[AKSA]

Dejan,

I'm sorry bwana, not paying attention.......

I use ultrafast soft recovery diodes and Nippon Chemi-con power caps.

They seem to work well, neither ruining the top end nor making the bass too prominent.

I don't (cowering in the corner with dunces cap!) use bypasses.  Ahem.

Cheers,

Hugh

[DVV]

Dejan,

I'm sorry bwana, not paying attention.......

No problem, Hugh, no problem at all.

I use ultrafast soft recovery diodes and Nippon Chemi-con power caps.

They seem to work well, neither ruining the top end nor making the bass too prominent.

Ah, another similarity. I am also strongly inclined towards Nichicon, as I find Elna too bass prominent, it tends to blow up the bass to larger than life, and in my book, that's called coloration.

Nichicon on the other hand, I find much more balanced and perhaps a wee bit faster, all other things being equal.

However, when I'm on the warpath towards the ultimate, nothing beats Siemens Sikorel caps, not surprisingly, at a hefty price. Extreme speed (better than 100 V/uS, better than 10A/uS), extreme clarity.

I don't (cowering in the corner with dunces cap!) use bypasses.  Ahem.

Cheers,

Hugh

No need, Hugh, we already established the sacrosanct right to one's own philosophy at least here, if not in world politics. No war drums beating here.

But I am curios - why don't you use bypasses? I mean, any personal reasons, experience?

Cheers,
DVV

[tmd]

I am a little confused. Were all the above comments about power supply capacitors? It seems that some of them were about capacitors in general.
I was wondering what the deal would be by taking two extremes. Say I wanted 40,000uF total capacitance for the power supply. What would be the difference between 80 x 500uF and one 40,000uF capacitors? What about one 20,000uF and 40 x 500uF? I understand that it doesn't mean I can use cheap ones. I am assuming good quality caps all around.
Thanks, Neil.

[DVV]

I am a little confused. Were all the above comments about power supply capacitors? It seems that some of them were about capacitors in general.
I was wondering what the deal would be by taking two extremes. Say I wanted 40,000uF total capacitance for the power supply. What would be the difference between 80 x 500uF and one 40,000uF capacitors? What about one 20,000uF and 40 x 500uF? I understand that it doesn't mean I can use cheap ones. I am assuming good quality caps all around.
Thanks, Neil.

Very hard to answer precisely, Neil. Some designers argue many small parallelled caps are better because of their improved delivery speed. Others claim big milk bottles of capacitors filter better and should thus be used. In my view, both camps are partly right, but each approach has its downsides as well.

As for your questions above, let's make this a little more precise. Say you want 25,000 uF per supply line, never mind what it feeds. You could buy a 22,000uF big capacitor - the obvious choice. Next, you could use two 10,000uF caps and one 4,700uF all in parallel, or you could use 50 470uF caps in parallel.

The first option is the traditional approach. The second option is an attempt to reconcile greater size which filters better with smaller size which is inherently faster. The third option is all in speed, but I feel with poor filtering effects when compared to the first two.

Other may feel differently. Depends on who you ask, no clear cut answers.

Cheers,
DVV

[tmd]

I think that this is one I definately will try when I get around to building an amplifier or two. It should be fairly easy to do, if a little labour intensive.
Thanks Dejan.
BTW, what make a capacitor good quality? Is it all about materials used or all about manufacturing technique or more likely, some of both?
I have never understood why people spend so much on caps. The most obovious one is spending $1,200 extra on the cap upgrade for the RM40 speakers from VMPS but that is for another thread.
As far as this thread goes, why do (most) more expensive caps sound better?

[BeatleFred]

Hello DVV:

I am a Sansui enthusiast, so I found your slew rate comments interesting to read. Are you familiar with the AU-X1 amp?

Please let me know when your site is available and you post your schematics. (My email is: purepower@nyc.rr.com)

www.sansui.us  

and: www.audiokarma.org

Regards, B/F.

[cjr888]

Could you tell me what a rise time of 1,3 volt per 1 nS (80 Volt in 104nS) is comparison to amplifiers you know?  Good, bad, amazing?

[DVV]

I think that this is one I definately will try when I get around to building an amplifier or two. It should be fairly easy to do, if a little labour intensive.
Thanks Dejan.

Not at all, Neil.

BTW, what make a capacitor good quality? Is it all about materials used or all about manufacturing technique or more likely, some of both?

Quality manufacturers are, in random order, Nichicon, Elna, Panasonic, Rubycon, Philips, Aerovox.  Each has several series, at different prices, so your choice is wide.

