Capacitors, where's the beef belong?

Ok, there's an immense amount of discussion and opinion about the role caps play in an amp/receiver (mine's the SX 950), but what section or what board are good quality caps a necessity? I guess the question is really where's the area good caps can have the most effect on the quality of the sound -- clarity, purity, color, etc? I'm mostly a spectator in this great hobby and have done cap replacements on a phono board only -- very rewarding!
I seem to remeber from other threads that certain caps are preferred in the power section while other areas of the signal path require (or it's worth it) to use higher end caps. Could anyone explain this in very basic terms? Many thanks in advance.:scratch2:

Power supply caps will filter AC making the DC smoother (less ripple) or (hum) and signal path caps will couple or pass the signal from one stage to another.
Thats just the basics , caps will pass AC and block DC (store)voltages.
Capacitors can build up voltages too by doing this..
A charged capacitor will hold it charge until it is discharged.
Hopefully if needed someone else can explain it better .
Replace with same or higher voltage rating and go with a higher temp rating ( 105c ) and a lower ESR when available IMHO.



Barney

The main thing here is using replacement caps with the most linear charge-discharge characteristic attainable as well as total discharge without retaining a residual charge which would affect linearity as well. Some in this group will be able to recommend the best of the signal electrolytic caps appropriate for these locations in the signal paths where the quality of the caps is important.

Low/no leakage is a 'given' as well. Usually this means polystyrene, polypropylene, polyethylene, and teflon film caps for the normal signal path values in tube equipment and some locations in solid state gear, and the best electrolytics obtainable otherwise. Modern large power supply size electrolytic caps are pretty good in linearity tho' good general practice is to also bypass the main power rails filtered/resevoired by these larger e'caps with film caps to extend their HF bypass capability. Some will also bypass the other e'lytics with the higrade film caps .

All caps in the signal paths can/will make a difference, remembering that the power rail bypass caps are the return path for the internal signal path. Whether or not it's worth the bother of upgrading the stock already-working in-spec caps depends on the source material, interconnects, and other gear including especially the competence of the speakers. Nothing can fix a lousy speaker system except replacement with a good well-designed and constructed one with good parts.


Some further thoughts concerning the inherent limits of performance in a given unit/system;

Something that is commonly overlooked is the signal wiring itself, inside as well as outside the unit which also has a profound effect on the sound. The most important thing here is that the signal return is NOT taken thru a shield, rather a separate path with the shield grounded at one end only. This is poor practice. The resulting signature sound is that "electronic/mechanical processed" sheen/glare that is prevalent with a lot of this period gear predating ribbon-cable internal signal path wiring, and the 'gimme" coax line interconects that came with most gear are the kiss of death for achieving a "hi-end" like natural sound in an otherwise good properly functioning system.
These points are not debatable, I have researched and demonstrated this for 25 years and the difference between the 'gimme' coax cables and the simplest replacement interconnects made from lamp cord terminated with the cheapest RCA plugs obtainable is readily audible as an improvement in naturalness and dynamic response. I don't necessarily recommend these as a permanent replacement, just a demo. Better would be the same plugs with twisted-pair mic cable with the shield grounded at the downstream end only so that it cannot carry any of the signal.

The point is to eliminate all it, inside as well as out, from an otherwise well sorted system for best results and a profound improvement in performance, dynamic as well as tambre, with what one has and without changing anything else.


It comes down to potential and expectations vs doing a restoration vs doing a modified unit.
I guess if I wanted to achieve full-out High-End sound, I'd start with something that was designed to aspire to it without the pitfalls requiring extensive reworking. Sadly my SX-1980 doesn't qualify without mod'ing it's signal wiring. Modified, nobody knows how good it could sound in a good system otherwise.
S.B.

Oerets said:

Replace with same or higher voltage rating and go with a higher temp rating ( 105c ) and a lower ESR when available IMHO.
Barney

Click to expand...

I'm working on my TX-9100 and ordering up some new caps as I found a burst one. While I have the AWE-026 Tuner board out, I figured I'd replace some of the caps as they are relatively cheap.

I can't find a 6.3V 470uf cap though on Mouser, so I'm guessing I'd be OK with a 10V version? It's the closest I find.

Just for conversation, I see that there are quite a few different voltages on many of these caps, so i'm a bit curious as to if it makes much/any difference with replacing them with the same voltage or a higher one; given the cost is approximately the same, i.e. a 16 V for a 10V or a 50V for a 35V.
Does an increased voltage rating have any real effect on the circuit other than overkill?

