Orgy of Capacitors: The Cap Thread

[dimkasta]

Ela re patrida
 
I am very interested in your K73-16 tests...
 
K75 alone is indeed very dark. And it was still relatively dark bypassed with FT-3 and SGM. The real difference came from the 4.7uF. This was much better than the 2.2uF and the 10uF ones. And when I soldered two together (bypassed with FT3 and SGM of course), it was like I removed plugs from my ears...
 
The reason I have not yet bought K73 is some really negative comments I 've read. I will try them though... I really do not like that big lump of capacitors in my preamp.
 
Unfortunately I need a big value at 10uF...
 
Any thoughts on EPCOS films at 10uF?

 

[JamieMcC]

   
Any thoughts on EPCOS films at 10uF?

 
I have used epcos 89uf films and epcos 100uf pio films and found they to be good alternatives to the normal audio branded makes.  Particularly liking the 100uf pio I have tried

 

[pelopidas]

K73-16
Well sofar I cant find a fault with them. It seems like the perfect input cap for a T-amp. Extremely detailed, even shockingly so, but not harsh. I am not getting any sibilance or harshness but it is just so open up top that I am spending way too much time listening to music thinking about all that I missed while I had the K75 in there. Well, because I cant help myself and leave a good thing alone, I bypassed these with some 0.22uf FT-3 and it does absolutely nothing for it. Almost feels as if I lost a few dB and maybe also some detail. I have to let it settle in and see what happens but I definitely am shocked by how good the K73 is in my setup. I was about to buy some Mundorf Supremes but now I am worried that that will be wasted money. I guess I need to keep going through my albums until I find something that does not work. DIY audio is an OCD friggen manic addiction.
Here are some 10uf K73-16  10 for 10$ from Russia, so expect 3 weeks delivery
http://www.ebay.com/itm/10-x-10-uF-63V-PETP-capacitors-K73-16-NEW-LOOK-/181209687754?pt=LH_DefaultDomain_0&hash=item2a30f066ca
 
http://www.ebay.com/itm/10-uF-63-V-MATCHED-RUSSIAN-PETP-AUDIO-CAPACITORS-K73-16-73-16-/271450530856?pt=Vintage_Electronics_R2&hash=item3f33b66428
 
2 for 5$ and get them in a week.
They are way too cheap not to try.

 

[dimkasta]

You got me really intrigued... I will get some

 

[pelopidas]

Hey dimkasta,
 
I am trying to figure out the teflons. I have the K72 and FT-3 and have tried them both but they are not what I was expecting. The 1837 as bypass is fairly easy to hear. I just opens up the top and gives everything air and also tightens up the image. I just got the teflons in the last week and have been playing around with them but im not sure I hear what I am looking for. With the K75 the K72 actually muddied the sound, made it less coherent. Now I am listening to the K73 with the FT-3 and it is also "less". The details are less, the air is less and the dynamics are less. Is this really just the need for burn-in and by christmas it will sound great? What has your experience been with the teflons? What is it that you hear when you hook them up to your K75 and SGM?
Did they originally suck detail out of the music before giving it back again?
I am just somewhat underwhelmed and thinking about pulling them out and just going for the K73 again, all by itself, because it sounded tight from top to bottom and now with the FT-3 its still very good, much better than the k75 combo I was running but also flatter?
So do you recommend patience or should I pull the teflon?

 

[JamieMcC]

The Teflons often require around 200hrs burn in put the cd player or iTunes library on shuffle for a week.

 

[dimkasta]

First of all, I will clarify that I am trying these at the output of a DCB1 preamplifier previously driving an Electrocompaniet AW120 and now a pair of Gainclones, who in turn drive a pair of Montana EP Signature speakers.
 
I have not tested K72 or 1837, only the FT3s, FT2s, K75, Mundorf supremes, and Mundorf supreme silver-gold.
 
I initially used a single FT3 at 0.47uF with my 220k input impedance amp and the results were phenomenal. There was nothing to be desired. Top to bottom the sound was excellent. The solo FT3s completely moped the floor with everything else, including the Mundorf Silver-Gold ones (not oil). Excellent clarity and bandwidth way beyond what the roll-off frequency calculations would suggest in bass.
 
