Cap Rolling

TurbOSquiD77 · 6606

0 Members and 1 Guest are viewing this topic.

Offline TurbOSquiD77

  • Full Member
  • ***
    • Posts: 103
on: March 03, 2017, 10:26:20 AM
What is an acceptable range of replacement values for the Dayton 0.1uF 400V capacitors on the low current board?

+ uF value (thinking 1uF)?
+/- voltage (3-600v)?

I'd like to try some 0.1uF  600V Copper Foil Paper & Wax from Jupiter :)

I don't see an output cap from them that would fit well. Maybe if wire them in parallel with other values/types it can add a bit of different flavor.

10uF flat stacks for the Mainline ;)
What stock film cap values are on the Mainline's pc board?

The builds from Zysonics are pretty cool.

Thanks!
Happy listening,

-T
« Last Edit: March 03, 2017, 03:54:38 PM by TurbOSquiD77 »



Offline Paul Joppa

  • Global Moderator
  • Hero Member
  • *****
    • Posts: 5751
Reply #1 on: March 03, 2017, 02:16:53 PM
The 0.1uF is in series with a 220 ohm resistor. The time constant (R times C) is chosen to keep the 431 chip stable, so the product should not change very much - maybe +30% to -20%? I favor keeping the resistance value near 200 ohms so the effective source impedance is similar at high frequencies to the regulator's low frequency source impedance.

Paul Joppa


Offline TurbOSquiD77

  • Full Member
  • ***
    • Posts: 103
Reply #2 on: March 03, 2017, 03:59:33 PM
Did some research - more tomorrow.

Thank you!



Offline Paul Birkeland

  • Global Moderator
  • Hero Member
  • *****
    • Posts: 19316
Reply #3 on: March 09, 2017, 12:18:16 PM
A 250V rated cap will do fine in that position.  As PJ said, the value of that cap is pretty critical.

Paul "PB" Birkeland

Bottlehead Grunt & The Repro Man


Offline kgoss

  • Sr. Member
  • ****
    • Posts: 326
Reply #4 on: March 09, 2017, 02:22:12 PM
I do like replacing some parts in BH kits with boutique parts. Maybe it's a waste of money, but it makes me happy. But one rule I have is to replace with the same value part with the same or a higher voltage rating. The last thing I want to do is change the design by altering parts.

For one thing, I am not qualified to make such changes. But more importantly PJ, PB, and Doc have a lot of science, testing, and experience invested in their wonderful circuits.  I have my fun switching parts while respecting their hard work by keeping the circuit unchanged.

Ken Goss


Offline TurbOSquiD77

  • Full Member
  • ***
    • Posts: 103
Reply #5 on: March 10, 2017, 03:25:57 AM
Dually noted, and respected  :)
Just need to be sure, not get too curious and change a part's value which could be catastrophic at times.

I'll have some findings to report over the weekend, hopefully!

Before I try, would a radial cap (.1uF) be okay to use on the low current board? I have a solid mounting idea I'm confident with, and these caps I have in mind are pretty special.

I am curious on the Russian teflon 3rd power cap bypass from the Crack. Which would be the 3rd cap in the C2A, terminal 30/31U? I can try a .22uF Russian connected in parallel there. I'll try to find a 2.2uF. I don't think these are polar/directional are they? Outer foil to least resistance... Etc. http://partsconnexion.com/capacitor_film_russian_ft3.html

Thanks guys  8)

« Last Edit: March 10, 2017, 04:02:49 AM by TurbOSquiD77 »



Offline Paul Birkeland

  • Global Moderator
  • Hero Member
  • *****
    • Posts: 19316
Reply #6 on: March 10, 2017, 07:24:46 AM
Before I try, would a radial cap (.1uF) be okay to use on the low current board? I have a solid mounting idea I'm confident with, and these caps I have in mind are pretty special.
Be sure the caps you have in mind aren't electrolytic or mylar (polyester), as those are common compositions for radial capacitors.  Also, big radial capacitors sometimes need a little more support than what's offered by their leads.
I am curious on the Russian teflon 3rd power cap bypass from the Crack.
The hybrid shunt regulator in the Crack provides extreme isolation between the power supply and the output stage.  Somewhere previously I calculated the hybrid shunt reg as looking like a 40,000H choke feeding a very large value capacitor. Consequently, putting a film cap in the power supply behind that isn't going to do a whole lot.  I would also mention that getting capacitors with metal bodies and installing them into your kit is a recipe for disaster.  I have had plenty of repairs come in where one of these capacitors is laying across several terminals and causing all sorts of operational problems.

