[Ears]

I am changing my ps caps in my amp to brighten up the midrange and highs a bit.

The current Jamicon 10k 100v caps [4] are getting swapped for something brighter.
I had decided on Blackgates but found them to cost 900-1k plus for four caps of the same value.
www.musicalconcepts.com has been modding amps for decades and they talk of this exact cap value to put there 25k custom made caps to shame and say that these blackgates are worth every penny and generally talk about them as the second coming.

If someone knows were i could get my hands on these for lets say 5-600.00, i would appreciate a heads up.

Any alternatives to these or ideas on what is close to BG sonics would be appreciated.
Also, does anyone know if bypassing the current caps, would net the same results as installing the  10k 100v BG's?

[audioengr]

If you want to save money, replace all of the other caps in the amp with Black Gates and put .1UFd Polypropylenes in place of the ceramics etc..  Most of the time, this will improve the dynamics sufficiently.  USe "FK" versions of the Black Gates.  Also parallel .1 caps with the black gates where there is no HF decoupling already.

[randytsuch]

Hi,
I was going to suggest with bypassing the big caps with something like 470uf blackgates.  A modded suggested I do this to an amp that he had modded for me, to further improve it.

An alternative is to use Nichicons instead of gates.  The audio Nichicons are supposed to be very good, Michael Percy sells them.

I use smaller gates, but was never willing to shell out the money for those big gates.

Randy

[Curt]

If you want to save money, replace all of the other caps in the amp with Black Gates and put .1UFd Polypropylenes in place of the ceramics etc..  Most of the time, this will improve the dynamics sufficiently.  USe "FK" versions of the Black Gates.  Also parallel .1 caps with the black gates where there is no HF decoupling already.

Replacing ceramic bypass caps (for noise control) with PP caps is a common audiophile mistake.  To get the lowest noise floor possible use good quality X7R ceramic bypass caps, PP caps have no damping and ring like a bell when trying to perform this task in a circuit. You can wind up with all kinds or resonance issues.

[audioengr]

[/quote]I was going to suggest with bypassing the big caps with something like 470uf blackgates. A modded suggested I do this to an amp that he had modded for me, to further improve it.

An alternative is to use Nichicons instead of gates. The audio Nichicons are supposed to be very good

The inductance between the big caps and the load will make the 470's ineffective.  Better off to put these right at the output transistors.

As for Nichicons, they are not nearly as effective as Black Gates.  Okay to use in the early sections (right after the bridge rectifier, but not after the voltage regulators in a preamp for instance).

[audioengr]

Replacing ceramic bypass caps (for noise control) with PP caps is a common audiophile mistake. To get the lowest noise floor possible use good quality X7R ceramic bypass caps, PP caps have no damping and ring like a bell when trying to perform this task in a circuit. You can wind up with all kinds or resonance issues.

This depends entirely on the circuit topology.  If there is very little inductance in the cap and on the circuit board, there should be no ringing.  I have been very successful in using these.  I believe in using High-Q caps wherever possible, but I rewire the entire power delivery system as well to lower inductance.  BTW - I do not use them for noise control as I interpret this term.  I typically remove all such caps when I do a mod.  I use them for HF decoupling.

[randytsuch]

The inductance between the big caps and the load will make the 470's ineffective. Better off to put these right at the output transistors.

I am not sure, as I never tried it, but wouldn't the inductance vary depending on the amp in question?

The inductance between the big caps and As for Nichicons, they are not nearly as effective as Black Gates. Okay to use in the early sections (right after the bridge rectifier, but not after the voltage regulators in a preamp for instance).

Based on the values he was talking about, I figured these were the big filter caps that would be right after the bridge, in an amp.  I was not recommending the Nichicons in any other application, but in an amp it costs a small fortune to use gates here.

Randy

[Curt]

This depends entirely on the circuit topology.  If there is very little inductance in the cap and on the circuit board, there should be no ringing.  I have been very successful in using these.  I believe in using High-Q caps wherever possible, but I rewire the entire power delivery system as well to lower inductance.  BTW - I do not use them for noise control as I interpret this term.  I typically remove all such caps when I do a mod.  I use them for HF decoupling.

I see now your only talking about the power supply, but let me comment on the noise issue for a moment.

I find noise control in all circuits (digital and analog) very similar. The amount of noise and causes of the noise can vary, but the control methods are all about the same.