Now I have switched to a 22K input impedance amp, so I have to use about 10x the capacitance.
 
The original suggestion came from Salas who proposed that I used a 2.2uF K75-10 bypassed with an FT3 for my phono. I never tried it there, but when I switched to the new amps it was the first thing that I tried.
 
Initially I used 2x2.2uF K75s alone. The result was as you said DARK... really vintage sound.
 
After a few minutes I added an FT3 at 0.47uF if I remember correctly. The change was big, but it still was very far away from what I had in mind from the FT3s alone.
 
After a week or so, not much had changed, but then my K75s at 10uF arrived. They were really big and impressive so I hooked them up the same evening. Being disappointed by the K75 alone, I bypassed them immediately with an 0.1uF FT3, keeping the proposed ratio of 1/100 to prevent smearing because of overlapping frequency performances. The result was significant improvement. Both in clarity and bass, the latter obviously from the lower roll off frequency. At that point I added a 1nF silver mica, and there was a very significant improvement in the high end as well.
 
I used those for another fortnight or so, and when my K75s at 4.7uF arrived, I was really not going to test them. I had some negative thoughts, but I tended to attribute them to other weakish links in my system.
 
I am not a guy that leaves stuff unchecked though, so the same day I hooked up 2x4.7uF again bypassed by 0.1uF FT3s and 1nF SGM.
 
The result was just WOW!! Currently I do not feel I am missing much once again. The sound is super detailed and focused in a huge and deep soundstage, with a slight emphasis on the mid-lows and mid-highs giving it a very smooth texture and slightly sweet sensation, making it really easy to the ear.
 
To be honest I would like a bit more active highs, but I was thinking things might come around with breaking in....
 
That is mostly why I am considering the K73s.
 
Anyway, I have to say that people keep giving opinions, but they seldom give more info about their setup, the most important being what input impedance they are feeding. This pretty much defines the roll-off frequency and angle, so it is pretty important. Especially when comparing caps with different values.
 
By the way, for other applications, I tend to prefer either low value FT1s or Silver micas.
 
For local filtering of high speed digital stuff for example I really like the 1 and 10nF SGM and cornel dubillier silver micas, while for other filtering duties like some spots in my shigaclone I really like the FT1s at 10nF again.
 
I also had excellent results in PSU output and decoupling duties from the Panasonic FRs. Although keep in mind that they are really low ESR, so they might not be suitable for use right after a series regulator chip. They are perfect for local decoupling though and their character seems to dominate if used close to the load.

 

[pelopidas]

That is an awesome setup
Mine is much more humble - a Xonar as dac (modded of course) going to a modded t-amp. I am using these caps on the input stage of the t-amp and the input impedance is 20k. The roll off is at about 37 hertz or so. This is not a problem right now as I am using a separate sub. My speakers are Martin Logan MLT-2. So it's a very simple setup for playing flac and cd's. It's at my desk and I'm using it near field as I am less than a yard away.
The dac is using ad797 in I/V and lme49710 for buffer. The output buffers on the dac have been bridged which is why I need the input caps on the amp which is without an attenuator- so it's working as a power amp. I'm trying to remove as many variables and veils between the source and the speakers. It is very sensitive to even little changes which brings me to another observation. The ft-3s have been burning in and they are sounding much clearer but overall they still seem to be sucking the upper bass and some of the dynamic range from the k73 s. I am using the isotek burnin cd and it really accelerates things but I still have many hours to go before a final verdict.

 

[dimkasta]

If FT3s are killing your mojo bypassing the K73s, why use them? What are you trying to do with them?
 
I have ordered some 10uF K73s by the way... 0,5 euros each

 

[pelopidas]

Ha! that is a good question. At this point I don't know why I am using them.
 
I have heard about how revealing the teflons are and seeing how new the K73's are to me, I wanted to see if there is anything I am missing, sonically speaking, with the K73. The K73 seemed to have a short break in period for me. After about 10 hours the sound was completely stable and stunning and having these teflons just sitting here staring at me got the best of my curiosity.
 
