Tripath Input Coupling Caps

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
There was a thread a year and some months ago regarding input coupling caps in the T-Amps.

Original Coupling Capacitor Thread

I have been in contact with Michael Mardis regarding testing circuitry and I will be testing a number of capacitors over the next few months.

The circuitry will have stereo (two gang) switches on each channel. Therefore, the cap under test will be the only cap in the circuit. There will be no parasitic influences from other capacitors with this arrangement. All other capacitors will be electronically removed from the input circuitry. Using two sets of switches will provide the capability to compare one cap on channel A and another on channel B.

It is my intent to have the various capacitors listened to by musicians, music lovers, and audiophiles, but not anyone who really has an electronics background beyond myself will be listening to the caps under test.

I would appreciate any other consideration that comes to mind that may be employed as I undertake this task.

The proposed testbed will be a modified SI T-AMP with the forementioned switches in each channel. At some later time I will undergo the test with a 41HZ Amp6 and possiby a TA-10.1.

The Caps I currently have at hand are:

Obligatto Copper Case 2.2 MFD
Obligatto Paper In Oil (PIO) 2.0 MFD.
Auricap 2.2 MFD
Blackgate 2.2 MFD
Russian PIO 2.0 MFD

I have a number of Russian Teflon 0.1 MFD caps which I may use at some point to bypass PIO caps.
 
I just spent some time comparing caps in a similar fashion, but making a high gain differential measurement between two channels, each with a particular cap. It was not possible to get a good null across the audio band unless the caps were matched in value to a high degree. IMO, if one can hear anything, one can hear slight differences in the frequency response caused by mismatched caps. When the caps were matched, little difference was measured or heard between various film types. I don't have any Black Gates or other exotics, but no electrolytics could produce a good null, leaving some obvious hash in the audio band. I believed I could hear the difference as well. Thus, my advice is to only compare caps that measure close to identical in value, say 1-2%, and never try to compare a cap to a wire- they don't result in identical signals for obvious reasons. There may also be mismatches between the channels of your amp, unless you confirm them to match with a high gain differential measurement. Just comparing outputs on a scope isn't good enough. Without these precautions, the results of the listening tests lose a lot of meaning.
 
Davet, if your Russian caps are hermetically sealed, and if you send me some, I'll inject them with Fluorinert and send them back to you. I've been experimenting with Fluorinert lately and find it to have premium dielectric qualities. Teflon/Fluorinert caps are better than anything I've heard (I've pretty much bought/heard/measured everything out there).

And if you have any hermetic sealed Russian caps you care to get rid of, I'll take them off your hands!
 
Administrator
Joined 2004
Paid Member
Dave, are you going to build the burn circuit I sent?

If you don't burn them in, you know you will get a lot of "these caps don't sound good until after 500 hours" type of thing. ;)

Just want to insulate you from an obvious complaint. It might also allow you to compare a fresh cap with one that has some "miles" on it.

Looking forward to your findings!

BTW, I second Brian's vote for Panasonic FC. 2.2 or 3.3uF
And maybe some generic electrolytic and film cap.
 
Capacitor Burn in Circuit.

Michael, I plan to build and use your burn in circuit.

I don;t have the Panasonics on hand; unless they are the ones that came with the Amp6. At this point I do not plan to purchase all the various caps available. Maybe over time I will test a great number of them, but initially I will test those listed above. In addition I do have some 5 MFD Polycarbonates, Solens 3.3 MFD, and some Dayton films that I may try as well.

The Auricaps have been in my SI T-Amp for the past 5 months. I swapped out the Auricaps for the Obligatto PIOs today. The PIO seem to provide fuller bass and smoothed the highs a little. My initial impression was the PIO don't have the focus/definition that the Auricaps yield. The PIOs provide a broader soundstage, but the musicians do not seem to have the air around them of the Auricaps.

This initial impression is based on 3-hours of my listening.
 
Ex-Moderator
Joined 2005
Dave,

I can send you a bunch of 2.2uF Panasonic FCs and some of the 2.2uF MKT films caps. Send me an email with your address and I'll mail them to you.

My ears have led me to like no coupling caps the best - this wasn't with any Tripath amp though ;) With Tripath based amps I like the Auricap the best. The MKT was my next favorite.
 
serengetiplains said:
Davet, if your Russian caps are hermetically sealed, and if you send me some, I'll inject them with Fluorinert and send them back to you. I've been experimenting with Fluorinert lately and find it to have premium dielectric qualities. Teflon/Fluorinert caps are better than anything I've heard (I've pretty much bought/heard/measured everything out there).

And if you have any hermetic sealed Russian caps you care to get rid of, I'll take them off your hands!
Better than V-caps? Isn't Fluorinert some sort of coolling liquid?
 
theAnonymous1 said:
How about no caps. It's so easy to do I don't see why more people don't do it. I've done it on a few AMP3's and AMP32's and haven't had any problems.

The sound could be described as "capless".:clown:

http://i19.tinypic.com/4u27puc.jpg
http://i19.tinypic.com/52efm7n.jpg


Unfortunately, Doing this modification, In Some Cases, where the source equipment has, for some reason Not got O/P coupling-caps, and has a voltage-offset , then the tripath will amplify this and produce a large DC voltage offset at its O/P....

This could easily ruin a speaker or the tripath chip, blow-fuses etc...

Not a good idea to remove the coupling-caps at the I/P....
 
soongsc said:

Better than V-caps? Isn't Fluorinert some sort of coolling liquid?

