Tannoy capacitor advice needed

P

pete693

Joined
Aug 4, 2008
Messages
143
Reaction score
0
Location
Stanmore Middlesex
Having had my Tannoy HPD speakers re-coned it got me thinking about the caps in the crossovers.They have been working away in their Berkeley cabinets since 1978 and I don't know to what extent they may have degraded in that time.
So this is the question.Should I leave well alone or change them.
I've seen lots of posts on the subject with many various opinions. If I change them I will want to replace them with something competent but not outrageously expensive.
My schematic shows the following:
1x4.7 Mfd,1x1.5 Mfd,1x3.3 Mfd,in the HF circuit.
2x6.8 Mfd,1x3.3 Mfd, in the LF cicuit.
Any suggestions of makes/types would be appreciated.I should add that I'm not really interested in trying a number of different types to test and see what I like best.I just want to put them in and forget about them for another 35 years,(I should be so lucky).
Pete.
 
Hi Pete,

Solen polypropylene from Falcon Accoustics.

Excellent quality, great value and they'll pair them to 1% tolerance for a very small fee.
They also stock odd values sometimes used in vintage crossovers and not generally available today.

I use them when I do a crossover recap.

However, I'd only change any electrolytic caps. If any original film caps are present I'd leave them.
 
Having recapped many HPD crossovers, the one addition to Rob's advice above is that the notch filter "smoothness" seems sensitive to the type of cap used. Whilst I generally advise Claritycap ESAs for the 4u7, Solens everywhere else but use Mundorff Evo Oils for the 3u3 notch filter. This cap design is meant to result in very low distortion compared with other types, and I've certainly found it to enhance the acoustic performance of the HPDs by offering a smoother and clearer response close to the 3K horn resonance point. Why that should be, I don't really know, but it works well.
 
Tenson said:
Hi, Do you have any response graphs to show the difference?

Cheers
Click to expand...

I've compared the responses on impedance curves measured through the crossover, and as you would expect, you wont see much of a difference in that response because the cap values are the same. However used in the notch filter, I can assure you that of all the caps tested (and there were many) the Mundorf for whatever reason consistently sounded the best one to use. As mentioned previously, I can only point towards the low plate distortion of the evo-oil design making that acoustic difference. Mundorf claim that the design and the way it is wound is key to its clean sound. I normally put getting the correct, accurate capacitor values over any form of "foo-ness" as I do believe that a well designed crossover using modest components always sounds better than a badly designed crossover using expensive components.

The point is that the Mondorf's aren't that expensive (not much more than the Solens really) and do make a difference to my ears when used for the notch filter. That is my experience and I've done fairly extensive work on new HPD crossovers for my own speakers, the Reference Fidelios.

Tannoy got it right with the circuit and their values so no need to deviate much from what they designed. The notch filter can be improved by altering LCR values to widen the flattening range, but you'll never eliminate that horn resonance completely.
 
On the subject of values.
My Berkeleys are fitted with 386 DCU speakers.The crossovers have different value caps than those fitted with the 385 DCU.
The first series cap in the H.F. circuit is a 6u8 and not the usual 4u7.The cap in the notch filter is a 4u7 and not the usual 3u3.
Many years ago as an experiment I changed the cap values to the more usual 4u7 and 3u3.The result was terrible.Bass appeared unaltered but vocals, especially female vocals were what at best could be described as spitty.A sort of extreme sibilance.I was not prepared to give them any settling time they were just too bad.I went back to the original Tannoy values and all was well again.
I should add that the values I put in my first post were the ones shown on a schematic that I had to hand.It wasn't until I looked at the crossovers again recently that I remembered my previous findings and the difference in values that Tannoy had originally fitted.
 
Last edited by a moderator:
pete693 said:
On the subject of values.
My Berkeleys are fitted with 386 DCU speakers.The crossovers have different value caps than those fitted with the 385 DCU.
The first series cap in the H.F. circuit is a 6u8 and not the usual 4u7.The cap in the notch filter is a 4u7 and not the usual 3u3.
Many years ago as an experiment I changed the cap values to the more usual 4u7 and 3u3.The result was terrible.Bass appeared unaltered but vocals, especially female vocals were what at best could be described as spitty.A sort of extreme sibilance.I was not prepared to give them any settling time they were just too bad.I went back to the original Tannoy values and all was well again.
I should add that the values I put in my first post were the ones shown on a schematic that I had to hand.It wasn't until I looked at the crossovers again recently that I remembered my previous findings and the difference in values that Tannoy had originally fitted.
Click to expand...


Tannoy have it generally right and for good reason. They spent many years R&D perfecting those values and knew what they were about. People experimenting with different values sometimes forget that the values were originally carefully worked out to the achieve the optimum crossover to avoid running woofer cones too high into cone break-up area, or running tweeters too low. The values are specifically matched to the specific impedance curves for the drivers. Often, the end result of differing from these carefully considered values is worsening the performance in some way. I'd say that the only exception is experimentation with the notch filter values as I do believe that these can be improved upon.
 

Latest posts

Back
Top