[kyrill]

wonderful link, thx Ian
but as i read yr explanation, it will not harm if you take >1% ( within reason) of the to be by passed cap. not for electronic reasons but for logistic reasons. Suppose 1% is 0.01uF but i have 0.022uF in the house, then i can use that one. There will not be a hole in the freq spectrum not-covered by one of the parallel caps.
However, a small spectrum will have two caps parallel to each other, in that little spectrum they will influence each value by adding ?

[ginger]

Kyrill,
The 1/100th "Rule of Thumb" came about quite a few years ago. Electrolytics are generally much better these days (their self resonant frequency is generally higher). We have switch mode power supplies to thank for that. They just needed much better electrolytic caps and given the huge volume of such supplies various cap manufactureres were motivated to get of their butts and design them. So using 1/200th or even 1/1000th will  be OK.

I'm a bit embarassed about the comment on using XR7 Ceramic bypasses. When I did a power supply upgrade I found that they added noise and sounded harsh and I ditched them. This is not surprising since ceramic caps have about the worst dielectric absorbtion of any sort of cap that you can get. Use film caps and preferably polystyrene as the bypasses. Where polypropylene is impractical due to physical size limits a Polyester cap will still help.

I will probably get back in the "sand box" in the next couple of weeks to see if I can't mod my LF55 to something like what I think is Hugh's 55LF2.

Off Thread:
I have a couple of "humungous" JADIS JA80 80W Class A Monoblocks to finish a repair on first and by the time they are done my back will be about "stuffed" from carting them to and from the workshop. Not to mention my fingers - for some reason the rubber feet on the amps are just less than finger width in height, which I curse everytime I set them down on a hard surface. They have the unfortunate symptom off wanting to blow up every couple of days which is also not improving my attitude. Its been interesting being exposed to some 1980's supposedly top end, French Tube Amps BUT so far they have chewed up and spat out about half the cost of a LF55 in KT88 tubes - and a LF55 leaves them for dead sound wise.

Cheers,
Ian

[andyr]

... and a LF55 leaves them (Jadis) for dead sound wise.

Cheers,
Ian

Nah, Ian you must be suffering from lack of sleep or too many slabs  ... a tube amp has to sound better than a lousy ss amp, right!!??   

Regards,

Andy

[ginger]

Andy,
I Know your post was "Tongue in Cheek" but thought I'd respond.

The last 3 years of design, build, mod of tube amps has taught me:

Rule 1: Never get into the slabs when working on tube amps with voltage rails of 300 to 500 Volts.
(For none Australians a "slab" is a carton of 24 cans of beer - no, not Fosters which in my opinion is only good for pouring back into the horse).

A lousy amp can be made using tubes or ss and I've made both.
Most lousy tube amps WILL sound better than a lousy ss amp BUT lousy is lousy!!!.
Very, very few tube amps sound better than AKSA 55/100N+ or LFs and that may just be my own particular preference in musical presentation preference. Certainly the "usual" (Yet Another) Williamson Clone or Mullard 5-20 Clone won't do the job.
A tube amp to beat the N+ or LF can be done, I've done it, a stunning little 10W per channel using 6SL7 diffamp front end phase splitter with SS current source tail, MOSFET source followers to drive 1940s AWV 6V6G Ultralinear plus balanced shunt feedback push pull outputs. So you could justifiably say its not really a tube amp but a hybrid - Parts cost around the same as a completed LF55.
The problem with tube amps is COST. Not just of the parts but the support equipment required. This week I bought a ex Channel 9, AVO MKIII Tube Tester - A$1000.
As for a superior tube amp with the power output level of even a 55LF - 3 years down the track and I'm still working on it, I thought I'd made it when I had something which gave the 55N+ some competition but then Hugh released the 55LF.
What did that mean? - Well I had to upgrade my "reference amp" to the LF and it was back to the drawing board on the tube amp.
Cheers,
Ian   

[kyrill]

thx for sharing Ian

couldn't you post until end of summer a daily story of yr wanderings in audio land,  choosing  routes and approaches, avoiding interferences and electrocutions, finding answers and solutions, supported by Uncle Hugh
in search for the creation of a perfect virtual gestalt that will bring us to the past to be the audience?

i will read it every day 
kyrill

[rabbitz]

I ended up trying the MKP1837 by-pass caps in my speaker crossover. The results were not favourable and tended to smear the sound and with loss of focus. There were some parts that did appear to have more air but the sound became fatiguing which has never happened in this system previously (driven by the LF55-2 which is never fatiguing). I don't understand why that happened and maybe someone could give me an explanation.

