• The move to the new server is done. There are some software and database maintenance updates in process. This has us passing the hat around to help out. We appreciate any donations. Seriously, even a dollar helps. The payment page may be found here - https://www.audiokarma.org/support.html

Low leakage electrolytic capacitors

dlucy

dlucy

dlucy67 (Doug)
Staff member
Moderator
Subscriber
Low leakage electrolytic capacitors are favored by many for use in the audio signal path since they have the least effect on the signal. The Nichicon KL or UKL series has been favored most as it has the lowest leakage stats (from their datasheet), but these have been discontinued and are now hard to source.

A list I compiled years ago from the Nichicon electrolytic capacitor data sheets may help you as you look for the next-best electrolytic cap in terms of leakage:

Comparing Nichicon 200 uF caps:

SeriesPackagePNuFVHoursSeries leakageUnit leakageVoltage rangeCapacitance rangeNotes
UKLRadialUKL1J221KHD220632,0000.0020.027726.3 to 1000.1 to 10000lower leakage for less noise in signal path
UFGRadialUFG1J221MHM220631,0000.0100.138606.3 to 1000.1 to 10000Better audio than FW
UKZRadialUKZ2A221MHM2201001,0000.0100.2200025 to 10010 to 1000
UFWRadialUFW1J221MPD220632,0000.0300.415806.3 to 1000.1 to 33000
UKWRadialUKW1J221MPD220632,0000.0300.415806.3 to 1000.1 to 33000Better audio than FG
UPWRadialUPW1J221MPD6220638,0000.0300.415806.3 to 4500.47 to 15000low impedance

Comparing Nichicon 2,000 uF caps:

SeriesPackagePNuFVHoursSeries leakageUnit leakageVoltage rangeCapacitance rangeNotes
UKLRadialUKL1V222KHD2200352,0000.0020.154006.3 to 100V0.1 to 10000µFlow leakage
UKARadialUKA1V222MHD2200352,0000.0100.770006.3 to 50V22 to 22000 µFhigh grade audio, high temp
UFWRadialUFW1H222MHD2200502,0000.0101.100006.3 to 100V0.1 to 33000µFstandard audio equipment
UHERadialUHE1H222MHD622005010,0000.0101.100006.3 to 100V0.47 to 18000µFlong life
UHWRadialUHW1H222MHD22005010,0000.0101.10000smaller, long life
UKARadialUKA1H222MHD2200502,0000.0101.100006.3 to 50V22 to 22000 µFhigh grade audio, high temp
UKWRadialUKW1H222MHD2200502,0000.0101.100006.3 to 100V0.1 to 33000µFstandard audio equipment
URSRadialURS1H222MRD2200502,0000.0101.100006.3 to 400V0.1 to 10000µFsmaller, compact and low profile
URZRadialURZ1H222MRD2200501,0000.0101.100006.3 to 400V0.1 to 10000µFcompact and low profile, high temp
UVKRadialUVK1H222MHD2200501,0000.0101.100006.3 to 35V6.8 to 220µFhigh temp, low impedance
UVRRadialUVR1H222MHD2200502,0000.0101.100006.3 to 450V0.1 to 33000µFsmaller size than VR
UBYRadialUBY1H222MHL2200503,0000.0303.3000025 to 50V620 to 12000µFhigh temp, long life, high ripple
UPJRadialUPJ1H222MHD2200505,0000.0303.300006.3 to 450V0.47 to 15000µFlong life, low impedence, for switching power supplies
UPMRadialUPM1H222MHD2200505,0000.0303.300006.3 to 450V0.47 to 15000µFlong life, high temp, for switching power supplies
UPSRadialUPS1H222MHD2200503,0000.0303.300006.3 to 450V0.47 to 15000µFsmaller version of UPJ, for switching power supplies
UPWRadialUPW1H222MHD2200508,0000.0303.300006.3 to 450V0.47 to 15000µFsmaller and lower impedence than UPM, for switching power supplies
 
Last edited:
Register to hide this ad
PabloTincho said:
What about tantalum caps by Kyocera

Audio F95 Series | KYOCERA AVX

AVX Audio F95 Series Solid Tantalum Electrolytic Capacitors are conformal coated chip capacitors that are designed for audio applications.
www.kyocera-avx.com www.kyocera-avx.com
or
https://partsconnexion.com/ has the
Audio Note and the lower brand SEIRYU Series
Click to expand...
This thread is for electrolytic caps that replace other electrolytic caps.

Alternate types of caps are covered elsewhere.
 
Now Doug, don't forget that Tantalum is an electrolytic capacitor, just not an "aluminum electrolytic". I haven't looked at the AVX series, but if designed for audio, maybe they've figured out how to tame the bad sound that Tants have a reputation for. They've always had "low leakage" as one of their benefits. I've been of the mindset to replace tants with film or UKL BECAUSE of the low leakage inherent in these two types, perhaps now that UKL have gone away (in the process) it might be time to "reverse the equation".
 
