I see cap datasheets that do not specify ESR.
Instead, it specifies tan ( delta ) at 120 Hz.
Are we supposed to calculate ESR using ?
ESR = Z x tan ( delta ) where Z = 1/( C x 2xPi x 120)
Instead, it specifies tan ( delta ) at 120 Hz.
Are we supposed to calculate ESR using ?
ESR = Z x tan ( delta ) where Z = 1/( C x 2xPi x 120)
Ok, I found ESR hidden under the name "impedance 100KHz".
A Panasonic FR 470uF 25V 8mm x 20mm is 0.030 ohm
*Really* the best/cheap/fast option is to replace them, period.
Measuring them is fine if you have nothing else to do .... but then what if (proper) measurement pronounces them dead/lossy/out of tolerance?
Back to square one: replace them.
Measuring them is fine if you have nothing else to do .... but then what if (proper) measurement pronounces them dead/lossy/out of tolerance?
Back to square one: replace them.
I understand.
However, I like to check things for myself, to see realities, the truth about capacitor aging and the actual need for recapping.
Thanks guy, seems I need to re learn my math....
Ask myself if the capacitances i have measured grom the two filters are not the genuine values since 85...left and right speaker have less than 5% between eachothers set of caps.
It is what is suggesting Elvee, 80s npe are reliable enough to display their initial uF values... while many display 10% more than their marking...but very consistant between left and right channel ...between zero and 5% left/right diference...
Maybe 35 or 40 of week-end only use is not so hard for rhose npe from the 80s????
While I read npe changing after such long period is always giving a clearer sound, less dark...some shunt caps may have changed perhaps...while still close to 10% tolerance...
I have the feeling I should just bypass all the caps with cheap 0.1 uf mkt instead swaping all the lytics maybe... Not sure I know someone around me with a GenRad stuff for such measurement and seems not usefull in case of muF scale for speaker filters...
Ask myself if the capacitances i have measured grom the two filters are not the genuine values since 85...left and right speaker have less than 5% between eachothers set of caps.
It is what is suggesting Elvee, 80s npe are reliable enough to display their initial uF values... while many display 10% more than their marking...but very consistant between left and right channel ...between zero and 5% left/right diference...
Maybe 35 or 40 of week-end only use is not so hard for rhose npe from the 80s????
While I read npe changing after such long period is always giving a clearer sound, less dark...some shunt caps may have changed perhaps...while still close to 10% tolerance...
I have the feeling I should just bypass all the caps with cheap 0.1 uf mkt instead swaping all the lytics maybe... Not sure I know someone around me with a GenRad stuff for such measurement and seems not usefull in case of muF scale for speaker filters...
Also, the measurement of speaker caps may be usefull for some speaker brands which have tight tolerance and sort put the caps within 1 to 3% from the 10% marked tolerance on the cap...for instance a 150 hz xo is not 190 hz...
Browsing with Mouser or Digikey component selectors you can see the available capacitor tolerances and prices.
To get matched pairs, one can buy ten 5% caps,
there must be at least one pair 0.5% matched.
To get matched pairs, one can buy ten 5% caps,
there must be at least one pair 0.5% matched.
For usual values, it's not a problem with film caps. They often are between the exact value and 3% for most.
sourcing is not a problem either. The only problem, is to know the exact values the designer sorted out because it's about knowledge as yu wrote above
Hence all the good advices given. If it was not about knowledge and respect towards the people that developped speakers (well it's our hobby, isn't it) I will just go for a standard swap... Learned many things since post 1 about npe and...also polar caps (which sees the simpliest of the both). So I assume I will go with Elvee"s tip to try to know if the capacitances values fall a little despite the certainly to huge esr due their old age... I like also the idea it cost almost nothing while I learn something in between after they gone.
sourcing is not a problem either. The only problem, is to know the exact values the designer sorted out because it's about knowledge as yu wrote above
Hence all the good advices given. If it was not about knowledge and respect towards the people that developped speakers (well it's our hobby, isn't it) I will just go for a standard swap... Learned many things since post 1 about npe and...also polar caps (which sees the simpliest of the both). So I assume I will go with Elvee"s tip to try to know if the capacitances values fall a little despite the certainly to huge esr due their old age... I like also the idea it cost almost nothing while I learn something in between after they gone.
Last edited: