I respectfully disagree - the issue he discusses has a significant bearing on how capacitors sound. The fact that he objectively demonstrates measurable non-linearity in capacitors of various dielectrics indicates that these capacitors are all adding significant distortion, and hence coloration to the signal. The only part that isn't really relevant to current hifi usage is the section on ceramic coupling caps and that is still useful to the guitar amp guys, and is a good demonstration of why you should avoid them in the signal path of hifi amps.
KevinKr,
I agree with your comments,
As stated in the thread, DC coupling does "sound better" than any capacitor. The less components in the signal path the better. As long as low frequency saturation is not a problem.
Interesting that the cost does not always mean "better" from a component point of view. Non inductive comes to mind. Dielectric type and conductor material.
Its easy to use metals like silver for wire tests, just go to the local jewelry suppliers and buy silver wire sleeve with PTFE any impurities tend to be copper. use the correct solder. Only problem is it seems to show any other weaknesses in the system! Good fun though.
Just for interest I tried replacing caps in the tone control on an Epiphone Les Paul to disk ceramic because they are used in the Gibson. Guess what it made it sound very similar.
Where do you get your: Quote :The Russian FT-3 is good and not too expensive a place to start. I use TFTs and Vcaps mostly
Regards
M. Gregg
I agree with your comments,
As stated in the thread, DC coupling does "sound better" than any capacitor. The less components in the signal path the better. As long as low frequency saturation is not a problem.
Interesting that the cost does not always mean "better" from a component point of view. Non inductive comes to mind. Dielectric type and conductor material.
Its easy to use metals like silver for wire tests, just go to the local jewelry suppliers and buy silver wire sleeve with PTFE any impurities tend to be copper. use the correct solder. Only problem is it seems to show any other weaknesses in the system! Good fun though.
Just for interest I tried replacing caps in the tone control on an Epiphone Les Paul to disk ceramic because they are used in the Gibson. Guess what it made it sound very similar.
Where do you get your: Quote :The Russian FT-3 is good and not too expensive a place to start. I use TFTs and Vcaps mostly
Regards
M. Gregg
Last edited:
The author demonstrated some non linearity in D-E curve. The photos taken from oscilloscope screen do not have any scale. What can be analysed base on those remains unclear.
Also I can not find from the article any straight and clear connection between audible distortion in an amplifier and the described unlinearity at D-E curves. Completely different case is if one believes that these has someting in common.
SB called out the voltage swings in the text- something like 70V across the caps, if memory serves, stating that this much voltage was needed to be able to readily measure the effects. Now, that's not exactly typical usage with respect to coupling caps, where the voltage across the caps in the passband is very, very small.
No, not typical in the passband, however at the LF corner the AC voltage across the dielectric is increasing as is the resulting distortion. This could be a significant issue in stages driving low mu power triodes in the presence of significant levels of subsonic energy. (LP playback for instance, and certain types of electronic music) I actually don't know how voltage dependent these effects are wrt to performance of the dielectric - however as a point of comparison under controlled conditions the experiment is valid if not completely documented. This is very relevant to passive cross-over design where the AC voltages across capacitors can be large.
I still think any electrolytic in the signal path is a more significant issue if a reasonably good coupling capacitor is being discussed.
Some of us here think that all capacitors at the signal path are harmful with respect to sound quality of an amplifier.
Should we then concentrate to construct dc-coupled tube amplifiers only ?
Would such amplifers sound better ?
How many capacitors have the music we play already passed thru before it reach our tube amplifier and its last two to three capacitors - those that are claimed to create distortion to the sound ?
What about these earlier "some 50 pcs." during the chain from the instruments via the studio etc. etc. and finally out from our CD player ? Don't these have the same effect ?
If they had, why should we then concentrate to get rid of these last few capacitors ?
I fully agree that zero-biased tantalum or electrolytics are not the best capacitors in the signal path.
Should we then concentrate to construct dc-coupled tube amplifiers only ?
Would such amplifers sound better ?
How many capacitors have the music we play already passed thru before it reach our tube amplifier and its last two to three capacitors - those that are claimed to create distortion to the sound ?
What about these earlier "some 50 pcs." during the chain from the instruments via the studio etc. etc. and finally out from our CD player ? Don't these have the same effect ?