I've never needed a low leakage aluminum electrolytic with a high value. It's more often coupling caps in the 1-47 uF range. The data sheet is certainly the place to start, but you just have to measure. Some ordinary parts can be excellent, or lousy. DC leakage also can increase with age. Or not. I have parts that have been hanging around for a decade and some series are ready for the trash; others are as good as the day they shipped. Low voltage parts like 10V aren't to be trusted. Can't escape measurements!
 
ConradH said:
I've never needed a low leakage aluminum electrolytic with a high value. It's more often coupling caps in the 1-47 uF range. The data sheet is certainly the place to start, but you just have to measure. Some ordinary parts can be excellent, or lousy. DC leakage also can increase with age. Or not. I have parts that have been hanging around for a decade and some series are ready for the trash; others are as good as the day they shipped. Low voltage parts like 10V aren't to be trusted. Can't escape measurements!
Click to expand...
You should also check the Kemet Tm series
My tech used the tm series in a accuphase tuner
 
Kemets are good and I'm not as afraid of tantalums as some are, but you still have to pay attention to the necessary derating. They have a table for that in the datasheet and it's the typical 50%. Sansui and others have used tantalums on the inputs of some stages, especially FETs, where they need really low DC leakage. Tantalums have almost no tolerance for reverse voltage and, if exposed to even low levels, will develop high leakage. I tend to use them in test equipment where I want to rebuild it once and never again!
 
I'm so confused now, I thought low ESR was important for coupling caps hence why film caps sound better than elects. I did see uvr, which as I understand, is a general purpose cap nothing special, in a Marantz cd48 recently that surprised me and I have heard that PW which is more special sounds quite good but the leakage figures above suggest it's worse. I put Panasonic FC in the signal path in one player as it's supposed to be good, and it was metalic initially, it gradually got better but I ended up taking it out. Some on the internet say its great others not so much.

Theres another older thread on this subject here
 
Last edited:
John3655 said:
I'm so confused now, I thought low ESR was important for coupling caps hence why film caps sound better than elects. I did see uvr, which as I understand, is a general purpose cap nothing special, in a Marantz cd48 recently that surprised me and I have heard that PW which is more special sounds quite good but the leakage figures above suggest it's worse. I put Panasonic FC in the signal path in one player as it's supposed to be good, and it was metalic initially, it gradually got better but I ended up taking it out. Some on the internet say its great others not so much.

Theres another older thread on this subject here
Click to expand...
The audio hobby is specifically designed to exploit all our biases and sensory deficiencies and drive us crazy! DC leakage is important only in coupling caps if it affects the bias of the next stage. DC leakage is time and voltage dependent so it's best to just have it low. Any film cap will be extremely low. Most new electrolytics are low, regardless of series, but if you need the lowest you need a special series. Tantalum is very low but tantalums can also have some weird sonic problems and are generally avoided in signal applications.

The importance of ESR depends on the circuit but it's rarely an issue. It's especially not an issue in 50/60 Hz power supply circuits. It's usually not an issue with coupling caps because the circuit impedances are so much higher than the miniscule ESR value. [broken record rant on] ESR should be used mostly by switching power supply designers, where it's the big factor in choosing filter caps. It's usually measured at 100 kHz. At very high frequencies the reactance of moderate value caps is low enough to be ignored and ESR determines the losses. Everybody else should be talking about Dissipation Factor. Aluminum electrolytics are typically measured at 120 Hz because in power supply applications they see 2X the line frequency. [broken record rant off]

IMO, the information you want about coupling caps isn't usually in a standard data sheet. That would be graphs of linearity and losses over the full audio range and beyond. Fortunately, most small AEs perform well there. I try to use caps with a dissipation factor well under 0.1 and with voltage ratings over 25V, even for very low voltage circuits. I like to see DC leakage under 2 uA for coupling caps. Do that and it will be very difficult to tell the difference between any of them! (Yes, I measure everything coming out and everything going in.)
 
Conradh thank you very much for explaining.

Trawling the datasheets
Range, Leakage, tan
Nich KZ 0.01CV 0.12
Nich KA 0.03CV 0.18 (for audio)
Nich FG 0.01CV 0.14
Nich PW 0.03CV 0.14
Pan FC/FR 0.01CV 0.14
Nich VR 0.03CV 0.16
Elna RFS Silmic 0.01CV 0.10

So for these measurements from the datasheets, Silmic wins followed by KZ, the general purpose VR is better than the audio specific KA? Both Panasonic FC and FR are equal and as good as FG. As you say tends to suggest there must be other factors outside the datasheets.
 
Last edited:
Though the line isn't as extensive as Nichicon UKL, Rubycon TWL part numbers are all still showing active and in stock at Digi-Key. There is still quite a bit of stock of UKL available as well.
 
You must log in or register to reply here.
Back
Top Bottom