-PB

Paul "PB" Birkeland

Bottlehead Grunt & The Repro Man


Offline TurbOSquiD77

  • Full Member
  • ***
    • Posts: 103
Reply #7 on: March 10, 2017, 08:36:52 AM
Awesome :) thanks Paul.

I'll skip on bypassing the power supply caps.

Were certain series of Nichicon caps chosen, or can any 220uF 250V+, or 10,000uF 6.3V+ lytics do well if I choose to rebuild those positions?

I also have an idea to stack a couple nice caps on the shunt regs for fun to get to 10,000uF, but that's just an idea for future. Even if it isn't audible I still like to experiment.

Some caps I can only find in the snap-in type. Can I extend the leads with Teflon coated wire on these?

Long life for power supply, and audio grade for shunt reg lytics?

Thanks  :)

-T

« Last Edit: March 10, 2017, 08:44:51 AM by TurbOSquiD77 »



Offline Paul Birkeland

  • Global Moderator
  • Hero Member
  • *****
    • Posts: 19316
Reply #8 on: March 10, 2017, 09:13:03 AM
Were certain series of Nichicon caps chosen, or can any 220uF 250V+, or 10,000uF 6.3V+ lytics do well if I choose to rebuild those positions?
The 220uF/250V caps were chosen for reliability and being reasonably sized.  Many other 220uF/250V electrolytics are so much bigger that they just don't fit in certain kits where we use these parts.  The 10,000uF/6.3V capacitors are not in the signal path, there is no benefit to changing them to something else.



Some caps I can only find in the snap-in type. Can I extend the leads with Teflon coated wire on these?

Long life for power supply, and audio grade for shunt reg lytics?
I've seen snap-in caps mounted as you describe with wires soldered to the legs, then those wires soldered onto terminal strips/boards. Not surprisingly, every time I have seen this, it's been on a repair where I've had to remove them and replace them with appropriate parts.  I would strongly advise against using snap-in capacitors in the C2A, and we even changed to leaded capacitors in the standard Crack because they are so much easier to work with.  Additionally, you're still swapping parts that are well isolated by the hybrid shunt regulator, so it will end up being an exercise in changing out very nice, brand new capacitors with different brand new capacitors.  There isn't a lot to be gained by doing this IMO.

Where I've spent my money in my own C2A is on 100uF Audyn film caps at the outputs, and the attenuator upgrade. 

Paul "PB" Birkeland

Bottlehead Grunt & The Repro Man


Offline TurbOSquiD77

  • Full Member
  • ***
    • Posts: 103
Reply #9 on: March 11, 2017, 12:40:46 PM
Awesome sounds good, thanks for the reassurance. I'd rather put my money where it matters, of course. Even though I have an obsession for replacing components just for fun  8)

I understand that the .1uF 400V caps on the low current board can be audible if swapped out, but how about the 2.2uF 250V PP caps?
These are sitting on my desk right now, and pondering if they'd be okay. They're a bit large, but can make it work. I should have looked at a different size series.

http://www.mouser.com/Search/ProductDetail.aspx?R=MKP1F042206F00JYSDvirtualkey50520000virtualkey505-M102.2%2f250%2f5

WIMA MKP1F042206F00JYSD - if the link is borked.

The stock PP's have (M)225K E250 with dots under the E, 2, and 0 - I've been having a difficult time deciphering this. Could someone please explain?

Thank you!



Offline TurbOSquiD77

  • Full Member
  • ***
    • Posts: 103
Reply #10 on: March 11, 2017, 01:17:28 PM
Also, is there a big difference between:
Ripple Current: 3130 mA and 2940 mA (6% difference)
Dissipation Factor: .2 and .15  (25% difference)

I'm learning about electrolytic caps and wondering if the difference in values such as these are minuscule or not.
For fun/learning, I'm comparing:
Nichicon UCS2E221MHD - http://www.mouser.com/ProductDetail/Nichicon/UCS2E221MHD/?qs=%2fha2pyFaduj0TJdaj%2fXN7j6r0nibbRIPYNEWr1mhhg%2fbrYKd6SH2wQ%3d%3d
Panasonic EEU-EE2E221 - http://www.mouser.com/Search/ProductDetail.aspx?R=EEU-EE2E221virtualkey66720000virtualkey667-EEU-EE2E221

The Nichicon's seem like they perform better than the Panasonic's, even though the Panasonic's are more expensive.



Offline Paul Birkeland

  • Global Moderator
  • Hero Member
  • *****
    • Posts: 19316
Reply #11 on: March 13, 2017, 07:00:43 AM
Ah yes, the 2.2uF bypass caps.  They will be about as influential as those 0.1uF caps on the same board. 