HF decoupling, isn't that getting rid of HF spikes, hash, RF, certain IC artifacts? The stuff that looks just like noise

There is always inductance, in the cap, in the component leads, and in the PCB traces. The power supply rails between active devices becomes a low-loss inductor as well. With any type of repetitive signal or pulse the whole PCB system can become excited as a complex resonator. When I sweep circuits while watching on the spectrum analyzer you can see some of this at different frequencies.

The little bit of ESR in X7F ceramics (0.1 ohms) along with a few tants (1 ohm)  here and there help lower the Q and help lower the possibility of resonance - ringing issues. Some designers even add a small value resistor (1-4.7 ohm) in series with a cap for damping problem areas on a PCB.

The only reason I replied on this thread is because several of our customers have talked about swapping all the 0.1uF or 0.01uF ceramic bypass caps in their audio components (every cap) for film caps thinking that because film is better for the signal path it must be better for everything, and it's  not. They wind up making their audio component worse instead of better.

Your just talking about power supplies here. For a high-power audio power supply, yes, low ESR film bypass caps (like PPs) bypassing the main filter caps can sometimes speed up the current delivery and sometimes improve the audio. This is a common practice and works well.

I still would not remove the original bypass caps (put in place for noise control), but just add the PPs where needed.  I like to put them directly across the large filter caps using the shortest leads possible.

[Ears]

So use a cap such as Aurri caps to bypass the  board mounted 10k 100v ps caps, with the shortest possible leads.

Is this what your suggesting?

[audioengr]

I don't think Auricaps are a reasonable option.  The leads are too long and inductive.  They are fine for coupling and speaker crossovers, but not for power decoupling.  You need very short leads, usually radial designs for this.  I like to solder them directly to the component legs when possible.

BTW - I much prefer polystyrenes to polypropylenes for anything in the signal path.

[Al Garay]

Are AudioCap Theta's suitable for power supply?

Al

[DVV]

Curt said:

I find noise control in all circuits (digital and analog) very similar. The amount of noise and causes of the noise can vary, but the control methods are all about the same.

I agree. In both cases, it starts from a decent line filter. In 34 years, I haven't seen a power supply yet which wasn't offloaded that way and when noise didn't drop quite substantially (depending on the line filter quality).

HF decoupling, isn't that getting rid of HF spikes, hash, RF, certain IC artifacts? The stuff that looks just like noise  

That's it Curt.

There is always inductance, in the cap, in the component leads, and in the PCB traces. The power supply rails between active devices becomes a low-loss inductor as well. With any type of repetitive signal or pulse the whole PCB system can become excited as a complex resonator. When I sweep circuits while watching on the spectrum analyzer you can see some of this at different frequencies.

The little bit of ESR in X7F ceramics (0.1 ohms) along with a few tants (1 ohm) here and there help lower the Q and help lower the possibility of resonance - ringing issues. Some designers even add a small value resistor (1-4.7 ohm) in series with a cap for damping problem areas on a PCB.

The only reason I replied on this thread is because several of our customers have talked about swapping all the 0.1uF or 0.01uF ceramic bypass caps in their audio components (every cap) for film caps thinking that because film is better for the signal path it must be better for everything, and it's not. They wind up making their audio component worse instead of better.

How true! And worse, some audio companies promote this by using mammoth film caps for decoupling, I suspect because they look good on the PCB, impressive, and you can always point at them and say - wow, lokk at that, theres guys are not into savings!

Your just talking about power supplies here. For a high-power audio power supply, yes, low ESR film bypass caps (like PPs) bypassing the main filter caps can sometimes speed up the current delivery and sometimes improve the audio. This is a common practice and works well.

I still would not remove the original bypass caps (put in place for noise control), but just add the PPs where needed. I like to put them directly across the large filter caps using the shortest leads possible.

I second this - this is good engineering practice combined with using brains rather than famous brand name products alone.

Cheers,
DVV

[audioengr]

Are AudioCap Theta's suitable for power supply?

The smaller values could certainly be used, such as the .01 and .1.  When they get to be 3/4"-1" long, they are not as useful.

[jesserparker]

bypassing the ps caps is something i have been meaning to get around to for a while now (i was thinking black gate, but opinions for that would also be appreciated), but i wonder how i should go about deciding what value caps to use to bypass???

should i just always use a very small (.1 ?) value, or is there some way to determine what value will be best for different size caps?

thanks...

