Orgy of Capacitors: The Cap Thread

[lehmanhill]

  In that case you really need to try the K73-16

Oh, I plan to, thanks.  Based on my good experience with Amtrans poly caps, I'm also going to try their mylar cap.  Of course, they are both on a long list of things to do
 
Speaking of projects, I do have a pair of the Jupiter HT caps.  I haven't listened to them yet because they are part of an unfinished project. 
 
To answer the question about their outside case, it appears to be cardboard infused with something, perhaps wax.  What is interesting is that the end caps appear to be epoxy where I was expecting wax.  Epoxy might be an interesting potting material for uncasing and potting an oil cap.  I would think that the epoxy would generate much less temperature in the oil and should act as a good seal.  I've seen this approach on the Rike S cap which seems to work pretty well.
 
Jac

 

[nhall]

Just for info., here is one of the K75-11's pealed .. these cap's just have a thin plastic wrap (the pale turquoise bits to the left of the bottom picture) over the inner foil/film innards .. the same green ceramic like end seals (very hard to break) - not a lot of work to strip one of these!
Haven't done the next step on these cap's yet as I just got some, but will report when I do.

 

[pelopidas]

Give the K73-17 a try as well. I was surprised to find that in my amp these actually are even better than the K73-16. The high end is just a bit more defined and the low end has more body. For input caps, these are the ones that tamed my "wanderlust" through capacitor-land.
 
Now I am repeating the same with electolytics.
 
I had some pos Jakson or some such in the psu. Replaced that with Panasonic FC and wow, night and day. But as you all know, us audiophools can just enjoy a good thing. I thought that just a bit more warmth could be had with some Silmic II.
 
I really hate breaking in capacitors and for the life of me I dont know why I decided to just put these Silmics in there without first breaking them in. 2 days of misery. So much sparkle is gone and especially the decay of delicate sounds but the mids are rich like butter pecan caramel ice cream. Voices are to die for. Everything else is to kill myself for. So I bypassed the Silmic with a K73-17 (because I have a whole bag of em now and it is much better. But these are not staying. Next up are the Nichicon KZ and Cerafine and if those don't float my boat, a Mundorf Mlytic AG. If that still isnt the answer I will have wasted 40 bucks and go back to the panasonic FC.
 
I am really amazed at how important the psu is to the overall sound of the amp. The electrolytics have as great an influence as the input caps, maybe even more. 
 
So this rambling brings me to a question. What electolytics are you running in your amps and psu and are they bypassed or not? Have you tried different caps and what effects did you find?
 
:edit:
bypassing the Silmic II with the 2.2uf K73-17 and then a 0.01uf 1837 has become something special, almost magical. The clouds have parted revealing amazing dynamics and crystaline highs, clean and deep bass with this lush mid that makes you feel the breath of the singers. Much, much fuller than the Panasonic FC by itself. However, the bypasses work from the top down and I feel they go to a ceirtain frequency and then drop off in the higher mid range.
 
This got me thinking and please correct me if I am wrong. I am not an electrical engineer, nor do I play one on TV.
 
With input caps, size depends on the input impedance of your amp and this dictates how much extension down you have. For a full 20hz -20khz my amp requires about 4.3 uf capacitor as my amp is 20k impedance.
It is my uneducated theory that in this case the electrical signals behave much like the audio waves in air itself. The high frequency signals can get "lost" in a big room (capacitor) while the bass is still loud and clear. This is why we bypass with small (ie 0.01uf) caps that pretty much only let those high frequencies through.
 
PSU capacitors don't just power the amp, they are the very electrons that the music is made of. From my brief experimenting it seems that the bypasses on the electrolytics should be much larger than what one generally considers as a bypass.
If the top end is closed up and there is not enough detail, perhaps it is because the electrolytic cannot feed a stream of 14-20khz.
So whereas the input capacitor is dealing with reference voltage the psu needs to move the cones of a speaker. So I ordered a few 22uf K73-16 to test this theory. These will have enough capacitance to make a bigger difference vs the 2.2uf that is already installed. My theory is that if 2.2uf brought clarity from 20 - 12khz (guessing) then maybe a 22uf will bring clarity from 20 -1khz (just guessing). I actually don't even know how to measure this but this is just my hypothesis that I will not be able to prove, except through subjective opinion.
 