Those Wima caps will be a royal pain to mount on the board, since they have tiny little legs intended to just barely poke through a PC board, but there isn't a whole ton of room even if the leads were longer. 

225K E250 is a 2.2uF 250V capacitor.

For the electrolytic capacitors, sometimes ripple current rating is really critical (in our low voltage DC filament supplies, this becomes a big issue), and sometimes it's completely trivial (especially true in power supply capacitors downstream from the first power supply filtration resistor).   A decent barometer of capacitor quality is the lifetime rating, and both of those caps are 10,000 hours at 105C.

Frankly, I would suggest that you buy the Nichicons.  I think they are a very high performing capacitor that will last about as long as anything on the market.  I should also mention that those are the exact capacitors we supplied with your kit...

Paul "PB" Birkeland

Bottlehead Grunt & The Repro Man


Offline TurbOSquiD77

  • Full Member
  • ***
    • Posts: 103
Reply #12 on: March 13, 2017, 04:25:13 PM
Awesome, thanks! I'll experiment with board caps.


Quote
I should also mention that those are the exact capacitors we supplied with your kit...

That's why I asked  ;) haha. The specs look awesome compared to most all others.


Can capacitor lead length affect the value of the time constant, or anything else? .1uF on the low current board - mounted with oh, say 30mm lead length from board to cap.

So the .1uF contributes to the calculation of time constant. What does the 2.2uF PP cap calculate? Can other caps of the same value be used here or were PP's chosen for a specific reason?

« Last Edit: March 14, 2017, 10:38:33 AM by TurbOSquiD77 »



Offline Paul Birkeland

  • Global Moderator
  • Hero Member
  • *****
    • Posts: 19316
Reply #13 on: March 15, 2017, 10:25:25 AM
Can capacitor lead length affect the value of the time constant, or anything else? .1uF on the low current board - mounted with oh, say 30mm lead length from board to cap.
No, you don't need to worry about that.

So the .1uF contributes to the calculation of time constant. What does the 2.2uF PP cap calculate? Can other caps of the same value be used here or were PP's chosen for a specific reason?
The 2.2uF cap is a way of feeding any remaining noise on the B+ line back into the input of the TL-431 regulator.  It is a 2.2uF cap and 2.49K resistor, though the requirement of the time constant isn't necessarily that it needs to pass all audio frequencies, but rather those frequencies that make up the leftover noise.  Anything from 1uF to 3.3uF should work just fine there.

Paul "PB" Birkeland

Bottlehead Grunt & The Repro Man


Offline TurbOSquiD77

  • Full Member
  • ***
    • Posts: 103
Reply #14 on: March 30, 2017, 04:33:48 PM
So in a way the 2.2uF position is like a filter, and the .1uF position is what sets the speed flow of the circuit?

Be sure the caps you have in mind aren't electrolytic or mylar (polyester), as those are common compositions for radial capacitors. 
-PB

Lately I've been digging into output cap (100uF) possibilities.

In short:
Duelund CAST PIO (copper) = thousands.
Duelund CAST PIO (silver) = six figs. Amazing isn't it!
Mundorf Evo Silver Gold Oil 100uF - few hundred each. Might go this route but scored only a bit higher than the Evo Oils in Humble Homemade Hifi's cap test.
Stacking Mundorf Supreme Silver Gold Oil (because they don't make 100uF) - Thousands.

Bypassing a cheaper cap with a pure copper and/or silver cap of 1% or less - MUCH more cost effective.
Example:
Mundorf Evo Oil 100uF bypassed with Duelund or Jupiter Copper Foil .1uF

Running Jupiter .1uF  caps on the small board has already made quite the difference. It's not as noticeable as a speedball, but worth it IMO.

My question is: Why might mylar (polyester) not be a good choice to use in the circuit, or is this just in the .1uF/2.2uF positions on the small board? What is the smallest value recommended to run in the output cap position?

Which Audyn caps do you run, PB? Cap Q4's? Currently I've put many hours on my Mundorf Evo Oil and Dayton 100uF caps, and am looking for other 100uF possibilites that I can bypass with Jupiter's or Duelund's. I'm up for paralleling if need be, and size is not an issue as I have begun an external mounting platform project. Kind of like external speaker crossovers. So far, using copper tongue set screw lugs to externally connect the output caps is working great! Running Duelund oil impregnated cotton covered copper from the terminals to external, then connected the wire and caps with the copper lugs. Yeah, no touchy.  :)






« Last Edit: March 30, 2017, 04:36:14 PM by TurbOSquiD77 »