[DVV]

bypassing the ps caps is something i have been meaning to get around to for a while now (i was thinking black gate, but opinions for that would also be appreciated), but i wonder how i should go about deciding what value caps to use to bypass???

should i just always use a very small (.1 ?) value, or is there some way to determine what value will be best for different size caps?

thanks...

Jesse, you can't miss with 0.1uF. However, that said, it should be obvious that while they may not be wrong, there's no guarantee that they are the optimal value either. Optimal could be 0.15, 0.22, 0.33 uF or some such, it all depends on where they are and what are the characteristics of the cap you are bypassing.

If you have the time and the nerves for it, I would suggest you experiment. Use 0.1, then 0.15, 0.22 and 0.33 uF, and listen. If you hear any differences, you're home and dry, because you will know which sounds better to you.

Cheers,
DVV

[WerTicus]

Ohh this topic is really interesting.....

i have found that adding more and more power filter caps of the same value in paralel with the existing ones - certainly same voltage or more. ...

gives you substantial increases in quality... and naturally capacitance.

[WerTicus]

Why dont we use an actual example for this discussion....

This topic is all very interesting - but I would like to see what you guys would do with a typical DIY amplifier found in plenty of lounge rooms im sure....  or something similar perhaps.

this thing is a class a/b - 210wrms a channel dual mono (totally separate channels)  power amp...  10hz to 100khz frequency response.

100db sn at 100w - it uses miniature mylar caps (mkt) to shunt electrolytic's in the audio path at high frequencies.  Its meant to be a cheap bang for your buck type of DIY project btw, but this gives us plenty of upgrade options I'm sure.

what would you do to upgrade this amp?  here is the design specs of this simple amp:  http://www1.jaycar.com.au/products_uploaded/KA1760.pdf

I myself would replace all the MKT caps with polypropylene or polyester and also replacing the electrolytic's with polypropylene or at least bi polar's......

then increase the number of power filter caps.... putting more in parallel.  I have found this to be effective in other amps.  

But it would be interesting to know if it would be more effective to shunt the power filter caps with polypropylene or what ever....

other ideas? or definite no no's?  perhaps the ceramics shouldn't be touched or else this 'ringing' issue will come up? I have no experience with regards to this effect.

[Occam]

this thing is a class a/b - 210wrms a channel dual mono (totally separate channels) power amp... 10hz to 100khz frequency response.

100db sn at 100w - it uses miniature mylar caps (mkt) to shunt electrolytic's in the audio path at high frequencies. Its meant to be a cheap bang for your buck type of DIY project btw, but this gives us plenty of upgrade options I'm sure.

But how does it sound? It appears to be an elegant design with the primary merit of difficult to bust high power. It verges on a Self style 'blameless Class B' design'. While I've heard excellent derivative BJT output designs of such, a MOSFet implementation is the worst of all possible worlds. I've not heard it but it might well be an a good subwoofer amp, or possibly the greatest thing since sliced bread.... or simply an exercise in 'polishing a turd'. If you choose your components to upgrade based on specs, rather than sound, there really isn't much to discuss.

myself would replace all the MKT caps with polypropylene or polyester and also replacing the electrolytic's with polypropylene or at least bi polar's......

Mylar(tm) MKT caps are polyester caps....  and please share your experiences  in replacing non coupling electrolytics with bi [non] polar lytics.

then increase the number of power filter caps.... putting more in parallel. I have found this to be effective in other amps.

You think 80,000mmf/rail/channel insufficient??? I can't speak as to the quality and characteristics of RICHEY caps.

But it would be interesting to know if it would be more effective to shunt the power filter caps with polypropylene or what ever....

other ideas? or definite no no's? perhaps the ceramics shouldn't be touched or else this 'ringing' issue will come up? I have no experience with regards to this effect.

Please read the schematic. The rails have an admirable amount PS decoupling, given the cost, on the amp boards. Replace the 10nf signal bypass with a 38cent polystyrene. Upgrade the C7 'speedup' cap to something more appropriate (an 80cent Dayton film and foil? or actually a high K ceramic migh be appropriate in this function) Replace the Zoebel's cap with a box polyprop.
Similarly, examine the function of the ceramics. They are used in the very critical application of feedback, both local and global. They are in the signal path, especially the Miller cap, C8, and C6. I'll assume they are COG (NPO) as they're low values, one doesn't want one's poles and zeroes wandering about, but they're certainly not mlcc ceramics. This is the first place that an upgrade is called for. But non surplus silver mica caps are quite expensive from a manufacturing perspective...