If they had, why should we then concentrate to get rid of these last few capacitors ?
I fully agree that zero-biased tantalum or electrolytics are not the best capacitors in the signal path.
Quite true, I'm not saying this is a realistic test..
It does serve to illustrate the effect, and some of these things can be measured at levels that can reasonably be expected in driver applications. Low mu DHTs may require swings of 300vpp and into 10K this would be 30mApp. The 50 dht is a perfect example as 10K is the maximum permissible grid circuit resistance - if not IT coupled will require grid drive approaching 17mApp under the stated conditions for full output. In some cases a grid choke will be used that obviously substantially reduces the current drive requirement, but I have also seen RC coupled 50 amps in the past.
In a passive X-O the currents can be far higher, usually the voltages are much lower except in pro applications where these sorts of voltages may be seen at very high currents. Bi or tri-amping makes a lot of sense in such applications, but isn't always done.
FWIW I more or less completely agree with everything you said in the above quoted post.
I have quite a few high quality film caps in the signal path of my electronics along with quite a few more transformers than are typical in most modern systems. Vintage recordings were recorded, mixed down and mastered on equipment full of transformers, tubes, and paper coupling caps. Newer recordings have some of the above in addition to sometimes mediocre op-amps and other less delectable items in the signal path. Even worse I mostly listen to vinyl these days using a table, arm and cartridge all designed more than 40 years ago.. 
What's the use of diyaudio anyway if you look at it that way?
I myself don't care whatever happened in the studio etc. The part of the chain I have influence on is what I try to get as good as possible.
To avoid making it worse - you don't know how it would have been without those caps.
Avoid them, and if possible, reduce the charge density as much as possible (i.e. work with high impedances).
I'm not here to get stuck in this kind of endless discussions
I stated earlier that getting rid of one cap improved things in my opinion.
If you call that 'belief' that's fine, I'm not a machine.
I have found similar improvement changing Output op amps on DAC's IE NE5534 to OP27. Even the make of the same Op amp has shown a difference in the same circuit. I think we have to accept the best source we can afford to buy or build and then say we must try to preserve the input to the amp in tact.(with as little addition or distortion as possible).
The next task for us is to raise the level of the source to a power that can drive our speakers in the best possible way.
Or with the most enjoyable presentation possible with the limitations imposed on us (money-component cost)
Then make it so good that as we improve the source the amp can still out perform it. (That is the hard part). We have to "forget" the source and say what ever goes in must be exactly what comes out nothing added or taken away!<<that is even harder if not impossible!
Regards
M. Gregg
If you are sure that the sound quality changes, then there are very potential chance that some measurable difference can be found with other "tools" than human ear too.
I sincerely strongly recommend you to go deeper into this observation.
If you found something, this would be very valuable and real science.
Just a thought,
I think DC coupling is sonically better, however it is not as much "fun" from a build point of view. With AC coupling each stage is "protected" from the failure of the tube in the previous stage by the capacitor. So measures would have to be put in to protect OP tubes and Tx's.
I like to think of topologies in categories:
LoFi, HiFi, SuperFi.
LoFi quick projects that are fun to build and give good results.
Hifi can be great fun and give real satisfaction to the builder. Built to good designs and using a few esoteric ideas.
Superfi can be a pain in the A** constant tweaking and weak sources showing their faults. However it can be very interesting. The trouble is you always think it could be better and each improvement is harder to get than the last.(cutting edge comes to mind).
So I think there will always be a need for good designs with AC coupling caps! That is why they are so popular now.
Regards
M. Gregg
I think DC coupling is sonically better, however it is not as much "fun" from a build point of view. With AC coupling each stage is "protected" from the failure of the tube in the previous stage by the capacitor. So measures would have to be put in to protect OP tubes and Tx's.
I like to think of topologies in categories:
LoFi, HiFi, SuperFi.
LoFi quick projects that are fun to build and give good results.
Hifi can be great fun and give real satisfaction to the builder. Built to good designs and using a few esoteric ideas.
Superfi can be a pain in the A** constant tweaking and weak sources showing their faults. However it can be very interesting. The trouble is you always think it could be better and each improvement is harder to get than the last.(cutting edge comes to mind).
So I think there will always be a need for good designs with AC coupling caps! That is why they are so popular now.
Regards
M. Gregg
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.