Tony did mention that it does not suit all crossovers and caps and I will try on another speaker in a different system eventually.

Has anyone else had a go with these in a speaker crossover?

[kyrill]

Well, I quote from the first post of this thread:
But there is one small problem: the bypasses all sound bad! They add a quality that at first sounds like an increase in air and detail, but after a couple of hours becomes an intrusive harshness and discontinuity in the upper treble. Remove the bypass..

But maybe the caps need more hrs to sound at its best?

[rabbitz]

I've never noticed a change in sound with MKP caps. They sound fine from start up not like those BG STD thingies.

[kyrill]

wll,

to unsolder, listen and then inserting the cap again is easy to confirm things

[vmed_cha_gr]

Dear Friend
I did the trick with bypass Vishay MKP1837 on my crossover. I found them on e-bay.
What i found out after listening same cds' again and again is this :
much more clear seperation. Also another thing that came accidently is the amazing quality of the ERO MKC1862 polycarbonate capacitors which are rare to find. They are doing amazing work in crossovers. I couldn't beleive it.
Thanks
Vassilis

[Occam]

Cool- I'm gonna try, because I've been getting nowhere with rs international and a dutch site that I was referred to.

Mark,

I used to have to import the MKP1837s from Europe and ludicrious prices, for my own use. For you Canadians, PartsConnexion is now carrying the 10nf value for 0.99USDea -
http://www.partsconnexion.com/t/catalog/CapacitorsFilm.html
scroll all the way to the right.

And a greater selection is now available directly from Newark Canada -
http://canada.newark.com/jsp/search/browse.jsp?N=500002+1000115&Ntk=gensearch_001&Ntt=Mkp1837&Ntx=

And in the States, Newark.com -
http://www.newark.com/jsp/search/browse.jsp?N=500003+1000115+511&Ntk=gensearch_001&Ntt=mkp1837&Ntx=&getResults=true&appliedparametrics=true&locale=en_US&originalQueryURL=http%3A%2F%2Fwww.newark.com%2Fjsp%2Fsearch%2Fbrowse.jsp%3FN%3D500003%2B1000115%26Ntk%3Dgensearch_001%26Ntt%3Dmkp1837%26Ntx%3D%26getResults%3Dtrue%26appliedparametrics%3Dtrue%26locale%3Den_US

Regards,
Paul

[andyr]

OK, ginger, Hugh and other ex-"purts", since this thread has been opened up again, I would like to ask a Qu.   

As I understand it, the issue of whether small-value "bypass caps" are a good thing or not, needs to be divided into two - completely different - scenarios:

1. electrolytics in PSes (ie. between power rail and ground), and

2. coupling caps in electronic componentss / highpass caps in speaker XOs (ie. these are in series, so the music signal passes through them).

In scenario 1, as has been described in previous posts , due to the chemical/electronic properties of an electrolytic cap, there is advantage in using bypass caps of around 1% value.  5 x 2,200uF caps in parallel will have a lower ESR than one single 10,000uF cap - and thus will perform better - but this is a much more expensive option - and takes up more real estate! 

So parallelling the 10,000uF cap with a 100uF/1uF/10pF combo is a compromise which will sound better than just the 10,000uF cap by itself.

However, in scenario 2, where the music signal passes through the cap, I had understood that bypassing results in smearing - essentially because the different sizes and construction of, say, a 1uF film cap compared to a 10pF cap cause the signal to take minutely different times to traverse the caps. 

Am I on the right track here ... or am I seriously deluded? 