I have never understood why people spend so much on caps. The most obovious one is spending $1,200 extra on the cap upgrade for the RM40 speakers from VMPS but that is for another thread.
As far as this thread goes, why do (most) more expensive caps sound better?

Because quality capacitors, in reasonable amount, can help the sound quite a lot, Neil, quite a lot. A poor cap can kill the amp, rob it of life and air.

Cheers,
DVV

[DVV]

Hello DVV:

I am a Sansui enthusiast, so I found your slew rate comments interesting to read. Are you familiar with the AU-X1 amp?

Please let me know when your site is available and you post your schematics. (My email is: purepower@nyc.rr.com)

www.sansui.us  

and: www.audiokarma.org

Regards, B/F.

The CD with it all, and all are schematics, PDF  data sheets for various op amps as used in audio, transistor data sheets, projects and some interesting texts, is in my bag and will go to my webmaster later on in the day. Since there's 41 MB of the stuff, he can't upload it all at once, but I expect him to start tonight.

I am not familiar with AX1 - I know about it, but no more than that.

Cheers,
DVV

[DVV]

Could you tell me what a rise time of 1,3 volt per 1 nS (80 Volt in 104nS) is comparison to amplifiers you know?  Good, bad, amazing?

1,3V per nanosecond is 1,300 V/uS. This would be amazing, and I'd like to see it done in real life in case of power amps. I know some advertise it, but I don't believe it until I see it.

I can accept that figure for the input stage, but not for the whole amplifier, input to output, as this implies incredibly fast output transistors.

Some manufacturers, like Goldmund of Switzerland and Spectral of USA advertise figures like this, but as I say, I don't believe it until I see it.

Cheers,
DVV

[JohnR]

I have never understood why people spend so much on caps. The most obovious one is spending $1,200 extra on the cap upgrade for the RM40 speakers from VMPS but that is for another thread.

I think what a lot of people are missing about that one is that fact that the crossover point in the VMPS is very low. Also, I believe it's a 4 ohm speaker. The result: a very large amount of capacitance is needed, and in a quality capacitor that is not cheap.

[audiojerry]

Since we are straying off topic a bit, is it possible to limit the low-end frequency response of an amp without limiting bandwidth at the other extreme? The reason for my question relates to driving a set of small monitors that are not capable of producing deep bass. An example is the GR Research Criterion which uses a 5.25" Eton midbass with an F3 of 58hz. I would like to prevent frequencies below 50 hz from getting to the speaker.  It plays beautifully down to this point, and I'd like to keep frequencies lower than this from mucking up the driver's excellent excursion.  I don't like the commonly used methods like electronic crossover's, or large capacitors.

In the past I've used small value caps placed between the preamp and amp to filter out low-frequencies, and this has worked ok. I then used a second output from the preamp without the limiting caps to send a full frequency signal to a powered subwoofer using its low-pass filter to shut out the high frequencies.

But I was wondering if a sonically more transparent approach would be to modify the amplifier's low frequency range.

Any thoughts on this?

[DVV]

Since we are straying off topic a bit, is it possible to limit the low-end frequency response of an amp without limiting bandwidth at the other extreme? The reason for my question relates to driving a set of small monitors that are not capable of producing deep bass. An example is the GR Research Criterion which uses a 5.25" Eton midbass with an F3 of 58hz. I would like to prevent frequencies below 50 hz from getting to the speaker.  It plays beautifully down to this point, and I'd like to keep frequencies lower than this from mucking up the driver's excellent excursion.  I don't like the commonly used methods like electronic crossover's, or large capacitors.

In the past I've used small value caps placed between the preamp and amp to filter out low-frequencies, and this has worked ok. I then used a second output from the preamp without the limiting caps to send a full frequency signal to a powered subwoofer using its low-pass filter to shut out the high frequencies.

But I was wondering if a sonically more transparent approach would be to modify the amplifier's low frequency range.

Any thoughts on this?

What you are saying is that you want a high pass filter, one which would filter out below a certain point, but would pass all above that point.

Of course it can be done. For example, you could exchange the capacitor at the input of your power amp with the same type, but smaller value. This would do the trick. Just which type I can't say without seeing the schematic and doing some maths.

As an example, say your input impedance is 47K, and say your current input DC blocking cap is say 100uF. The turnover frequency would be 0.22 Hz. If you were to reduce this cap to say 2.2uF, this would move to 1.54 Hz, and further reducing it to say 0.1uF would move this to about 39 Hz. If you were to insert a high quality film or polycarbonate cap, you'd gain in overall sound quality as well.

