rated caps or greatly derated DC caps - typically a DC 600V-rated should only see 200V max of AC

[schoolie]

The Wima DC Link caps are DC-rated. They are OK in power supplies only after the first node, which sees a lot of pulsed AC from the recto and associated ripple.

AC taxes the cap power-wise. It's dissipation factor is much higher - up to 20 times for some film dielectrics, and the cap can overheat real fast, melting it's dielectric or provoking internal corona discharges.

You should use use suitable AC-rated caps or greatly derated DC caps - typically a DC 600V-rated should only see 200V max of AC

Thanks, FYL!  The datasheet states 0.006 Ohms ESR at 10KHZ and Irms of 10A.   Am I interpreting the specs incorrectly?   This is the datasheet:

http://www.wima.com/EN/WIMA_DC_Link_MKP_4.pdf

[jjasilli]

IMHO the value changes you mention are too small to make a real difference. 

Choke: 1 - 10H is fine;  or 20 -25H for Marshall territory;  Caps:  44 or 38; 25 or 19 are within the margin of error of cap manufacture.

Perhaps you'll hear a distinction if you A/B those different values in your living room.  But in actual use I think it would be "a distinction without a difference". 

[schoolie]

IMHO the value changes you mention are too small to make a real difference. 

Choke: 1 - 10H is fine;  or 20 -25H for Marshall territory;  Caps:  44 or 38; 25 or 19 are within the margin of error of cap manufacture.

Perhaps you'll hear a distinction if you A/B those different values in your living room.  But in actual use I think it would be "a distinction without a difference". 

Thank you, jjasilli!  I think I will try with the original choke first.  I ran the differences through DUncan PSU II and there was a 40% increase in screen node ripple, but it was 257mV vs 176mV.  Hopefully not audible.

[jjasilli]

Some noise in a guitar amp is may be OK.  High gain amps can be pretty noisy.  But add bass, drums, and ambient noise in a venue, and noise in a guitar rig becomes inaudible anyway.  For hi-fi amps, noise is much more of an issue.

[HotBluePlates]

...  I ran the differences through DUncan PSU II and there was a 40% increase in screen node ripple, but it was 257mV vs 176mV.  Hopefully not audible.

257mV a.c. at a node of ~400v d.c. = 0.06% = nothing

Not even worth thinking twice about.

[schoolie]

Thank you, HotBluePlates!

[HotBluePlates]

Would I expect a difference in tone between electrolytic caps and film, possibly. 

I've built 2 amps that used polypropylene filter caps in place of electrolytics.

Values stayed essentially the same as the original spec (maybe 22uF instead of 20uF per the schematic), and I can hear no difference in tone of the amps.

The only reason I chose film filter caps was I had the space available, and I'd never have to replace the caps. The 630v rating also gave good safety margin for start-up surges.

[schoolie]

Drgonzonm,  I bought the MKP caps from Mouser, and I remember that the price for the 800V/20uF model is around $9.

[HotBluePlates]

HBP,  glad you added to the thread.  Out of curiosity, What was difference in prices between the polys and the electrolytics?  (Cost ratio would be fine).  (Am I too lazy to peruse Dougs listed prices?) 

It's been a while since I bought them, so I don't know what I originally paid. Going price for a Solen 22uF 630v cap seems to be ~$12, though better deals can probably be found. I also found a private seller let 4 go for $15, so shop around.

A 20uF 500v Atom cap costs $15 from Doug, but a 22uF 500v Illinois costs $3.75.