Then again, I've no doubt that spending scads of $ on Blackgates and audiophile approved polyprops would yeild substantial subjective changes. And certainly, replacing that .47mmf MKT input cap with the obligatory Auricap would impress everyone, but with no dc bias on it, would it sound significantly better?

[WerTicus]

You sound like someone who knows his stuff Occam...  However 8000uf of capacitance is enough, i just suggested adding more to increase the filtering quality of the 75v going through them.... I have recently read that you can just shunt them with .1uf polyproplyene's and get similar results or perhaps even better results... also pretty cheap to test out.

Im not interested in 'audiophyle' approved, just what the best upgrade options are for a bang for your buck type effect - Im learning as i go here!

I know mylar is polyester i ment to say polystyrene!

I have no experience in replacing non coupling bi polar electro's with non polar , I was under the impression that you could do this unless there was some special reason it HAD to be polar... if you can save me from making something explode here that would be nice

which caps are the zoebels' cap?

so replacing:
c7 with a film and foil
c8 and c6 are NPO's yes.... so replace with equal value silver mica...
and replace all bypass caps c3,13,15 with polystyrene... no worries mate!

thanks for your advice -I'm going to continue to research this for a few more months before i do anything anyways... I always research upgrades of this nature very heavily.

[Occam]

WerTicus,

I'd like to apologize for the snarky tone of my previous response. I'd mistakenly thought we were discussing a hypothetical amp, and didn't realize you actually had the amp and were an Aussie (the 'no worries,' gave it away)

So here is what I'd initially do....

1. Replace the those feedback caps c6 & c8 with same value silver micas
2. Replace C5, 10nf with an inexpensive polystyrene, it will probably be listed as a 10,000pf (1,000pf = 1nf, 1,000nf = 1uf.[1micofarad])

I've found that these inexpensive polystyrenes are generally unavailable in sizes larger than 10nf (I'm not talking about the excellent MIT RT film and foil polystyrenes, they're very expensive and large). Also, they are very fragile, if your soldering iron touches the body, its ruined. When you solder them, its a good idea to heatsink the lead(s) where they go into the cap itself with a heomostat, or other metallic device to keep them cool.
If you can obtain these inexpensive, but very good styrenes (generally marked LCR or Xicon) in 100,000- pf (100nf, .1uf) for rail bypass duties, use them, but make sure they're 100v or higher, due to the rail voltage...

3. Change the signal bypass cap C7, .1uf, to a good quality polypropolene, prefferably foil as opposed to metalized, but I don't know whats available for reasonable cost in the land down under.

4.I'd also change C2, a bandwidth limiting cap to either a polystyrene or mica (1nf=1,000pf). It might not help, but it will not hurt.

5.Replace the Zoebel cap, C12, 22nf (.022uf), with a good quality polypropelene. Although its a shunt element, there with the resistor to help ensure stability when the amp is on but not connected to speakers, it also contributes to the feedback networks operation.

6. Replace the board mounted PS decoupling  caps, C14 and C16 with same voltage and capacitance lytics, Hitano EXRs. Although unobtainium in the States, these are highly thought of by some Aussie audiophiles, obtainable, and  inexpensive, made in Taiwan. Same goes for the feedback network electolytic, C4,  with a Hitano ESX or a EXR if you can't source the ESX. Please remember to observe the proper polarity when replacing electrolytics, C14 and/or C15 could blow quite spectacularly!

Obviously, you might get better performance with (with increasing cost), Nichicon Muse KZs, FAs, FGs... (descending order of quality and cost) or  even BlackGates, but not knowing the amp, or your tastes, your wallet and prefferences, I'd be loathe to suggest such.

And obviously, the input coupling cap, C1, has a major impact on the subjective quality of the amp. I'm not about to join that debate....

And lastly, the PS decoupling (bypass, shunt) caps can be replaced. My preference would be box style polycarbonates, but they are very expensive, moreso than most 'audiophile' caps, and only make sense if you can find them on surplus. These caps, C17 & C18 which are bypassing the main 80,000uf electrolytics, and C13 & C15, which bypass the board mounted rail electrolytics, could be replaced with similar capacitace polypropolenes, but I'd strongly reccomend box style (or ones that don't necessitate long leads), as having long leads could defeat much of their purpose.

Ideally, I'd reccomend doing these upgrades in stages, so you can ascess their individual impact, but we all know how difficult it is to restrain ourselves (myself included).

Hope this helps.