By trying different bypasses vs stock I am looking for information in songs that I know very well (which my wife hates by now) which I have not heard without the bypasses. Also the usual "Angst und Drang" to try something new in the odd chance that it might work out well combined with the fame of the russian teflons.
 
So I guess it is just simple sonic greed. The enthusiasm of a neophyte vs the temperance, patience and wisdom of experience.
 
Now I am very curious to hear about your impression of the K73. Fortunately these are cheap experiments and that makes them all the more fun as I feel no obligation to like them unlike spendier caps where it can be harder to be objective when not only the stereo but also the money is crying.
 
So I have been burning in the caps every night by removing the speaker wires from the amp and hooking up an 8ohm resistor with the K72 in line ( to burn those in as well ) and letting it play the Isotek on repeat for the rest of the day. The K72 are so resonant I can hear them across the room. I can hear the frequency sweeps loud and clear. Anyone else have "singing" capacitors?
 
Before posting this I played around some more:
 
I removed the FT-3 and listened to just the k73 again by them selves. I am not missing any transparency or detail and so I a/b a few times with the ft-3 in and out. My wife has a new song she hates. The interesting thing is I cant even hear the FT-3. It is doing nothing.
 
So I decide to try the K72 and start A/B in and out testing. This capacitor actually has an effect and it wasn't what I was expecting. There is not more detail or air because it seems the K73 are that phenomenal. What the K72 did do was level out the soundstage. This took a while to figure out what I was hearing. I was testing with Espers Byss & Abyss and towards the end there are two main guitars and the center guitar just sounded as if he were several feet higher than the left guitar, but with the K72s in there suddenly the stage is more correct. I dont want to say flat because it is anything but flat but it sounds as though the musicians are on the same stage playing next to each other rather than one sitting on a ladder.
 
So FT-3 got voted off the Island and and K73 plus 0.1 K72 is king of the cap hill for now. I have some silver micas on order thanks to observations in this post.  I guess once the micas are here and I can see what they sound like I was thinking about getting more to replace the bypass capacitors for the opamps on the Xonar. I got that idea from some other thread where they said it made a significant improvement.
This got me thinking about your Gainclone. Dont they also have opamps that need bypassing? Hahaha...

 

[dimkasta]

Hmm
 
I always thought of differences in height to be interactions with the room, and usually considered them a good thing. Makes things like reverb or other guitar effects more pronounced and impressive.
For example in "What do you want from me" from Division Bell by Pink Floyd where the guitar joins the bass/drums line in a huge sound-wall of awesome
Quite different sound though...
I cannot say what you are hearing is good or bad... I do not know the song you are quoting...
 
Anyway... Whatever floats your boat
 
About micro-phonics, I never did frequency sweeps, but I can confirm that some FT-1s I modified with wood and wax had a significant improvement.
 
Here are some photos...
 

 
The metal casing is very easy to cut on the side, and the teflon thingy just slides out.
 
Below is the wax sealed cap, measured again for confirmation that I did not screw it up. As you can see it once again measures a perfect Q.
 

 
K72s have a steel casing, so I guess they would resonate more than the FTs which are in aluminum.
 
In any case, attaching them firmly with tie-wrap to a solid base also make a biggish difference.
 
I got some more wenge wood recently, so I am going to try treating the K76s too and see what happens...
 
Oh and make sure that you clean the leads before soldering. Those silver leads tend to oxidize having a blackish appearance. Flux cleans them a bit,but I have noticed a significant improvement if I sand them lightly with ~600 grit until they are nice and shiny again before soldering
 
Oh and about the Gainclones, the chips used are just AB push pull amps integrated in a single thing, not actual opamps. They just have the same pinout.

 

[pelopidas]

You know, seeing beautiful mods like that is not helping my obsession
Jon L said the nude teflons had a "creamy" sound. Did you also find this to be the case? What does it mean? Are there less details or are the details less sharp?
About the Espers song, you can find it here (if I am allowed to post links) http://www.youtube.com/watch?v=grEjtIdsras
at the 5:30 mark you will hear the guitars I was talking about.
 
