Charging super capacitors

[Pixel5]

Since yesterday i have a pack of 6 Supercapacitors that i soldered onto a balancing board that also avoid them being overcharged.

The capacitors are rated for 2,7V and 500F, after being put in series this leaves me with up to 16,2V and 83,33F

 

Now the big questions, whats the best way to charge them without destroying my charging equipment with the high current they can draw?

I already tried to wire them up to a 12V battery but without a power resistor which i dont have on hand right now the current draw will melt any wire i have here and also puts the 12V battery at its limit.

 

any ideas how i can charge them without getting too high currents? 

[MrMarriarty]

any car battery charger will do the job nicely, it will have a fuse if it draws too much

[cj09beira]

you could try a dc to dc boost converter as they usually have a current limiter, or a lab bench power supply, 

[iamdarkyoshi]

I'd suggest sourcing a power supply with both adjustable contant current and adjustable constant voltage. You could also use a car battery as a source and put a 12v lightbulb in series, which will act as a current limiting resistor.

[Pixel5]

1 hour ago, cj09beira said:

you could try a dc to dc boost converter as they usually have a current limiter, or a lab bench power supply, 

already thought about that dc to dc converter as well but the ones i got done have current limiting, they will just overheat and fall apart.

[Unimportant]

5 hours ago, Pixel5 said:

Since yesterday i have a pack of 6 Supercapacitors that i soldered onto a balancing board that also avoid them being overcharged.

The capacitors are rated for 2,7V and 500F, after being put in series this leaves me with up to 16,2V and 83,33F

 

Now the big questions, whats the best way to charge them without destroying my charging equipment with the high current they can draw?

I already tried to wire them up to a 12V battery but without a power resistor which i dont have on hand right now the current draw will melt any wire i have here and also puts the 12V battery at its limit.

 

any ideas how i can charge them without getting too high currents? 

Even tough its a supercapacitor, it's still just a capacitor, the charge is rather small compared to a battery, you won't need special equipment to charge it.

Let's do the math, to charge a capacitor with a fixed current:

dT = dU * C / I

dT = delta time
dU = delta voltage
C = capacitance
I = current

for dU = 12V and C = 83.33F, charging at 1 amp we get:

12 V * 83.33F / 1 = 1000 seconds or 16.6 minutes

So you can charge it in about 15 minutes with 1 amp, only 5 minutes with 3 amps...

These are relatively low currents any decent linear (bench) power supply should be able to deliver, no need for switching DC/DC converters. A car battery charger, typically capable of delivering far greater currents will charge them in a handfull of seconds.

[Pixel5]

1 hour ago, Unimportant said:

Even tough its a supercapacitor, it's still just a capacitor, the charge is rather small compared to a battery, you won't need special equipment to charge it.

Let's do the math, to charge a capacitor with a fixed current:

dT = dU * C / I

dT = delta time
dU = delta voltage
C = capacitance
I = current

for dU = 12V and C = 83.33F, charging at 1 amp we get:

12 V * 83.33F / 1 = 1000 seconds or 16.6 minutes

So you can charge it in about 15 minutes with 1 amp, only 5 minutes with 3 amps...

These are relatively low currents any decent linear (bench) power supply should be able to deliver, no need for switching DC/DC converters. A car battery charger, typically capable of delivering far greater currents will charge them in a handfull of seconds.

its not really how long it will take its more a problem on getting way too much current, specially when i use the 12V battery i get from nothing to red glowing battery terminals in a few seconds.

 

guess for the best way to charge them i should get a bench power supply.

[Unimportant]

1 minute ago, Pixel5 said:

its not really how long it will take its more a problem on getting way too much current, specially when i use the 12V battery i get from nothing to red glowing battery terminals in a few seconds.

 

guess for the best way to charge them i should get a bench power supply.

Don't you have a charger for that 12V battery ? Charge the caps with that - it'll have current limiting, because that's (part of) how you charge batteries.

[Pixel5]

55 minutes ago, Unimportant said:

Don't you have a charger for that 12V battery ? Charge the caps with that - it'll have current limiting, because that's (part of) how you charge batteries.

currently no charger in place as the one i had before just stopped working 

[mariushm]

You could just use a basic linear regulator like LM1085 for example.

LM1085 is designed for maximum 3A and has an internal current limit which varies from part to part, but it's guaranteed to be within 3.2A .. 5.5A .. so worse case scenario you'd cap the current at 5.5A

 

Of course, you can wire two regulators in parallel and then you have max 6A and ~ 6.,4..11A current limit, or you can use one as pre-regulator in current mode to limit the current (see datasheet)

 

[Unimportant]

40 minutes ago, mariushm said:

You could just use a basic linear regulator like LM1085 for example.

LM1085 is designed for maximum 3A and has an internal current limit which varies from part to part, but it's guaranteed to be within 3.2A .. 5.5A .. so worse case scenario you'd cap the current at 5.5A

 

Of course, you can wire two regulators in parallel and then you have max 6A and ~ 6.,4..11A current limit, or you can use one as pre-regulator in current mode to limit the current (see datasheet)

 

An alternative would be a LM317 (1.5A max), which is so common you should be able to buy one everywhere.

A lot of power will be dissipated when the capacitor is empty tough (12V * 1.5A = 18W), you'd need a bulky heatsink.

A el-cheapo car battery charger would probably be cheaper/less hassle.

[Pixel5]

i found an old laptop power supply that i wired up to charge the capacitors.

 

takes forever because when the current exceeds 2.1A the charger start pulsing the charge instead of continuous charge.