Fluorinert is essentially liquid teflon with a very low dielectric constant (1.75) and probably correspondingly low DA, which is a function of dielectric constant. My paper/Fluorinert caps (despite the paper!) handily bettered VCaps---no contest. My Russian teflon/Fluorinert caps seem to me, after several days' listening, even that much better.
 
Administrator
Joined 2004
Paid Member
Remember those experiments a few years ago where they had rats breathing an oxygen bearing liquid?
That was Fluorinert, IIRC.

So what are you dong with the stuff? Just injecting it into caps? So that it mixes with the dielectric? Doesn't that change the capacitance?
 
It was indeed Fluorinert.

Yes, I drill a small hole in the hermetic seal and inject Fluorinert until the cap is full. Fluorinert, because a form of teflon, is evidently chemically compatible with teflon, as after being injected, slowly seeps into the tight teflon windings. After four days, capacitance increases 4-5%. The increase is evidently from displaced air, which can be heard, several days post injection, if the cap is shaken.

The end result is a PIO cap ... without the paper, which is a terrible dielectric, and without the oil, which is but a moderately good dielectric, while theoretically retaining all the advantages of a liquid dielectric (reduced microphonics? self-cancelling DA?) with the superiority of teflon. Match made in heaven.
 
Coupling Cap Shoot Out

I have received a number of capacitors inclusion in this test from contributors to this thread. Have built a HZ41 AMP6 Tripath based amplifier for the testing. In addition I have built a test bed just for the ability to switch between the capacitors.

Currently, I am burning in a series of capacitors based on a circuit provided by Michael Mardis. I plant to let the caps burn in for a minimum of 72 hours.

The first test I will conduct is to compare caps that have been burned in to the same cap that hasn't been burned in. I then plan to move onto comparing various capacitors against each other.

At some later point I will test various bypass caps on the coupling caps.

At this point I would like your input. The proposed test will be subjective for those not consumed by the hobby. Music of their choosing will be played through my system and I need criteria for them to rate in.

PRAT is not going to get it for them. I thought descriptors like lifelikeness; bass sounding like a acoustic bass; a piano sounding like it is in the room and proportioned to a piano; breadth of soundstage; air around performers, depth of soundstage; etc. Anything else you can think of that should be included please post or email me.

I will use terms measuring categories such as: really like: somewhat; like; dislike like somewhat; dislike; and hated it.

Once I get your imput and the test are underway I will start a new thread with the testing and results

TIA
 
Test Scenerio I

Scenerio: Compare Panasonic to Backgate electrolytics

I burned in a group of caps for at 336+ hours using a circuit provided by Panomaniac. I wanted to hear for myself if there was a audible difference between burned in (BI) and non-burned (NBI) in caps.

The first thing tested were Blackgate (BG) to Panasonic 3.3 MFD electrolytic caps. I compared (BI) BGs to (NBI) first and there is a definite difference in the sound of the BI BGs to the NBI BGs.

The BI BGs have a lot of body depth and width of soundstage compared to the NBI BGs. The NBI BGs sounded thin and the soundstage sounded compressed from front-to-to rear and from side-to-side.

I hear very little difference between the BI Panasonics and the BI BGs. The difference seems to be in the air around the performers. The BGs get the edge here.

The NBI Panasonics are tubby in the bass and the highs sound recessed.

Overall, the BI Panasonics and BGs are a wash to my ear. There is a subtle difference in the BGs, but for this application there is not a pronounced enough difference that would warrant the effort and expense, in my opinion, for the boutique cap.

Jan's inclusion of the Panasonics in the 41HZ kit is a winner once burned in.

Film-BG-AMP3-AMP6.jpg

The caps are the MKT films, BGs, AMP3, and AMP6 Panasonics.
 
Test Bed

The test bed for these test are as follows:

Input is from either a tweaked Samsung HD-841, or a Bobwire Tweaked Samsung 709 with a SuperPro Dac.

A 41HZ AMP6 T-AMP with a RS ALPS VR.

CAT5 wire is run from the input capacitor pads on the PCB to a 5 gang 11 position Russian Switch (Currently 4 gangs and 8 positions are being used.) Each switch position has caps for left and right channels.

CAT5 wire runs to European terminal strips. This allows for quick substitution of caps in the test bed. Current capacity is for 8 different caps that may be compared at the same time.

Only the capacitors under test are in the circuit at any time. The caps do not share any wiring aside from the CAT5 wire that connects the switch to the PCB.

Note: I have to use the mute switch(sleep) on the AMP6 to switch between caps. There is a horrendous click that comes from the speakers if I try to switch directly.

I will post a photo of the test bed and AMP6 in a subsequent post.
 
Administrator
Joined 2004
Paid Member
Good work, Dave!

Nice to know that you did find a difference in the burned in caps. Over 300 hours is a long time. I don't know of any other AB test of BI vs NBI caps. Well done.

There also seems to be a change in sound with butn in of power supply caps, too.

Looking forward to reading your opinions of the other caps.
 
Russian Caps

At this point I don't have a definitive answer for you, but this is one of the test that is planned. I have a number of PIOs which I will A/B with .01 flourinert .01 caps, regular Russian .01 caps, .01 Russian PIO caps, and generic Radio Shack film caps. However, there will be some time before I get to it. I have a lot of caps to test prior to getting to bypassing.

Outside of formal testing I have used the Russian .01 mfd caps for bypassing on any number of caps and I have always heard a positive difference.

I would suggest you give it a shot and let us know what you hear.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.