Regards,

Andy

[ZIG]

Andy, you can all call me nuts or whatever..........but.......I have made a large dual capacitor block of solid bees wax containing: 2X Samwha audio grade 10,000mf electros(removed from their cases, wrapped many times in rice paper and sealed), bypassed by 2X Solen 27mf polyprops, bypassed by 2X Sprague 0.22 orange drops. All this mish mash was placed in a cardboard square art box and molten bees wax poured in, effectively encasing all the capacitors in the wax .

Result?: the fastest cleanest power supply capacitor ever.....and guess what?.....I'm still bypassing with giant GE P.P oil caps and the sound is even better. There are wires all over the place and I should be picking up some local radio stations, CB radio and noise from the planet Jupiter......but I'm not!

Power supply bypassing works

Don't know about signal paths though....haven't tried it.

[AKSA]

Andy,

The jury is out, no one can give you a definitive answer.  So, if you have a coupling application, you need to experiment.  I've had good experiences bypassing power electros with 6.8uF block polyester caps, but in all my designs I go to some lengths to specify a particular cap without much bypassing.  The GK1 is an exception, where there is massive bypassing on both the power rails and the supply to the input stage.  It really is empirical, because to measure the differences you'd need very expensive machinery and there is no guarantee you would be able to correlate the measurements with the sonic results.

This is not helpful, but as good as I can make it......    
 
Hugh

[kyrill]

Andy, you can all call me nuts or whatever..........but.......I have made a large dual capacitor block of solid bees wax containing: 2X Samwha audio grade 10,000mf electros(removed from their cases, wrapped many times in rice paper and sealed), bypassed by 2X Solen 27mf polyprops, bypassed by 2X Sprague 0.22 orange drops. All this mish mash was placed in a cardboard square art box and molten bees wax poured in, effectively encasing all the capacitors in the wax .

Result?: the fastest cleanest power supply capacitor ever.....and guess what?.....I'm still bypassing with giant GE P.P oil caps and the sound is even better. There are wires all over the place and I should be picking up some local radio stations, CB radio and noise from the planet Jupiter......but I'm not!

Power supply bypassing works

Don't know about signal paths though....haven't tried it.

Well done

Electrolytes were never invented to sound well, but to store (enormous) amounts of current ( F) So the effects of your careful and detailed approach does not surprise me a bit 

In the line of your approach you could go further by clamping the caps not in a cardbox but clamp them somehow in a (not thin) wooden box.
Twofold reasons

1) try to dampen their mechanical resonances this really helps you could use tube dampers that fits the electrolytes or damp them creatively in another way
2) try to "contaminate" the residual resonances with those of wood. Zero resonances are not possible so pick the best sounding ones: wood

PS engulf them in beeswax?  dielectric constant of Beeswax 2.7 - 3.0 is close to PVC

I would try dampen the caps with raw cotton  (dielectric constant Cotton 1.3-1.4 ) in a wooden box

[terminator]

For me, I don't like to bypass especially in power supplies. Never like the sound of it. Bypassing for
speakers can't comment as I've no experience. I love to use Blackgates & they don't like to be bypass at all.

[zygadr]

Have had to change my Moniker for some reason....sorry about that....I'm still ''ZIG''.
Thanks for your suggestions on encasing the bees wax block. Wood is a great idea as is the wool.

As the block was removed from the cardboard box, it's basically nude wax as such. Was thinking about a multiple cardboard ''wrap'' around it with the cardboard lined with a soft foam or rubber.

[SamL]

For me, I don't like to bypass especially in power supplies. Never like the sound of it. Bypassing for
speakers can't comment as I've no experience. I love to use Blackgates & they don't like to be bypass at all.

Well.. interestingly my experience is opposite of yours. To get my TDA1541 DAC to sound good, I have to bypass the power supply that uses LM317 like mad. Before the bypass, the blackgate std cap on IV board sound horrible. After 3 weeks of running, it still sound hard, unfocus and cold and I have to replace it with Muse. After the power supply bypass, a new set of BG std sound great on day 0. No burn in required.

[zygadr]

During my Kenwood bypass everything (unecessary signal paths/I.C.'s) experiment, I just connected up the cap/bees wax block without the large GE oil cans to see if everything worked.It sounded great, but as soon as I connected the oil cans........wow! .......that made a significant improvement - especially dynamics and bass.