Silmic II is supposed to be extremely low ESR, even better than Panasonic FC http://www.audiokarma.org/forums/showthread.php?t=490633
so this cannot be the final word in what I am hearing.
 
Is this worth pursuing or am I way off?

 

[dimkasta]

We usually think of it the other way around.
Electrons are not "pulled" by the circuit.
They are allowed to flow through it (least resistance path etc). If the psu does not have the ability to provide the speed that frequencies need, we have rolled of highs, or fluffy bass, or whatever.
A very good example in this is the Gainclone. The usual schematic gives virtually non-existent bass. Add to it some nice Panasonic 10.000uF electrolytics, and you get a very nice boost.
Or add some Sikorels and it will blow your mind
 
Input caps are a different thing. An ideal load would have infinite input impedance, and the capacitor would not stop any frequencies (perhaps only some because of its dissipation factor, dielectric loses etc).
In real world though, input caps form a first order highpass filter with the finite input impedance of the next stage, blocking some bass.
 
My current choice for 22K amps is 10uF and they have a significantly deeper bass in my system in comparison to 4.7uF
 
Unfortunately no K73-17 caps at 10uF... I might try them on my phono though...

 

[lehmanhill]

Next up are the Nichicon KZ and Cerafine and if those don't float my boat, a Mundorf Mlytic AG.

 
I have used the Mundorf Mlytic AG in the power supply of the MyRef Fremen Edition.  In comparison to the Panasonic FC, they make an audible improvement, especially in the bass.  Apparently, the improved sound quality comes in part from low ESR as the Mundorfs are significantly lower than many.  Worth it in the case of the MyRef, but definitely pricey.
 
I haven't used bypass caps on the power supply.  There is an alternative.  In my, as yet unfinished, Aleph J amp, I am using 10 Nichicon FW (3300 uF) in parallel for each rail.  By paralleling electrolytics of smaller value, both ESR and ESL are significantly reduced, to the point that the 33,000 uF combined capacitor has an ESR about the same as the Mundorf.  The price of the parallel caps are about 1/2 the cost of the Mundorf of the same size.
 
Have fun exploring.

 

[dimkasta]

Compared to Panasonic FC I tend to prefer the new line the FRs
 
They have a better temperature rating and lower ESR
 
I haven t used them in a main psu filter though...
 
In my shigaclone I currently use Sikorels at 10000uF and results were phenomenal...

 

[dimkasta]

Has anyone tried the KSO mica caps?
 
I am currently using Cornel Dubilier silver micas on my phono, but I am going to channel match them soon and I am considering alternatives.
 
Any other suggestions for a phono?
 
I need 1000 and 3000pF

 

[pelopidas]

Those are some pretty small values.. I tried the SGM 2400pf silver micas but they are too small for for application. They are pretty though
http://www.ebay.com/sch/i.html?_odkw=KSO+Silver+Mica&_osacat=0&_from=R40&_from=R40&_trksid=p2045573.m570.l1313.TR0.TRC0.H0.Xsgm+Silver+Mica&_nkw=sgm+Silver+Mica&_sacat=0
 
Why not some FT-1 teflons?
http://www.ebay.com/sch/i.html?_from=R40&_trksid=p2050601.m570.l1313.TR0.TRC0.H0.Xft-1+teflon&_nkw=ft-1+teflon&_sacat=0

 

[nhall]

For small values I've been using the good ol' Russian Polystyrene cap's - those block ones that tend to come in pale blue, green, beige .. you can easily peal the outer plastic casing off them with a blade if you feel it worth while, and inside there is just a block of solid resin material (very hard indeed - couldn't easily damage with a blade - it shatters rather than cuts) - only thing to watch out for is that a few I've pealed the case off have not been entirely encased with the resin, and I presume an air bubble in the casting process (looks like they just put the cap leads and innards into the outer shell and pour in the resin mix) has formed exposing the inner foil coils, so I've then potted them up in the beeswax.
 
These are the chap's I'm talking about:-

 

[dimkasta]

Yes I was considering the FT-1s too. I was just wondering if someone has done some more extensive testing on RIAA networks. Micas have a nice reputation there.

