Can you show a drawing of that? I don't get how you connect the bias or why you have to add a DC bias.
Value will probably be around 1000/2200uF at most. Low voltage for size, like 5v to 10v probably. Probably not bypassed, or are there any values you would recommend for that? The UES types look interesting for that application, though maximum is 1000uF.
Have you tried Ciryl Bateman recommendation of two BP in series?
How is that? It's a differential power amp? Which is it?
Last edited by a moderator:
Not tight I Think.You mean it's not enough, less pronounced or not "tight"?
I find it pretty balanced...
All, watch out for using too high a value for your first filters. A high capacitor value reduces the conduction angle of the charging pulses and will generally make the noise problem a lot worse. You are best to leave replacement filter capacitors at or very near the original, designed in values. With anything except a power amplifier, I will often reduce the value of the filter capacitors. It is rare that designers use too small a filter capacitor these days.
-Chris
-Chris
Those gold/brown Silmics look like the ars version, not to be confused with the regular silmic II, which are usually rfs I believe. Hard to find now. For main filter caps in pre amps and dacs, I use Epcos B41505. Another that I am using in that same position, but have not compared yet to other parts is the Nichicon HE.
Another is the Nichicon VX, 100uf 35v that has worked very well to damp the output of an AMS1117 regulator, in a digital circuit. I had found these at radio shack way back, are discontinued however maybe still available on auction sites.
Another is the Nichicon VX, 100uf 35v that has worked very well to damp the output of an AMS1117 regulator, in a digital circuit. I had found these at radio shack way back, are discontinued however maybe still available on auction sites.
Last edited by a moderator:
Anatech is right about first cap thingy. Don’t know why many people have the conception of more farads the better. I always try to get away with using the least farad as possible.
You’ll hear that the sound is more natural with highs not being spiky or splashy especially on percussion instruments.
You’ll hear that the sound is more natural with highs not being spiky or splashy especially on percussion instruments.
Do caps have the same general sound signature when used on the power rail as they do when used for coupling? Or is it totally different? And is the signature as strong?
For instance, Gentlevoice, you say that polymers sound "up-tilted" and without weight in the bass. I found the opposite to be true when I tried the newer Panny OSCON for headphone coupling. They lacked air up high but had powerful bass. But I did get the same "smeared" transients you mention. These have a warm, buttery, fuzzy sound to me.
I am considering upgrading power smoothing caps on a DAC, and just wondering if any of my many experiments with electrolytics for coupling will also apply here.
For instance, Gentlevoice, you say that polymers sound "up-tilted" and without weight in the bass. I found the opposite to be true when I tried the newer Panny OSCON for headphone coupling. They lacked air up high but had powerful bass. But I did get the same "smeared" transients you mention. These have a warm, buttery, fuzzy sound to me.
I am considering upgrading power smoothing caps on a DAC, and just wondering if any of my many experiments with electrolytics for coupling will also apply here.
Found this site that made a roundup of "not low end" capacitors. They are all fairly expensive except the "Cornell Dubilier 940C", which starts around €2,50 on mouser and got a 8,5 score in his roundup (same subjective score as the Munford Evo Oil as a comparison). The Cornell Dubilier 940C series scored higher on this subjective scale then some higher end capacitors in this roundup. I am a novice or newbie, but the proposition looks like an interesting deal when you are on a tight budget and want the most bang4buck design you can get.
Mouser-list Cornell Dubilier 940C
Mouser-list Cornell Dubilier 940C
Last edited by a moderator:
Some other situations where bang for the buck can be had is with combining different parts to take advantage of the various characteristics. This is easier said than done however. A good and thrifty combo for coupling is to use a 10uf Nichicon Muse, and a .22uf Wima MKP10.
I’ve found coupling and power applications generally to be more different than similar for electrolytic caps.
I’ve found coupling and power applications generally to be more different than similar for electrolytic caps.
Last edited by a moderator:
Ahhh, the cardinal rule. Replacement parts must absolutely fit where the old component was originally mounted. I often see this "rule" ignored, bringing broken traces or drilled out plated through holes. Often the equipment will run for some time with reduced performance (due to noise pickup), and finally a broken connection. Equipment that had plated vias drilled out is often not working the first time it is powered up, and the cause can be difficult to locate. Time = money at the service centre. The thought of multi-layered PC boards being drilled out brings visions of instant junk. Modified to death.
-Chris
-Chris
Okay but why then in the datasheet they do not mention it? It is good that the Elna silmic are brown and white writing, it is not by chance.The data sheet and global if a special version would be released it would be mentioned
I recently discovered similar parts, had come on a dac board from years ago, and they sounded awful. I wasn’t sure if that was just the application or if they were fake, I have my suspicions however, since I recently tried them again in another circuit, still awful.I thought the really good ones are made for oems, not as likely to have a published or accessible data sheet.
Last edited by a moderator:
I'll agree and go further to state that there's no capacitor as good as "no capacitor". I'm designing a device now that uses DC servos instead of coupling caps. It's a heck of a lot more complex, but if you get it right, then the circuit works without coupling caps that can age and decay. DC servos can get a bad name if they're poorly designed, but if you use a pre-filter to roll off audio into the servo integrator, use a low distortion zero drift amplifier, and also use post attenuation to attenuate the servo noise, there's no reason to expect that a DC servo will work poorly. Yes, they might be expensive and slightly challenging to design, but compared to tweakazoid audiophile caps, they're inexpensive and compact, and have no wearout mechanism, unlike a wet Al coupling cap.
Servos are not able to be used for every application, but if you can solve DC problems with a quality servo, you'll be much better off than with any sort of coupling cap, film or electrolytic.
Well in ways your correct Monte but there’s no free lunch right. Lol
Caps are here to stay regardless. A simpler approach might be to clean upp as much as possible spike, noise etc at the ps caps then use a very small value cap for decoupling.
That way the influence of caps is minimized. Having decoupling caps of 100uf as on many circuits is not good anyway cause it also affects the transient response of the regs especially on analog side.
Caps are here to stay regardless. A simpler approach might be to clean upp as much as possible spike, noise etc at the ps caps then use a very small value cap for decoupling.
That way the influence of caps is minimized. Having decoupling caps of 100uf as on many circuits is not good anyway cause it also affects the transient response of the regs especially on analog side.