What I haven't seen in this discussion is any mention of amount. COntext is everything.
Let us say we have a 10uf cap. Now either this low voltage effect happens or it does not, but assuming it does, what are we talking about? 10uf becomes 20? 10uf adds 10pf? Certainly that matters. If the cap doubles, I would consider that significant, but adding 10pf or 20pf, I will ignore.
It is all well and good to measure 30 year old 10uf caps and find them at 12uf, but they could easily have measured 12uf when new as well. Same thing when testing 63v caps at 80v. They probably worked at 80v off the shelf. Most caps don;t fail at a volt over their rating.
Let us say we have a 10uf cap. Now either this low voltage effect happens or it does not, but assuming it does, what are we talking about? 10uf becomes 20? 10uf adds 10pf? Certainly that matters. If the cap doubles, I would consider that significant, but adding 10pf or 20pf, I will ignore.
It is all well and good to measure 30 year old 10uf caps and find them at 12uf, but they could easily have measured 12uf when new as well. Same thing when testing 63v caps at 80v. They probably worked at 80v off the shelf. Most caps don;t fail at a volt over their rating.
Regardless, imho the order of failures in electronic gear is roughtly: solder/mechanical, caps, resistors, everything else.
Caps might be up there next to solder, close second.
They fail.
Electrolytics do last longer if they get charged up periodically. Folks who own or work with antique or "boatanchor" (ham radio) gear know this. Turn that thing on a few times a year and it will keep workin' - don't and it might not turn on next time.
Very technical comments, but...
_-_-bear
Caps might be up there next to solder, close second.
They fail.
Electrolytics do last longer if they get charged up periodically. Folks who own or work with antique or "boatanchor" (ham radio) gear know this. Turn that thing on a few times a year and it will keep workin' - don't and it might not turn on next time.
Very technical comments, but...
_-_-bear
The increase in capacitance with age probably isn't real. It has to do with measurement and the model used. Let's say you start out with a 100uF cap having a dissipation factor of .01, measured at 1kHz, series model. The parallel model will be almost identical.
Now, as the capacitor ages, the dissipation factor is likely to go up. Let's say it goes up to 0.5 (time to junk it in my opinion, but do as you will). The capacitance also goes up to 125uF. Convert to, or switch the LCR meter to the parallel model and what do you have? 100uF. In real life almost anything might happen, but seeing high values isn't that uncommon when servicing old equipment.
I've seen some really bad capacitors with normal or high capacitance readings, but they certainly don't have the filtering properties of good capacitors with the same value. IMHO, having a cap meter that doesn't measure dissipation factor is ok for matching and building new equipment, but pretty much useless for servicing old stuff. That's where a proper bridge, or a high frequency ESR meter shines.
CH
Useful impedance converter download on my site.
Now, as the capacitor ages, the dissipation factor is likely to go up. Let's say it goes up to 0.5 (time to junk it in my opinion, but do as you will). The capacitance also goes up to 125uF. Convert to, or switch the LCR meter to the parallel model and what do you have? 100uF. In real life almost anything might happen, but seeing high values isn't that uncommon when servicing old equipment.
I've seen some really bad capacitors with normal or high capacitance readings, but they certainly don't have the filtering properties of good capacitors with the same value. IMHO, having a cap meter that doesn't measure dissipation factor is ok for matching and building new equipment, but pretty much useless for servicing old stuff. That's where a proper bridge, or a high frequency ESR meter shines.
CH
Useful impedance converter download on my site.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.