It works for now but its kind of slow.

 

i also already completely decharged the capacitors with my dummy load and found out the capacity is only 260mah at a voltage dropping from 16.2V down to almost zero.

That turned out to be only 2.13Wh but theyre capacitors so thats to be expected.

[Unimportant]

1 hour ago, Pixel5 said:

i found an old laptop power supply that i wired up to charge the capacitors.

 

takes forever because when the current exceeds 2.1A the charger start pulsing the charge instead of continuous charge.

It works for now but its kind of slow.

 

i also already completely decharged the capacitors with my dummy load and found out the capacity is only 260mah at a voltage dropping from 16.2V down to almost zero.

That turned out to be only 2.13Wh but theyre capacitors so thats to be expected.

Well, the math did say you'd be able to charge them in about 15 minutes @ 1Amp. That works out to 250mAh. Yeey math

[Pixel5]

On 2.9.2017 at 6:00 PM, Unimportant said:

Well, the math did say you'd be able to charge them in about 15 minutes @ 1Amp. That works out to 250mAh. Yeey math

next math problem for you

 

how big of a cable cross section in square mm will i need if i want the cable to hold up to about 300A without melting the isolation?

[mariushm]

Depends on how long you're gonna have that current, how long the wire will be, and what insulation you're gonna use on the cable and how much you're comfortable with when it comes to heating up the wires.

 

There's cable insulation which starts melting at 100c and there's insulations which can do 200-250c or even more (impregnated fiberglass for example)

Then if we're talking about pulses of a few seconds then you could use smaller gauge because wire wouldn't have time to reach critical temperature

 

You can start with https://en.wikipedia.org/wiki/American_wire_gauge#Tables_of_AWG_wire_sizes

 

[Pixel5]

1 hour ago, mariushm said:

Depends on how long you're gonna have that current, how long the wire will be, and what insulation you're gonna use on the cable and how much you're comfortable with when it comes to heating up the wires.

 

There's cable insulation which starts melting at 100c and there's insulations which can do 200-250c or even more (impregnated fiberglass for example)

Then if we're talking about pulses of a few seconds then you could use smaller gauge because wire wouldn't have time to reach critical temperature

 

You can start with https://en.wikipedia.org/wiki/American_wire_gauge#Tables_of_AWG_wire_sizes

 

seems like i can get away with using 4mm² wires to connect some handles for spot welding.

Probably should build a circuit that pulses the output for like 1-2 sec max on button press to avoid problems when i cant open the circuit for some reason.

 

195A Transistors are already on the way, i guess ill use at least two of them and should be good to go.

[mariushm]

You can see here how it's done

 

1. https://www.eevblog.com/forum/reviews/sunkko-709a-spot-welder-teardown/

2. https://www.eevblog.com/forum/testgear/battery-tag-welder-from-china/

 

Note that more modern spot welders don't use DC pulses but rather some specific waveform that results in better quality of the welds and safer to spot weld on batteries at same time.

 

[Pixel5]

11 minutes ago, mariushm said:

You can see here how it's done

 

1. https://www.eevblog.com/forum/reviews/sunkko-709a-spot-welder-teardown/

2. https://www.eevblog.com/forum/testgear/battery-tag-welder-from-china/

 

Note that more modern spot welders don't use DC pulses but rather some specific waveform that results in better quality of the welds and safer to spot weld on batteries at same time.

 

seems like they are just directly using the AC that is coming in stepped down to decrease voltage while increasing the current.

And full bridge rectifier would need to be way too beefy to hold up to that current so they just gone straight AC for the weld.

 

as it seems they go from 110V down to like 12 or 13V so i doubt they get more than 200A on the output so it should be good with that capacitors.

Ill try out if using AC to spot weld actually makes a difference and try to create my own AC with some circuit magic.

[Unimportant]

3 hours ago, Pixel5 said:

seems like i can get away with using 4mm² wires to connect some handles for spot welding.

Probably should build a circuit that pulses the output for like 1-2 sec max on button press to avoid problems when i cant open the circuit for some reason.

 

195A Transistors are already on the way, i guess ill use at least two of them and should be good to go.

If it's for spot welding you might want to opt for some thicker cables. Even tough 4mm² cable probably won't have time to heat up much for a spotweld, they might drop a significant amount of voltage at those current levels. That's less voltage arriving at the weld spot, wasting potential. Flexible 10mm² cable should be do-able.

Material also plays a huge role, go for quality copper wires, not cheap wires with aluminum mixed in.

 

What do you mean with "195A Transistors" ? Do you mean MOSFETS capable of 195Amps ? If yes, then take care to drive them properly with a high current gate driver. Even tough a MOSFET's gate is high impedance, it acts like a capacitor. At the currents you're talking about, you need to turn the MOSFET on and off FAST. If the transition between on/off or off/on takes too long the MOSFET will dissipate a lot of heat during the transition and possibly die. In order to turn a MOSFET on or off fast a gate driver with high current capability is required so the "gate capacitor" can be quickly charged (turn on) and discharged (turn off). Especially so for high current MOSFET's as they probably have a high gate charge. (and off course, multiple MOSFET's in parallel also increases the gate charge). A proper PCB layout which minimises parasitics around the MOSFET and driver will also be needed as otherwise the gate can ring too much. Ringing will also make the MOSFET dissipate a lot of heat as it's not cleanly turning on/off but oscillating between on and off.

 

Example of ringing:

https://cdn.hackaday.io/images/1633061463951942352.png