 

[nhall]

Sorry Dimkasta, I couldn't truthfully say specifically one cap v's another here. Problem is I really don't get enough time to do every cap comparisons, but my feelings are that Silva mica's were always rated as good sounding and stable for small value caps in guitar circuits (so I too have a load of them at home, for that job), however, in the past the older generation (were discontinued, but now I understand back by popular demand) European poystyrene's were always rated as a good sounding stable option for RIAA use, so I figured that the ol' Rusky one's would be at least as good, and unlike some of the older European polystyrene's don't have a metal outer jacket.
 
If you hold the Russian beige one's up to the light you can see what's inside, and where I've taken the jacket off a few  green ones, there's been that air pocket that revealed the innards to be tightly wound long thin cylindrical coil innards (some values appear to be made of two tightly wound coils in parallel between the legs), and then all cast in the resin within the outer box jacket.
 
I did however build the first phono stage I use with PIO's as per the original, then went to a mix of PETP's / Teflons & these poysterene's (where the values were smaller the poystyrene's fitted the bill), and I can say with absolute truth that the second build was far more detailed, everything was better defined, and bass was nowhere near as overblown, and in fact although initially I thought the bass has got leaner, it really was just properly defined with great timbre and did in fact go lower where it was recorded like that (the previous build gave the impression of far more bass, much heavier and fuller, but anyone could here by comparison that it was artificially so). The highs of course were also more defined and extended.
In that circuit I did use a 1000pF polystyrene in the EQ, and also used a second one as a paralleling cap over a PETP to get a correct value.
 
My next build isn't far away (time needed to do all the denuding and potting), and will include Russian PETP's and polystyrene's, all denuded and beeswax potted. I have even potted up the poystyrene's, although they are a solid cast block inside their outer jacket, because I reckon some of the beeswax potting 'majic' may well come from additional damping  .. oh, and from your recommendation I've ordered some Sikorel's to try this time for the on-board electrolytics (just to be consistent I'll denude them too after measuring & pairing).
 
I think given a bank account full of whatever cash I needed, I'd be trying the Duelund cap's, but I don't feel those Russian poystyrene's have let me down in any way, or held back any detail .. I'll be honest and tell you what happens with the all beeswax potted version one way or the other, regardless of all the time I'll have put in, if it doesn't improve, or isn't that much better than the standard version I'll say .. but looking forward to hearing what all the buzz-zzz is about!

 

[dimkasta]

The biggest value I have seen on Duelund caps is 10nF anyway, so rather not adequate for such RIAA filters...
 
One of the plus sides of RIAA is that it can be designed for really low values, going into few hundreds or picos, allowing the use of smaller and better quality caps.
 
I remember Thorsten suggesting the use of variable air caps in his EAR 834p clone. I always wanted to try that, but I am not that comfortable adjusting RIAA values...

 

[nhall]

Those Rusky cap's are so cheap, you won't loose by trying them, I reckon you'll like em. They do low values, so a cheap bag load of them and you can experiment a bit.
 
[Just a bit of a venting session coming up!]
I do admit (not here on this forum of course) that I've got to take some of the general stuff out there with a pinch of salt .. too much is said in order to 'join in' and there is a kinda one-upmanship going on with a lot of them out there.
I've read a lot, and you can easily get sucked into going down a blind alley or following a route and willing yourself to hear differences other wax lyrical about. 
 
Case in point, I just recently read a thread where someone was citing specific songs and saying how much a set of different phono plugs made in terms of detail retrieval on his system, saying how suddenly he was hearing all this additional low level detail .. well, I wouldn't doubt he heard differences, and those did improve on what he previously had, but come on .. in context ..  unless he'd swapped the cable as well, I sincerely doubt it was 'that' huge an upgrade .. the way it was put you'd think all of a sudden he'd just built a complete new phono stage!
 
I realise we all want the best for our efforts to be worthwhile, and hope to find the answer out there, but us what build a lot know that some of the gains are quite low level, and it's the cumulative effect we're striving for - but, and it's often the big but - sometimes there's only so far you can push a particular circuit you are working on, and greater gains will be found elsewhere in the chain.
 