I ask because somewhere along the line someone hinted that the increase voltage capacity could have an effect in the circuit, but I'm not sure how or why...

As far as the signal path goes, you need to replace them all with good quality caps - the quality will only sound as good as the worst cap - 'weakest link in the fence' as it were. So it wouldn't make much sense to recap the preamp but not the power amp, or vice versa.

Regarding the voltage question - you are fine replacing the 6.3 with a 10. Generally you are fine upping the voltage within reason. I wouldn't replace that 6.3 with a 100 volt cap but then you're not likely to fit that one the PCB either. Generally for the signal path caps I stock them in 50 volt / 105C values. I replace plenty of 16 volt caps with 50 volt caps. I don't think it will make any difference good or bad.
I would definitely make a point of replacing all of the tantalum caps (they look like red/blue/orange epoxy coated teardrops) with electrolytics. The mylar caps you should leave alone.
JimB

JimB

You MAY want to read the tuning fork guides on capacitors:

http://www.audiokarma.org/forums/showpost.php?p=3012505&postcount=11

I recently worked on a tx-9100, and there are a few spots where solid aluminum or tantalum caps are used, and while they MAY have been used for small size, they ALSO could have been used for their low leakage characteristics.

I didn't think it would make much difference until I had my eyes opened when working on a SA-9500 amp, I used ALL UPW's in it, and there was a bit of noise in the preamp section, perhaps more than when I started.... I then replaced all caps that had previously been low leakage "ceanl" caps, tantalum and solid aluminum caps with Nichicon UKL (85 c unfortunately) low leakage caps, and it DID make a noticable/measureable decrease in the noise floor. It went to DEAD QUIET with the inputs shorted.

As the tuning fork guides attest, the CEANL and solid aluminum caps (powder blue usually) have a shiny epoxy seal visible around the leads as opposed to a visible rubber seal, while the Tantalum caps are in a very distinctive class of their own in appearance.

This in the TX-9100 will be more apparent as the S/N (decreased noise) ratio in the decoded stereo signal.

I only spoke up because of my direct experience on the 9100.

OH, and any caps between 0.1uf and 1uf or so I replace with metalized poly stacked film caps, their leakage is comparable or better than the UKL's, and handle audio perhaps even better, and I make EVERY EFFORT to shoehorn them in.

you now have one of my "biggest secrets".....

....I dont mean to derail the thread but I just got done posting a question under the "vintage Solid State" forum that asks what is an "Alisicon" cap? All I can tell is it's made by Sanyo and in the description it's called a 'sintered aluminum anode solid" something or other. Mark, is this the same as what you are calling a "solid aluminum" cap? They are common in Fisher RS 10xx and i saw somewhere they were used in a Rotel receiver. Never heard of it.

day67 said:

....I dont mean to derail the thread but I just got done posting a question under the "vintage Solid State" forum that asks what is an "Alisicon" cap? All I can tell is it's made by Sanyo and in the description it's called a 'sintered aluminum anode solid" something or other. Mark, is this the same as what you are calling a "solid aluminum" cap? They are common in Fisher RS 10xx and i saw somewhere they were used in a Rotel receiver. Never heard of it.

Click to expand...

More than likely, Yes.

Solid aluminum caps are NOT wrapped foils with wet "cardboard" and aluminum foil layers in between, if you were to cut one in half you would see the sintered construction that makes the metal like a microscopic sponge, it would look like a dull solid.

Read the tuning fork guides, wikipedia for more information.

Like tantalum capacitors they are very intolerant of adverse voltage conditions (reversed polarity, overvoltage, overcurrent & slew rate), and their failure modes (hard short circuit) destructive (explosive).

markthefixer said:

I didn't think it would make much difference until I had my eyes opened when working on a SA-9500 amp, I used ALL UPW's in it, and there was a bit of noise in the preamp section, perhaps more than when I started.... I then replaced all caps that had previously been low leakage "ceanl" caps, tantalum and solid aluminum caps with Nichicon UKL (85 c unfortunately) low leakage caps, and it DID make a noticable/measureable decrease in the noise floor. It went to DEAD QUIET with the inputs shorted.

Click to expand...

I feel validated now for spending a few extra pennies for Nichicon KLs to replace low leakage caps and tantalums. I was starting to believe that maybe I was just overthinking the whole thing.:scratch2: