Capacitor question for KEF speakers.

[vadimgal]

Hi To all,
I have problem with replacing the caps in my speakers KEF REF. TWO
One months ago I realized that one tweeter sounds bad. I pulled out the x-over which has all NON polar Bennic and ALCAP caps. I tried to replace them with Clarity Cap polypropylene ones and sound is not good.
IS it good Idea to mix NON polar caps with polypropylene ones or all caps has to be the same .
Thanks.

[Bobken]

Hi,

Caps made from different materials and by alternative makers usually 'sound' a bit different from each other. As you are talking about all non-polarised film caps here, there is no underlying reason not to 'mix and match' these in a x'over, but they will be likely to give a different overall balance or 'voice' to the results.

Possibly you have struck an unlucky combination here, if you have caps of the correct values, but makers like KEF are careful enough to select their caps where necessary, to suit the application.

Therefore a 10% tolerance Alcap of say nominally 4.7uF, could be anywhere in value between around 5.2uF, or maybe as low as 4.2uF, and this will make lot of difference to the overall 'voicing' in most cases. Component and entire x'over suppliers/manufacturers like Falcon Electronics, which was run by Malcolm Jones the ex-chief engineer from KEF, supplied a lot of high-end speaker manufacturers with their parts. Falcon was adept at this component 'selection' exercise, and I am sure they did this for many of their customers, to satisfy their special needs.

If you have the means, it is always best to measure any caps before replacing them in a x'over like this, and try to obtain something of a similar value, and then you will only have the slight variations in the individual cap's 'sonic' response to worry about.

Regards,

[AndrewT]

Hi Vad & Bob,
I think that replacing an electrolytic with all it's non ideal characteristics with a plastic based capacitor of the same value (measured and matched) will sound different.
I further think that saying "only"

then you will only have the slight variations in the individual cap's 'sonic' response to worry about

understates the size of the project involved in changing a well designed combination of crossover and drivers.

Sorry Bob, but this could be another area where we do not see "eye to eye".

[KP11520]

Maybe "Burn in" is a factor here. The old caps have hundreds and maybe even thousands of hours on them.

Many caps get better after the first 24 hours and some continue to change (slowly) for hundreds of hours.

Bob would be able to make a better assessment on that but maybe it could be a factor too!

Regards//Keith

[AMV8]

Hi

Polypropylene and bi polar caps have different charactoristics which can make it difficult to change over.

Having said that I think you will find that adding some resistence in series with the polypropylene capacitor will help a lot ( Kef recommended this when they changed from bi polar to poly caps ).

Kef recommended, and I also agree, you need IR for each 10uf of capacitance.

Don

[Bobken]

Originally posted by AndrewT
Hi Vad & Bob,
I think that replacing an electrolytic with all it's non ideal characteristics with a plastic based capacitor of the same value (measured and matched) will sound different.
I further think that saying "only" understates the size of the project involved in changing a well designed combination of crossover and drivers.

Sorry Bob, but this could be another area where we do not see "eye to eye".

Hi Andrew,

There's no need for any apology here, and there is no reason for us always to see matters exactly the same, as we have had different experiences in life (and in electronics).

You might well be right here, as we are talking about magnitudes of change, which will perhaps impress us slightly differently.

Over the years I have successfully replaced quite a few Alcaps in all kinds of x'overs (Alcaps were all the rage at one time, of course) including a couple of years ago in some monster TDL Reference speakers, which I have mentioned in another thread.

I haven't seen any drastic changes in overall balance which I was able to measure on these occasions, but to some extent it will relate to which 'substitute' caps are used, their individual values, and more importantly whether they are in shunt or in series with the signal.

On no occasion would I have described the subjective results as "not being good" as vadimgal has said, and when used in series I found it will generally 'enhance' the frequency-range being affected, but where used in shunt locations, it has had the opposite effect.

Without knowing anything further about this particular x'over, nor precisely what vadimgal's dissatisfaction is based upon, I wouldn't care to pass any other opinions on this. In other threads quite recently, I spent a lot of time impressing upon others just how difficult and time-consuming 'voicing' speakers and x'overs can be, and to end up with exceptional results is usually quite heavy going.

Regards,

[Bobken]

Originally posted by KP11520
Maybe "Burn in" is a factor here. The old caps have hundreds and maybe even thousands of hours on them.

Many caps get better after the first 24 hours and some continue to change (slowly) for hundreds of hours.

Bob would be able to make a better assessment on that but maybe it could be a factor too!

Regards//Keith

Hi Keith,

Burn-in will have some affect, of course, but with this poster's description it seems like a much more serious problem than is likely to be improved much by any burning-in process, in my view. In general (here I go again!) I would expect a fresh film cap to sound more like an old electrolytic in characteristics, with the usual harshness and lack of transparency observed with electrolytics, and a similar expectation with most 'new' film caps.

It would (will?) be interesting to learn more about this problem, whereupon it should be possible for us to give some more relevant and useful advice here.

Regards,

[Bobken]

Hi vadimgal,

If you would tell us a bit more about the caps which you replaced, together what you are unhappy with after this substitution, I am sure that we would be able to offer some more-worthwhile advice to you.

You see, so far none of us know the value of the caps, or even which ones they were (unless I am half-asleep), and possibly quite wrongly, I originally assumed that you probably meant the Bennics. You mentioned a problem associated with tweeters, and these are more likely to be the lower-value caps there, which knowing of some KEF x'overs, I would have guessed would probably be plastic films, anyway, just like the majority of yellow film Bennics I have seen before. I felt obliged to guess here as I am not personally familiar with these particular KEFs, whereas others may be so.

It is a slightly different matter 'soundwise' if they are Alcap electrolytics, which some posters here assume you replaced, but, either way, the symptoms you describe appear to be more severe to me than a minor change in the overall 'voicing'.

Regards,

[sreten]

Originally posted by AMV8

Kef recommended, and I also agree, you need IR for each 10uf of capacitance.
Don

hmmm....... so how much for 47uF ? /sreten.

[Bobken]

Originally posted by sreten


hmmm....... so how much for 47uF ? /sreten.

Actually, I was aghast to read that apparantly KEF recommend *any* resistance in series with *any* drivers except maybe where part of an 'L' pad attenuator, or similar, is concened.

I certainly would not have added anything like 3+ Ohms in series with my ATC mid-domes, when the speakers concerned had passive X'overs, as I know what it would have done to the sound. It would certainly have countered most of the benefits I achieved when replacing polypropylene Chateauroux caps with a bank of all MIT RTXs, these being polystyrene and tin-foils!

Who am I to disagree with KEF, though.

Regards,