Indeed, that is how I do it... select a slightly lower value and top with smaller SM or styroflex.
In the case of the trebble cap (11nF) the difference in cap used might produce dramatic results... micas are ver well defined but stark... styroflex sounds duller sometimes.. it is a question of taste.... If we could get teflons in the pico farad range
In the case of the trebble cap (11nF) the difference in cap used might produce dramatic results... micas are ver well defined but stark... styroflex sounds duller sometimes.. it is a question of taste.... If we could get teflons in the pico farad range
I must confess I never experimented with V-caps or other expensive teflons.... I use FP-3 CCCP teflons for interstage with superior results but these "cheap" military units do not work for me in the riaa trebble cap position... the extension is awesome but lacks the "fullness" of a tin foil RT cap for example.
If you didn't you are missing something, performance of thse caps is a lot superior to anything else you ever tried.
It is something I have to experiment with once boards get in.
It never happened to me that a 0.5% or even 1% tolerance on the RIAA could cause such a high displacement.
I guess that this high sensitivity to the nominal values might be due to the high output impedance of the gain stage.
Anyhow, like I said, parts I get are usually better than 0.5%
And even if it wouldn't be enough, do you have a precise LCR meter that can go lower than 0.1% tolerance on C measurements?
I think mine gets down to 0.05% I have to double check.
Last edited:
IMHO 0.5% = 1% / 2 = 0.25% x 2
The point is 11.11nF is NOT 'way different' from 11nF it is 1% we, being technicians, should not lose our self’s in Hi-end-techno-babble. We may agree that 1% is too much, we I am never going to agree that 'depending on situation 1% could be more (or less) then 1% (in some other situation).
Just MHO
The guys from the ATL do agree, and are mounting a fresh set of 0.0000001u% capacitors in there record player (Note: u% = 1/1000000 of an %) and complaining that the f% capacitors did not arive.
I already had some heavy discussions about this subject with Thorsten and got no concensus.
Anyway, I can assure you 10pF difference in 11n will be audible in the trebble region. Maybe because these 10p come from a different kind of cap bypassing the main one... but it matters.
I also know that technicians tend to base their conclusions based on a math truth alone.... In that case a MKS should not differ from a teflon but........
In the begining I believed this design was so heavily "filtered" that it was inpervious to cap type changes.... now it seems the idea is growing in another way... this is cool.
I am already very happy that some of us agree on the idea that the riaa values can be trimmed to perfection and 9kr is different from 9.9kr.... fortunately we now have a very good base for testing.... If possible I will try the teflons (If I can afford those)
Fascinating subject
Anyway, I can assure you 10pF difference in 11n will be audible in the trebble region. Maybe because these 10p come from a different kind of cap bypassing the main one... but it matters.
I also know that technicians tend to base their conclusions based on a math truth alone.... In that case a MKS should not differ from a teflon but........
In the begining I believed this design was so heavily "filtered" that it was inpervious to cap type changes.... now it seems the idea is growing in another way... this is cool.
I am already very happy that some of us agree on the idea that the riaa values can be trimmed to perfection and 9kr is different from 9.9kr.... fortunately we now have a very good base for testing.... If possible I will try the teflons (If I can afford those)
Fascinating subject
Last edited:
My point was/is that 1% is 1% and it is not depending on circumstances. But I do agree (like the ATL guys) the 0.1% (even less) can make a big difference, but it is not depending on circumstances, there is a technical reason for that (even if we do not understand the reason).
There is a lot of multiplication going on the the mirror. Question is if the mirror holds that 0.1% tolerance too. I prefer to build and then measure. I can only tell you that the new Pardise sounds swell. I use Styroflex in the treble and Wima FKP2 in the other positions.
Especially the midrange cap has a resistor in series. For me it makes no sense to use Teflon there out of the reason of lowest DA. That will be totally swapped by the series resistor. In my best commercial stages i use air trimmer caps to fine tune.
Especially the midrange cap has a resistor in series. For me it makes no sense to use Teflon there out of the reason of lowest DA. That will be totally swapped by the series resistor. In my best commercial stages i use air trimmer caps to fine tune.
I believe you hear this difference
The only thing that I try to set firmly when I conduct this experiments are:
1) make an objective measurement on the frequency response to see whether the 10pF has really altered and helped the response or not.
2) Verify if the change is more capacitor's like. In other word if the sonical difference you hear is because the capacitor has introduced a sonical signature that you might like better or just sounds different to you.
Now I meadure that RIAA has gone off, I don't even consider the change and go back to the test bench.
I usually don't tune the filter by ear because it would be way too subjective.
Conversely if curve improved, then I listen and I try to change caps with same value to see if the sound I hear is mostly due to the qulity of the cap.
Usually I tend to prefere 1 cap and no parallel to increase value.
It is always better to have one good quality cap.
What I do, I find best value on the bench and then, if possible, I try to get a custom value.
If I can't I found that I can get the next higher value for Aura T and I strip some Teflon off and close the cap.
If works pretty good.
Work is clean and still sounds much better than any WIMA you can buy.
In this specific case I am particularly lucky since I got 11nF and 32nF from the factory as custom values which is even better than having to strip material off.
I also believe (I will measure those again) that they are 11.008nF and 32.014nF which is like you can see higher precision than 1%
The only thing that I try to set firmly when I conduct this experiments are:
1) make an objective measurement on the frequency response to see whether the 10pF has really altered and helped the response or not.
2) Verify if the change is more capacitor's like. In other word if the sonical difference you hear is because the capacitor has introduced a sonical signature that you might like better or just sounds different to you.
Now I meadure that RIAA has gone off, I don't even consider the change and go back to the test bench.
I usually don't tune the filter by ear because it would be way too subjective.
Conversely if curve improved, then I listen and I try to change caps with same value to see if the sound I hear is mostly due to the qulity of the cap.
Usually I tend to prefere 1 cap and no parallel to increase value.
It is always better to have one good quality cap.
What I do, I find best value on the bench and then, if possible, I try to get a custom value.
If I can't I found that I can get the next higher value for Aura T and I strip some Teflon off and close the cap.
If works pretty good.
Work is clean and still sounds much better than any WIMA you can buy.
In this specific case I am particularly lucky since I got 11nF and 32nF from the factory as custom values which is even better than having to strip material off.
I also believe (I will measure those again) that they are 11.008nF and 32.014nF which is like you can see higher precision than 1%
And there was me thinking I was the only relic from the previous century in town.