Hi,
Just bought some brand new Nippon Chemi-Con electrolytic capacitors from Mouser. Is a KMZ series, 470uF 420V.
I am surprised that all the 4 pcs have value 412....414uF. So is below with 13% from nominal value. I agree there is on +/-20% limit, but being brand new, I expected to be closer to the nominal. I tried to apply voltage for 24 hours and still the same, no change. I measured it with a RLC meter.
The date code is 1NT058. After some research, I found that the first digit represents the year of manufacturing, so is 2001, 2011, 2021....??
Any comment on this?
Just bought some brand new Nippon Chemi-Con electrolytic capacitors from Mouser. Is a KMZ series, 470uF 420V.
I am surprised that all the 4 pcs have value 412....414uF. So is below with 13% from nominal value. I agree there is on +/-20% limit, but being brand new, I expected to be closer to the nominal. I tried to apply voltage for 24 hours and still the same, no change. I measured it with a RLC meter.
The date code is 1NT058. After some research, I found that the first digit represents the year of manufacturing, so is 2001, 2011, 2021....??
Any comment on this?
Welcome to the modern world, this is not just electrolytics, plastic film caps also suffer from this nowadays, my guess is that process technology have made it possible to control the capacitance very precisely so always being in the low end of the tolerance span saves a couple of turns of foil for each capacitor made, since the large manufacturers make millions of capacitors each year this is a significant saving for them.
Last edited:
Pressure on the margins....
Stockholders that want more....
Extremely tight competition....
Stockholders that want more....
Extremely tight competition....
The value is well within spec. Not sure why anyone would expect a higher value.
Electrolytic capacitors used to be -20/+50% in the past and in reality they were practically always higher in value. In fact it was suspicious when they were spot on or in the -20% region.
So that stays in the mind of most just like tantalum caps always being unreliable, SMD being inferior to TH, LM317/LM337 being best standard regulators etc.
So that stays in the mind of most just like tantalum caps always being unreliable, SMD being inferior to TH, LM317/LM337 being best standard regulators etc.
Last edited:
Being on the + side left a good feeling about the durability, with the decrease of capacity over time, that left still plenty of the originally intended capacity available. Unless the intended lifespan is only 10 years anyway. Then one might not care.
But being a genuine Nippon ChemiCon, from oficial distributor I would expected better tolerance.
Last edited:
The lower tolerance together with better specifications could mean that the traditional drawbacks of electrolytic caps have been partly overcome.
Last edited:
The only thing remaining is for designers, for example, what remaining ripple to calculate, when your cap is specified at 100 microfarad. Should one calculate for 95 microfarad average, or 100?
If you want a tighter tolerance, then you pay more for it. You already stated the caps you bought are within tolerance.
jeff
Still don't know how old are.
Date code :1NT058
The first digit should be the Year, second the month. So is from November but 2001.....2011....2021 ??
Date code :1NT058
The first digit should be the Year, second the month. So is from November but 2001.....2011....2021 ??
Just bought some others , Nichicon this time, same distributor. 470uF 350V.
Measured value is around 387 ....408uF so some is very close to the lowest limit. .
Measured value is around 387 ....408uF so some is very close to the lowest limit. .
Maybe they are older than expected. In my cap collection there are a number of older electrolytic capacitors from various well-known manufacturers, all from official distributors, which have all lost capacity over the years (10 years and more).
I'm impressed all four are at -12% (412uF). Would you feel better if two were +12% (526uF)? If you checked enough of them, a statistician could tell if they sorted out the precise ones for their Audio-File Label caps.
In the Tokyo subway you can sometimes see Nichicon ads...
In the Tokyo subway you can sometimes see Nichicon ads...
I recall (circa 1970 and the cap may have been 20+ years old back then) seeing an electrolytic with tolerance +80/-20. Manufacturing tolerances apparently have gotten better since then.
This happens (or so I've heard) with all components. One might order a bunch of 2% resistors, maybe expect them to be equally distributed from -2% to +2%, but they may be missing all values between -1% and +1%, as they may be measured and higher precision removed and sold for a higher price. The guarantee is that they're within the tolerance limits, it says nothing about what the distribution of values in between might be.
This happens (or so I've heard) with all components. One might order a bunch of 2% resistors, maybe expect them to be equally distributed from -2% to +2%, but they may be missing all values between -1% and +1%, as they may be measured and higher precision removed and sold for a higher price. The guarantee is that they're within the tolerance limits, it says nothing about what the distribution of values in between might be.
For ALL designs, using the worst-case values on the data sheet is the only way to design. Anything else is relying on parts being BETTER than what the data sheet guarantees. I saw Bob Pease say this about 24 years ago - I almost couldn't believe it, it seemed obvious to me, but clearly many younger engineers don't know that only the Min and/or Max datasheet values are guaranteed.
It was on DC 315V for 24 hour. The value increased from 387 to 390uF. Anyway...that's the life 🙂