So my SSG-3 silver micas showed up and, of course, I had to try them. But first I needed to mess around with my perfectly brilliant sounding system for no good reason.  Based on my input impedence being 20k I should be using 4+uf to be getting below 20 hrz without roll off. I have some 4.7 K73 on order but I am impatient, so I hooked up the 2.2 K75 with the 2.2 K73 to get 4.4uf. If the good part of this experiment is a fuller mid and low end (which I hope the 4.7 K73 will provide) the bad part of this experiment is that the K75 totally dominates and muffles all the sparkle and life out of the K73.   
 
So while the lower and midrange had (warning audiophool language) a luscious wetness and immediacy to it, it was under a blanket, where such terms belong. So I could not wait to try the SSG's. I removed the tabs and solderd the K73 to the SSG. The value of the silver micas is 0.142 uf.
 
My impression is similar to what Jon L said about them. I guess "bathed in soft white light" is the overall feeling, but it also seemed to constrict the soundstage compared to the K73 + K72 setup. Just overall narrower. The sound is not as sharply detailed as with the K72 but with the right music, ie vocal and instrumental such as Kings of Convenience it is just fantastically listenable, beautiful and soft in the shoulders. Then, switching genres to rock and electronic it just does not have what is needed. All that awesomeness from Dark side of the Moon was constricted and left me flat, unexcited. Thievery Corporation and Kruder and Dorfmeister all felt like the day after the excessive party.
 
It is very possible that the value of the SSG is too large and I do have smaller silver micas coming to try out, but after playing around with different songs and genres, this combo does not suit all music. I ended up adding a 0.01 vishay 1837 to it to give it more sparkle and basically found myself looking for the K73 + K72 sound again. So that is what I am going back to.
 
This should not be construed to think that the silver micas are not good. On the contrary. If I had a stereo just for instrumental and vocal, such as Arianna Savall, Kings of Convenience or soft, lovely music like Andrea Bocelli this would be my setup. For this genre it is like sitting in a velvet over-sized recliner after 2 glasses of wine. So "bathed in soft white light" is not from a 120w bulb but from a 40w bulb.

 

[dimkasta]

Quote:

You know, seeing beautiful mods like that is not helping my obsession Jon L said the nude teflons had a "creamy" sound. Did you also find this to be the case? What does it mean? Are there less details or are the details less sharp?

 
Wait till you see my latest mod which uses a wenge cylinder to house the cap
 
I wouldn't call it creamy. Rather less distorted. However please note that I use this cap not in coupling, but for the gain time constant of the tracking servo in my CD transport. Basically it allows the eye to follow the pit line with more accuracy.
I wanted to try this on my coupling caps as well, but with the 10uF oils total this is rather prohibitive. I never opened one to see if they are stable like that, but since oil slowly polymerizes with air I am a bit reluctant. Teflon is totally stable on the other hand...
 

 So my SSG-3 silver micas showed up and, of course, I had to try them. But first I needed to mess around with my perfectly brilliant sounding system for no good reason.  Based on my input impedence being 20k I should be using 4+uf to be getting below 20 hrz without roll off. I have some 4.7 K73 on order but I am impatient, so I hooked up the 2.2 K75 with the 2.2 K73 to get 4.4uf. If the good part of this experiment is a fuller mid and low end (which I hope the 4.7 K73 will provide) the bad part of this experiment is that the K75 totally dominates and muffles all the sparkle and life out of the K73.   
 
So while the lower and midrange had (warning audiophool language) a luscious wetness and immediacy to it, it was under a blanket, where such terms belong. So I could not wait to try the SSG's. I removed the tabs and solderd the K73 to the SSG. The value of the silver micas is 0.142 uf.

 
My experience is that for my speakers 10uFs are better. And yes I can confirm that the 2.2uF K75 is DARK and no amount of bypassing fixed that for me... I would keep those for PSU bypassing.
For bypassing, I would keep the ratio to about 1/100
Oh and I have not tried so big micas. I tend to like these more in phono stage RIAA network or local decoupling/filtering but in way smaller values.
 