However, Jerry, I must warn you that as any filter, so this filter also would affect your overall phase response with unpredictable results. For example, while you may remove frequencies below 39 Hz, you may also find that your remaining bass has gone soft and mushy, or even that your midrange response is not what it should be, or used to be.

Beware of filters, they are always a problem, no matter how well designed and made of which components.

Obviously, my advice to you is to let mother Mary come to you, speaking words of wisdom - let it be, let it be.

Cheers,
DVV

[audiojerry]

Thank you Mother Mary; I was afraid you'd say something like that. But given all filters are compromises, wouldn't this still be a better compromise than using a much higher value cap before the midbass driver on the crossover?

[DVV]

Thank you Mother Mary; I was afraid you'd say something like that. But given all filters are compromises, wouldn't this still be a better compromise than using a much higher value cap before the midbass driver on the crossover?

I don't follow? What has the cap before the midbass got to do with it, when it simply cuts off the lower range for the tweeter?

Jerry, those are two different, very different, crossover points.

Cheers,
DVV

[audiojerry]

Caps are normally used to restrict low frequencies from the tweeter, yes.
But the same techinique is used for midbass/midrange drivers when a larger woofer is employed. In a 3 way system, a cap may be used to pass frequencies 100hz and above to the midbass and tweeter. Oftentimes a large value electrolytic is used for this, which is of course a crappy technique. In a 2 way system like the Criterion with only a 5" midbass, I'd like to keep frequencies 55hz and below away from the midbass. I know I can do this with a cap on the positive lead of the speaker terminal, but I was wondering if a smaller value cap at the amp input would be better.
Am I making sense?

[DVV]

Caps are normally used to restrict low frequencies from the tweeter, yes.
But the same techinique is used for midbass/midrange drivers when a larger woofer is employed. In a 3 way system, a cap may be used to pass frequencies 100hz and above to the midbass and tweeter. Oftentimes a large value electrolytic is used for this, which is of course a crappy technique. In a 2 way system like the Criterion with only a 5" midbass, I'd like to keep frequencies 55hz and below away from the midbass. I know I can do this with a cap on the positive lead of the speaker terminal, but I was wondering if a smaller value cap at the amp input would be better.
Am I making sense?

Yes, you are, but what you suggest (a cap in between to the mid/bass) is, in my view, easily the worst thing to do. The entire sound coming out of your speaker will then be at the mercy of that capacitor. Also, how large? What it says on the data sheet is nice, but yo really know, you have to experiment, which implies measuring gear.

Ultimately, you want to do what the manufacturer of that speaker did not see fit to do, and this is no small, idle tweak, this is as capital as it gets.

Don't do it, Jerry.

Cheers,
DVV

[audiojerry]

I may not do it, but it is already being done my manufacturer's. How do you explain a 3 way system with a passive crossover, where the bass driver operates in the 30 to 150hz range, with the higher frequencies going to the remaining two drivers? I've seen this, and the cap that was used was electrolytic.

[OBF]

audiojerry,

My first semi hi-fi system consisted of a Rotel receiver, Paradigm 7SEs, and a Velo sub.  I'm not totally positive of where the cap went, but I believe the dealer installed a cap between the preamp and amp stages in the Rotel to spare it the bottom frequencies the sub was playing.  It basically amounted to a 1st order filter at 60hz.  I'm sure I didn't get a perfect blend between the mains and sub, but it seemed to work pretty well and was similar to what you're asking about.  The system would play darn loud and we put that Rotel through all kinds of abuse.  My roommates did manage to wreck 3 of the 4 woofers in the Paradigms by overexcuding (sp?) them, even WITH the filter.  Geez, so I guess it's not a good idea to turn your 50 watt receiver ALL the way up?  

[DVV]

I may not do it, but it is already being done my manufacturer's. How do you explain a 3 way system with a passive crossover, where the bass driver operates in the 30 to 150hz range, with the higher frequencies going to the remaining two drivers? I've seen this, and the cap that was used was electrolytic.

Jerry, that is different insofar that the speaker designer knew EXACTLY what he was dealing with, because he presumably measured his circuits beforehand.

This is true not only of speakers, but of everything. I can tell you in words how to make a current generator, but I can only guess at the values, even if the basic circuit is well neigh perfect. I can give you values which will work well in one ciruit, but not so well, or even very poorly, in all others but that one.

I agree the ear is the final judge, but before we get to the ear, we have some walking to do.

Cheers,
DVV