Replacing caps in a mini USB DAC

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Disabled Account
Joined 2002
You were obviously an early starter. At 12 or 13 I was still learning Ohm's Law and other physics relating to electronics. Some people try to skip this.

Using the wrong parts can certainly make a difference. Swapping one appropriate part for another appropriate part can usually only be heard when the listener knows that a swap has taken place.

Truth is I started at 8 or 9, parents differ in opinion on that. Ohm Law also then was relevant. Using the right parts also makes a difference. I think anyone can swap coupling caps with different dielectrics and let you hear the difference. We tested this by also tricking the listeners ("OK now we now swap to another capacitor pair" but did not change a thing) and results were consistent. Then it were phono preamplifiers. IMO you really go around the phenomenon by not trying, it is fun to do. many have observed the same, it is not voodoo. Parts simply do not sound the same. On the contrary, it would be very surprising if different materials would sound the same. In analog circuits it was observed easily. You simply ignore something that is quite known under many audio enthousiasts. Anyone that has replaced bipolar caps on loudspeaker filters for decent film types knows they sound different. I admit many exaggerate in the extreme and make a lifetime hobby of changing parts to boutique stuff. Also cables do sound different but for me differences are too small to put time in it. Case material steel or aluminium even changes sound, wood sounds best in many cases but the lack of shielding makes it difficult to use. We even had the same PCBs made from the then popular pertinax and the more expensive epoxy and even boards had an influence. All parts were the same. Some things are hard to explain (I stopped trying) but that does not make the results useless. It is a strange area and one can hide behind formula but there is more to audio than calculating and measuring. There will be a scientific explanation for it but I don't have it. I just hear it and simply choose what sounds best. We had a shop here that sold surplus stuff and MKT,MKC and MKP caps were very cheap there. You simply filled a bag with the stuff you wanted to try and came home with all brands and types of caps.

Nowadays with FDA it does not make much sense anymore to roll caps except when they're unbranded chinese knockoffs that will fail anyway.

Amplifiers which measure the same (including frequency response and output impedance) appear to be indistinguishable when used unsighted below clipping. I have not done the experiments myself, but others have. I have no reason at present to doubt these results.

Further discussion is useless I see :)
 
Last edited:
Very few people on here have done listening tests which are sufficiently well structured to give meaningful results. The difference between me and those who have done some listening tests is that I know that the results of most tests are meaningless; I have better things to do with my time. In addition, my wife is rarely in the kitchen and she has no interest in audio - so another common audiophile test is unavailable to me.

jean-paul said:
Anyone that has replaced bipolar caps on loudspeaker filters for decent film types knows they sound different.
I accept that changing caps in filters used with strong signals may be audible - physics says so! Changing caps used as couplers with small signals is a completely different issue.
 
Disabled Account
Joined 2002
The difference between me and those who have done some listening tests is that I know that the results of most tests are meaningless

Audio is not a religion. There are beliefs and there are tests. Knowing that something is meaningless without even having tried is eh... well forget it. One can't know without trying out. You can not dismiss tests because you think they aren't enough well structured without even being there at the time. These are simply assumptions to avoid doing something yourself. Such tests are so simple that even I can show them to you. I guess you have never built 2 exactly the same devices but one with (for instance) Wima MKT caps and the other with Wima MKP caps ? Even the most hardened engineer with all the formulas and stubborn non believer attitude will admit they sound different. I have joined numerous tests of others and have put many years in gaining experience by putting a lot of time in "meaningless" tests. I know an open mind is the only way to accept matters. Not willing to put time in it ? Fair enough. I have put much time in it and still am baffled by some matters that are showed to me.

Some years ago I designed a DAC with another member. I was the one that put many hours in choosing parts. Designed over 50 revisions of the PCB. Built several tens of them myself. Still I got feedback from builders to try out part X. I did and changing just one cap made the DAC sound different. The DAC design is proven to be sturdy as hundreds of them are still being used. The PCB design is good, the PSU was extremely good (measured). It was very well decoupled but choice of caps still made a difference. I have kept a few and recently sold some of the last excess ones. Last week I connected 3 all built with same parts except electrolytic caps and some with film caps where the others had polymer caps. They all sounded different. Not by a large margin but it was easily noticed.

It is so basic and simple to try out that dismissing this phenomenon is ostrich policy.

False reasoning kills literally every discussion:

The difference between me and those who haven't done some listening tests is that I know that the results of most tests are meaningful
 
Last edited:
jean-paul said:
I guess you have never built 2 exactly the same devices but one with (for instance) Wima MKT caps and the other with Wima MKP caps ?
No, why would I want to do that? Polyester and polypropylene are different dielectrics so I would choose the appropriate one for the task; polyester for uncritical locations (which may include some coupling capacitors) and polypropylene where there might be significant signal voltage across the cap and the larger physical size would not create problems (such as unwanted stray capacitance). I would expect that swapping to an inappropriate one would degrade circuit behaviour, unless the application was such that no difference would be noticed.

In mixed signal circuits the details of decoupling matter. The cap dielectric is probably less important, except that it also sets ESR and physical size (and hence inductance). Given that many digital circuits have to cope with some degree of ground bounce it is not too surprising that a change of cap might make a difference. However, this would mean that the original cap was inappropriate - which is what I said right at the beginning. On the other hand, simply swapping from one random cap to another random cap (or a 'recipe' cap) is as likely to degrade things as improve them. A full simulation (possible EM as well as circuit) might be needed to correctly choose an alternative cap.

I am not an ostrich. I don't design (or modify) mixed-signal circuits because I know I don't know enough to do it properly; maybe knowledge of my own ignorance sets me apart from some others? I just stick to audio and RF.
 
Hi Everyone,
I have to put my two cents in to this topic. In my opinion manufacturers are saving their money with using as cheap parts as they can. Another situation is very high end stuff. I don't have any problem to hear the difference with different type of capacitors in dac, amplifier, loudspeaker. I always try to replace electrolytic caps with film ones. I also fully agree with jean-paul that everything plays role in sound. Resistors, cabinets/enclosure materials, wires, simply everything. Some of them are more to be heard, some less. But everything is important. This is obvious that You have to have good equipment to hear any differences. If not, making any changes is meaningless.

Best Regards,
Daniel
 
Administrator
Joined 2004
Paid Member
<snip>In addition, my wife is rarely in the kitchen and she has no interest in audio - so another common audiophile test is unavailable to me. <snip>

Damn, I just had to say this is EXACTLY the problem I have.. I will confess to having a preferred dielectric, and it's not paper, but since I am a tube guy it's probably irrelevant.. ;)
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.