Electrolytic Filter caps.... brand new but out of spec already! What the heck?

5F6-A

Member
Messages
2,589
I have bought 2 Nichicon 47 uF 450V caps.


One reads 57 uF and the other 60 uF. I knew that electrolytic caps were a bit less strict in terms of tolerance but this seems excesive. What do you think? What'd you do?

Looking at the spec sheet tolerance is 20% (i.e. 37.6 uF- 56.4 uF)..... :(
 
Last edited:

VICOwner

Silver Supporting Member
Messages
1,649
I don’t worry much about them when they test high. I know it’s nice for things to be at their ratings but consider that if you are testing them with a DVOM that it’s charging them up with it’s battery voltage, maybe 9V, and clocking the charge time to determine uF rate. This is a test for sure, but you are about to use them at near their voltage rating, hundreds of volts. The accuracy of DVOM testing high capacitance is negligible. Just be glad they reached their ratings.
 

5F6-A

Member
Messages
2,589
I don’t worry much about them when they test high. I know it’s nice for things to be at their ratings but consider that if you are testing them with a DVOM that it’s charging them up with it’s battery voltage, maybe 9V, and clocking the charge time to determine uF rate. This is a test for sure, but you are about to use them at near their voltage rating, hundreds of volts. The accuracy of DVOM testing high capacitance is negligible. Just be glad they reached their ratings.
Helpful answer bro.
 

VICOwner

Silver Supporting Member
Messages
1,649
I also have to set my meter to delta mode when testing pF caps. My leads or whatever it is read enough hooked up to nothing and will make a 47pF or 100pF nearly double.
 

zenas

Member
Messages
8,136
It's also fun to test old electrolytics with a DVOM. Most of them test pretty close in UF even with bulges and stuff oozing out.
Unfortunately some people actually trust their readings and don't change 50 year old caps. Then other things get blamed for issues, like speakers for farty bass. ;)
 

5F6-A

Member
Messages
2,589
The meter I have here (Fluke 77 iv) can't read small capacitance (pF). Starts with 1nF.... very annoying as this is an expensive unit. Am I doing something wrong?
 

VICOwner

Silver Supporting Member
Messages
1,649
Does it also have delta? Looks like a triangle, it should. Try it on a 250pF. It will read .250nF. Delta should zero out the meter, then hook up to the capacitor.
 

Pete Cage

Member
Messages
1,420
The meter I have here (Fluke 77 iv) can't read small capacitance (pF). Starts with 1nF.... very annoying as this is an expensive unit. Am I doing something wrong?
On the lowest range (1000nF) a Fluke 77 only has 1.2 nF resolution. I wouldn't expect it to be measuring pF very well.
 

rolandk

Member
Messages
476
The meter I have here (Fluke 77 iv) can't read small capacitance (pF). Starts with 1nF.... very annoying as this is an expensive unit. Am I doing something wrong?
I say this as a 30 year audio electronics repair tech and in the nicest way possible... don’t worry about it. Buy new high quality caps, install, and move on with your life. The odds of it being defective is extremely low. The only real cap issue is electrolytics start to fail when they get old and/or heat stressed.
 

5F6-A

Member
Messages
2,589
I say this as a 30 year audio electronics repair tech and in the nicest way possible... don’t worry about it. Buy new high quality caps, install, and move on with your life. The odds of it being defective is extremely low. The only real cap issue is electrolytics start to fail when they get old and/or heat stressed.
Cheers buddy!
 

killer blues

Member
Messages
3,020
I have bought 2 Nichicon 47 uF 450V caps.


One reads 57 uF and the other 60 uF. I knew that electrolytic caps were a bit less strict in terms of tolerance but this seems excesive. What do you think? What'd you do?

Looking at the spec sheet tolerance is 20% (i.e. 37.6 uF- 56.4 uF)..... :(
Were they tested at operating voltage? That's all that matters.
 

teemuk

Member
Messages
3,051
Are you filtering a power supply or building a precision electronic device?
This was a rhetoric question.
 

pdf64

Member
Messages
6,856
On the face of it, yeah, the golden age typical power supply cap value spec of +100 to -50% is probably fine for our application.
But why publish specs if you can’t achieve them? I see it as being akin to Van Halen’s ‘no blue sweets’ contract rider; if blue sweets were provided, it was a warning that the venue hadn’t verified that every clause of the contract had been checked / fulfilled; hence key things, like the capacity of the power supply, would need to be verified.

So if (assuming the cap measurement is accurate) the manufacturer’s QA system has allowed an item to pass despite it failing on something that’s easily identifiable, how can the customer have confidence in the manufacturer’s processes? (Of course even a well controlled 6 sigma process will produce a few ppm outliers, and with cheap parts we can only expect sample, not 100%, testing.)

Hence my query about the part’s provenance.
 




Trending Topics

Top