supercapacitor ESR?

[coppercone2]

I have a 2000F super capacitor. I don't know how old it is etc, so I slowly charged it to 2.7V (its rating) with 1A current (did 500mA for 0 to 1V).

Using the time formula with 1V I got something along 1950F capacity (32.5 min @ 1A for 1V to 2V). However when I hooked it up to a rapid discharge circuit (spot welder), I am only getting 15A out of it. The rating claims 0.5 mOhm ESR.

I can leave these for a long while connected to a resistor to discharge them, and maybe measure ESR with de5000 (is this appropriate? scared I will fry meter. I should have measured ESR before I charged em up first time.. but it will be slowly decaying on the 10 ohm load for a long time).. but I am wondering for stuff thats been on the shelf for a long time is there any kind of restoration procedure.. like hooking it up to rapid charge or something (not 'trickle' charging it). Would a few cycles help... or am I just wasting my time? Never worked with supercapacitors before. I thought maybe like 50A charge current might 'electrolytically clean' something inside.

I thought to short it out on a clamp meter or something but I need to hook up some wiring (kind of scared its going to explode on me). My wiring is OK I measured it to be around 900mOhms per cable, less then 3mOhm total. Only other thing that comes to mind is terminal corrosion on the cap, its aluminum...

Is it possible to see such a huge ESR increase while measuring capacitance ok? Or is that capacitance measurement method flawed? Thought I was good when I got the high farad number from the charge time test. For that I hooked up a multimeter to a power supply, current limit to 1A and waiting for voltage to rise to 2.5V with a stopwatch.

[inse]

Some time ago, I was given a set of supercaps from an automotive starter boost application.
I figured out what fancy stuff to do with it, but had to find out, that the ESR was too high to get more than 150A out of it, iIrc.
Probably they had degraded.
They were too bulky for a flashlight as well, so in the end, I tossed them.

[jonpaul]

bonjour

you have mixed up storage capacity, short circuit current and ESR.

ESR is a small Signal AC parameter measured on a meter or bridge and is the delta V/ delta I

in power supply bulk capacitors it affects  the ripple voltage

Storage capacity of a battery or super capacitor  in ampère hourss is rated at a modest load, eg AH/10.

At high load current the  internal DC résistance ( not ESR) decrease the current from theoriical

Bon courage


jon

[coppercone2]

so is the DC surge/short resistance a parameter that can be some how revived? I don't want to buy this shit, these are expensive.

[Marco]

ESR is a small Signal AC parameter measured on a meter or bridge and is the delta V/ delta I

in power supply bulk capacitors it affects  the ripple voltage

He's only putting a small short duration load on it with a spot welder.

Much shorter duration and for a smaller percentage voltage drop than you'd see on a bulk capacitor in fact. Even though I disagree with it, with your reasoning ESR is exactly the specification he should be concerned with.

[T3sl4co1l]

ESR depends on frequency (or if you like: discharge rate).  That's why it's equivalent!

Supercaps have a substantial diffusion effect, i.e. ESR and C are proportional (or inversely) to sqrt(f) over a wide range of frequencies (<mHz to Hz).

Batteries also have a lot of diffusion effect, so their ESR varies in a similar way, on top of depending on state of charge.

Tim

[coppercone2]

well ok, I think this is the problem. I discharged it over night on 10 ohm load and then short circuit it for 1 hour with a alligator clip and measured it on DE-5000 meter. I got 0.08 Ohms t 120Hz, which is 80 Miliohms, which is 160 times the specification. At the low voltage the math shows that the current would infact be limited to like the tens of amps with this figure with 2 in series with similar ESR. I am not used to working with such low numbers so I must have ignored it at first glance when I spot checked it years ago.


So I guess its not possible to reduce ESR and that means they bad. Never heard about restoring capacitor ESR.

[Marco]

Supercaps have a substantial diffusion effect, i.e. ESR and C are proportional (or inversely) to sqrt(f) over a wide range of frequencies (<mHz to Hz).

The ESR for a short term high current discharge can not be out over 2 orders of magnitude from spec because of this. The differences for the DC and AC specs are more like 10s of percent.

For some reason his capacitor is way out of spec ESR wise, AC or DC it don't matter.

[coppercone2]

the other one reads 0.00 ohms on 120Hz but 0.07 ohms on 1KHz

They were also not impressive when I was sparking aligator clips, it basically reminded me of a D cell.

Should I bother trying to charge cycle them or does this have to do with dry electrolyte/unsolveable problems

[bdunham7]

the other one reads 0.00 ohms on 120Hz but 0.07 ohms on 1KHz

They were also not impressive when I was sparking aligator clips, it basically reminded me of a D cell.

Should I bother trying to charge cycle them or does this have to do with dry electrolyte/unsolveable problems

No, they simply won't work as you hope they will.  Maxwell made a replacement for automotive starting batteries using their 3000F supercaps and I've seen them fade away like this.  I think they turned out to be a lot less durable than plain LA batteries. You can buy these supercaps on the surplus market--I know someone who literally had a truckload--and while many are usable for some applications, what you are seeing in high-current uses is typical.  Capacitance and leakage often seem to be OK for a while.

[coppercone2]

Ok, I bought some new supercapacitors (eaton). At least I fed the trash can

160x out of spec.. I don;t want to keep them around. Seems like they get worse at any time. Doubt its region of stability?

[T3sl4co1l]

Supercaps have a substantial diffusion effect, i.e. ESR and C are proportional (or inversely) to sqrt(f) over a wide range of frequencies (<mHz to Hz).

The ESR for a short term high current discharge can not be out over 2 orders of magnitude from spec because of this. The differences for the DC and AC specs are more like 10s of percent.

For some reason his capacitor is way out of spec ESR wise, AC or DC it don't matter.

Actually I didn't specify the direction (let's see... ESR and C both drop with rising f, isn't it), though that was probably implied by the question, and offering such information in response.

OP didn't post the datasheet anyway, so, no way to tell if it's meaningful anyway.  This might be it:
https://www.mouser.com/datasheet/2/257/Maxwell_K2Series_DS_1015370-4-1179730.pdf
which uses a 5s pulse at 0A, following an unspecified discharge phase at specified current.  So, ESR should be similar under OP's conditions anyway.

Presumably, they're terribly out of spec due to age and/or abuse?

Tim

[741]

For testing with a normal ESR meter, maybe use a series 'normal' capacitor, whose ESR you have already measured?

[wraper]

On DE-5000 you can use resistance range to measure ESR of large capacitances which cannot be measured properly at high frequencies in capacitance mode or within circuit where normal capacitance measurement is disturbed.

[coppercone2]

yeah ac r is what i used but those caps are history now