I have Google searched without luck for the formula ( or even a good rule of thumb ) for how DC leakage current varies with temperature for common electrolytic capacitors - ie the usual 85C and 105C types.
My own testing suggests it is strongly positive and increases by a factor of
Anyone help ?
....... Phil
It doesn't seem to enter into most design guide;lines, or even models.
The old (2004) Rifa app notes ( page 13 - not in new app notes ) there's no formula given, just a table:
temp C mult 40 1 85 5
105 20 125 100RL
Some data here (in the front section)
Best regards, Spehro Pefhany
"Spehro Pefhany" "Phil Allison"
Some data here (in the front section)
Worth anyone's time to read.
Just the info in pages 6 to 15 is all.
The " bathtub curve" is a treat !!
...... Phil
rule
There's strong history dependence; after storage, the oxide can perforate, and a quarter hour at max temperature and voltage makes it heal. The 'healing' process generates lots of nonrepeatable leakage current.
So consistent measurement requires a temperature/bias prestress cycle (I've used a curve tracer and hot air blower). After that, the measured leakage might be meaningful . Manufacturer's data on leakage is ... sketchy at best. There aren't even any limits except at 25 C, in the datasheets I see.
"whit3rd" "Phil Allison"
" Manufacturer's data on leakage is ... sketchy at best. There aren't even any limits except at 25 C, in the datasheets I see. "
** The DC leakage figures I see published are ambiguous or just plain wrong - particularly in relation to the temperature.If you apply the formula quoted ( ie 3.sq.rt CV uA) and that is for 25 C - then allow the leakage to rise by a factor of 5 or possibly 10 at max temp of 85 C or 105C - a high voltage cap will be dissipating several watts & EXPLODE.
( note " C " in the above formula is in uF )
....... Phil
Have something to add? Share your thoughts — no account required.
Ask the community — no account required