Hello,
Recently, while looking for some electrolytic NP capacitor for relatively large values in speaker crossovers, I came across info on the foil inside.
Rough foil vs smooth foil. The smooth foil capacitors are 4-5 times more expensive.
Smooth foil are suggested for HF circuits, rough foil is suggested for LF circuits.
Mundorf BG50 vs Mundorf ECap100.
What is your take on this?
Recently, while looking for some electrolytic NP capacitor for relatively large values in speaker crossovers, I came across info on the foil inside.
Rough foil vs smooth foil. The smooth foil capacitors are 4-5 times more expensive.
Smooth foil are suggested for HF circuits, rough foil is suggested for LF circuits.
Mundorf BG50 vs Mundorf ECap100.
What is your take on this?
Are you trying to distract yourself from the ugly and expensive coils that also go along with these passive crossovers?
Use an active xo!
Use an active xo!
Smooth foil bipolar electrolytic capacitors are supposed to be less lossy than rough foil electrolytic capacitors, so a filter with smooth foil bipolar electrolytic capacitors will behave a bit closer to theory than one with rough foil bipolar electrolytic capacitors. When there is a resistor in series with the capacitor, like in impedance correction networks, you may be able to roughly correct for its losses by using a slighly smaller resistor value.
Have you found any datasheets for the capacitors you are interested in? They should state the loss angles.
Have you found any datasheets for the capacitors you are interested in? They should state the loss angles.
There is too much supposedness on the internet. Good you are pointing out to the datasheet!
I have an '90's copy of an Elektor Loudspeaker Special in which they tested various cap's, amongst others bipolar smooth and rough foil types. Amongst them were two from Visaton, both 2u2. Both scored pretty bad compared to MKP/MKT's. Interestingly, the smooth one had a 5 times worse Dielectric Absortion figure than the rough one. Go figure.
Bipolar electrolytic capacitors are all bad compared to MKT and MKT is bad compared to MKP. Then again, MKP capacitors are a whole lot bigger and more expensive than bipolar electrolytic capacitors when you need really large values.
https://www.hificollective.co.uk/sites/default/files/2021-07/mundorf-ecap-datasheet.pdf
For the BG50, the only info I found is :
I'm hoping to reach somebody with real life experience on rough vs smooth...
For the BG50, the only info I found is :
- Tolerance: 5%
- Terminal diameter: 0,8 mm
- loss angle: tan(delta) = 0,05
I'm hoping to reach somebody with real life experience on rough vs smooth...
A tan(delta) of 0.05 for rough foil, 0.025 for smooth foil according to the pdf you linked to. Both values are pretty good compared to the 0.1...0.2 that I'm used to for electrolytic capacitors (but not quite as good as the 0.001 of a typical MKP capacitor).
For Jantzen Superior Z :
I'm wondering how tan(delta) values reflect/translate in to sound?
(I found no tan(delta) mentioning in the datasheets of some of the most cherished capacitors available)
If Chat GTP suggest smooth film are for HF circuits, then "A tan(delta) of 0.05 for rough foil, 0.025 for smooth foil" may contradict with "the less is better" or else?
- Loss angle tan = 0.00002 to 1K
- Loss angle tan = 0.00001 to 10K
I'm wondering how tan(delta) values reflect/translate in to sound?
(I found no tan(delta) mentioning in the datasheets of some of the most cherished capacitors available)
If Chat GTP suggest smooth film are for HF circuits, then "A tan(delta) of 0.05 for rough foil, 0.025 for smooth foil" may contradict with "the less is better" or else?
Educated guess: I expect that the quality factor of your crossover filter will drop somewhat compared to what you would get with ideal capacitors, causing a bit of a dip around the crossover frequency, but that the effect will be small whenever tan(delta) << 1/Q. For a fourth-order Linkwitz-Riley filter, that means tan(delta) << √2.
ESR value is not the same at 1 kHz and at 5 kHz. You will notice that by measuring two different high passed
responses of a tweeter, one with an NPE and another with a higher quality film cap, by overlaping these graphs.
Tan(delta) of 0.05 means ESR value at the measured frequency is 5% of capacitive reactance value (Xc=1/(2*pi*f*C).
C=capacitance in Farads.
100uF NPE, Xc=1.59, 5%=0.08R
I would like to thank you all for your inputs. It looks like lower ESR may lead to "issues" unless capacitance of the replacement MKP and the next R value in the signal path are taken in consideration.
I'm getting vertigo to be honest here...
There is my simple question:
20 y old Hanlan 100v , +-10% capacitors,
-HF-- 5.7uF (4.7uF + 1uF film), 15uF, (next in the signal path is 2R2, then tweeter...)
-LF-- 22uF
Space is an issue!
Suggestions for electrolytics , please?
I'm getting vertigo to be honest here...
There is my simple question:
20 y old Hanlan 100v , +-10% capacitors,
-HF-- 5.7uF (4.7uF + 1uF film), 15uF, (next in the signal path is 2R2, then tweeter...)
-LF-- 22uF
Space is an issue!
Suggestions for electrolytics , please?
I think the Mundorfs with rough foil and with smooth foil will both be perfectly acceptable. Both meet tan(delta) << √2. The rough foil ones probably take less space.