 It is very possible that the value of the SSG is too large and I do have smaller silver micas coming to try out, but after playing around with different songs and genres, this combo does not suit all music. I ended up adding a 0.01 vishay 1837 to it to give it more sparkle and basically found myself looking for the K73 + K72 sound again. So that is what I am going back to.

It kind of is large. I would try something like 0.47 FT3 with 4.7uF main cap and a 1nF mica. Oh and keep the bypass connections in a star fashion with its center to the main cap, and use thick wire for them, so that the total looks more like one thing. This made a big difference for me, probably because I use too thin interconnect wire with resistance that created different filters with each cap.
 

  About the Espers song, you can find it here (if I am allowed to post links) http://www.youtube.com/watch?v=grEjtIdsras
at the 5:30 mark you will hear the guitars I was talking about.

 
This is an electric guitar, so it would make sense to hear it from the back and higher, since the sound comes from an amplifier and not the guitar itself

 

[pelopidas]

Well, the 4.7uf K73-16 arrived.
 
The good news is that it sounds just as clear as the 2.2uf but a bit "richer". What was interesting is that on this quest for audio bliss I have always found that each step towards better caps and sound I found more details in the upper frequencies. Air, sparkle, all that language one uses for a more revealing and intimate presentation. What surprised me this time was hearing more subtle and ambient sounds in the lower and mid range. These caps are so balanced and clear without any harshness that I am just listening to music without any desire to fiddle around with my system anymore. The details just immerse you in the recording. This one song that has sleigh bells used to be, well, you know, shh shh shh.. but now I can actually hear the little balls inside the bells rolling around. Or Dustin O'Halloran scooting around on his bench while playing piano, things like that which I had never heard before, even with headphones. And now I have this coming from my speakers. The soundstage feels really wide. Much wider than the actual location of the speakers. 
 
So at this point I have a serious DIY addict dilemma. I don't know if there is anything else "affordable" that I can do. Obviously I can spend the GDP of a small nation on getting better speakers but as far as this amp is concerned, I don't know that it can sound better. However, I also must admit that I have not tried the pricey Mundorfs or the exorbitant Duelands.
 
Dimkasta, did you receive your K73s yet? You mentioned that you had tried Mundorfs etc. I would really like to know what you think. How do these stack up to the best you have had including your current K75 + teflon + mica setup. Am I too easily pleased, or are the K73 really that good? 
 

 
 
here is a list of the Russian caps I have played with sofar. I have also placed a fuse along with a Vishay 1837 in the picture for size reference. For those who don't know, from top to bottom is as follows.
0.22uf FT-3
2.2uf K75-10
0.1uf K72-P6
0.142uf SSG
4.7uf K73-16
2.2uf K73-16
0.0024uf SGM

 

[dimkasta]

  So at this point I have a serious DIY addict dilemma. I don't know if there is anything else "affordable" that I can do. Obviously I can spend the GDP of a small nation on getting better speakers but as far as this amp is concerned, I don't know that it can sound better. However, I also must admit that I have not tried the pricey Mundorfs or the exorbitant Duelands.
 
Dimkasta, did you receive your K73s yet? You mentioned that you had tried Mundorfs etc. I would really like to know what you think. How do these stack up to the best you have had including your current K75 + teflon + mica setup. Am I too easily pleased, or are the K73 really that good? 

 
I have tried the plain Supremes at 2.2 at the output of my Phonoclone and VSPS phonos against K75 at 2.2uF bypassed by FT3 at 0.2uF (if I remember correctly)
 
The mundorfs sounded worse by a good margin. Rather dry and thin, even after lots of breaking in.
 
And I have tried the 0.47uF Silver-gold (not oil) against the 0.47 FT3 at the output of my DCB1 before my 220K amp, and the FT3s mopped the floor with the mundorfs.
 
and it kills me to say it because I would want to sell them at some point
 
My K73s are not here yet. The Russian fellow delayed the shipment, so I hope to get them in the next 10 days or so.