[Little true store from times of yor]
I had a really shocking experience many years ago when I'd spent some time (I was single then and had more time to play too) taking a particular (well thought of) valve stage to it's ultimate level of reg'd this that and the other, with expensive cap's and you name it (actually thinking back this was pre the internet really happening for the common man), and I met someone who I became mates with who had taught himself to design and build his own transformers .. anyway long storey short, I had just been to the big Ramada HiFi show (as it was then just outside London) and had been very impressed with what AudioNote was then demo'ing, and the huge sounds from a full range horn system also demo'd there, anyway, I went round his gaff to hear his system (see what he was doing with his transformers), and was completely knocked over with his system .. it literately looked like a pile of old crap 70's bit of HiFi equipment, bits of wire and tatty old cases, etc.. but boy, it's sound was the most realistic presentation I'd ever heard .. and he being a classical music fan got a lot of instrument separation in full orchestral works .. some other guys I later met (one had the then full AudioNote system with top silver IO cart / top Helius arm / Voyd ref turntable and huge outboard psu/ silver AN phono & pre and the Ongaku amp.. and he was also floored by this guy's system ..
Well, even now I know what's in that system circuit wise, I still don't know how it sounded as good as it did .. he only had a mid priced Ortophon mc, a hugely tweaked up Thorens t/t, and all the components for all his home built equipment were salvaged parts as he rarely had two ha-pennies to rub together .. one guy who came around one time decided to copy all his front end equipment, only thinking he'd be clever, bought a lot of very expensive cap's resistors, etc .. guess what .. his never ever sounded anywhere near as good as that guy's equipment!! - Now I'm not trying to claim there's 'magic' out there (I don't believe in it), and I'm not saying it's a bad idea to get the best you can afford - but what I'm trying to demo is that sometimes you can go too far on small things and miss a bigger picture!
 
BTW I still don't really see what made his system so good (yes, for his gear you'd really claim the players were in the room with you) .. but there were a couple of advantages he had - 1. He'd taught himself how to use the old test equipment he had thoroughly, to the point where he could really use it to see deep into what he was doing - 2. by designing his own transformers actually with the output values he used gave him a huge advantage over most DIY'ers - 3. he couldn't ever afford expensive gear, so learned were to find good quality parts, and how to make what he couldn't afford to buy (not cap's .. he just learned what old gear he could pick up for nothing that had good industrial grade parts in them .. though he did sometimes make his own resistors when need be)  .. 4. and he never got hung up on fancy parts - if the circuit didn't get him where he wanted to be, he'd dig in and try another route (his place was just littered with old crappy chassis and different builds!
 
BTW I did build a few of the Thorsten circuits a few years back - although clearly good, I personally didn't think they were as hugely impressive as the billing made them out to be .. so I went other directions..
 
Don't get me wrong, I do personally like all this tweakery, and will be doing this beeswaxing business throughout the next build, but after that will leave it there, as I think I'll be in danger of chasing stardust.
 
I know it's off topic, but in a parallel way to the above, thought it was a worthwhile read to bring one down to earth again:
http://www.roger-russell.com/wire/wire.htm
 
Chow for now..

 

[JamieMcC]

  For small values I've been using the good ol' Russian Polystyrene cap's - those block ones that tend to come in pale blue, green, beige .. you can easily peal the outer plastic casing off them with a blade if you feel it worth while, and inside there is just a block of solid resin material (very hard indeed - couldn't easily damage with a blade - it shatters rather than cuts) - only thing to watch out for is that a few I've pealed the case off have not been entirely encased with the resin, and I presume an air bubble in the casting process (looks like they just put the cap leads and innards into the outer shell and pour in the resin mix) has formed exposing the inner foil coils, so I've then potted them up in the beeswax.
 
These are the chap's I'm talking about:-

I tried some of those polystyrenes in my head amp and was very impressed with the initial results, they ultimately proved to be to fatiguing for extended listening anything more than a hour with them was to much for me with my T1's  my HD650 was a better pairing. 
 
Ultimately the Russian teflons combined with some pio proved a more refined experience in my case even besting (just) some expensive Mundorf M-caps.
 

 

[dimkasta]

What values do you use and what is their position?