Ultracapacitors / Supercapacitors in series to attain required voltage

[leftcliq]

Hi Guys,
Similar to other threads, my power supply is 50V, can I use 10 5.5V supercapacitors in series to store this energy. I know that in series, capacitance is divided by the number of caps. Just want to make sure if I go to the expense of buying these caps, that I won't blow them up immediately. If I am on the right track, does anyone know where I can source these supercaps for a good price? I worked out, I will need 100F for each supercap as 100F/10 Caps = 10F.
Regards,
Dale.

 

[phinds]

Your sizing is correct but I don't know where to get them. Surely, if the are available that large, you can find them on the internet.

Also, you'll want to be careful how you charge them since that's going to start out as a huge current sink on your voltage source and could damage it unless it is properly protected against overcurrent.

 

[leftcliq]

Hello Phinds,
Thanks for your quick response. I thought that put the caps in series with balance resitors in parallel would do the trick. In my experience with caps in series, the first capacitor gets all of the energy and the last one in the series gets very little. Don't know if this is the same with super caps.

 

[leftcliq]

Will the super caps handle a charge of 0.3 second and the same discharge?

 

[phinds]

Hello Phinds,
Thanks for your quick response. I thought that put the caps in series with balance resitors in parallel would do the trick.

Sounds right.

In my experience with caps in series, the first capacitor gets all of the energy and the last one in the series gets very little. Don't know if this is the same with super caps.

Hm ... that's not what I would expect. That seems to imply that current flows through the first cap at a different rate than through the last cap and that's not possible.

 

[leftcliq]

Glad to be corrected. Do I need balancing resistors if the caps are in series?

 

[phinds]

Glad to be corrected. Do I need balancing resistors if the caps are in series?

You need a series resistor when doing the charging otherwise the power supply will think you have put a buss bar directly across it. This would make the power supply very unhappy :w At the very least it would probably blow a fuse and if it could talk it would likely say very unkind things to you.

 

[leftcliq]

Sounds like I should give the power supply the name of an ex girlfriend. I assume there is a formula for the resitance?

 

[phinds]

Sounds like I should give the power supply the name of an ex girlfriend. I assume there is a formula for the resistance?

I'm going to have to leave that to someone more currently in tune w/ electrical engineering. I haven't done that stuff in decades and my memory isn't what it used to be. In fact my memory was NEVER what it used to be but don't tell my children. :p

 

[leftcliq]

I have been reading, instead of using resistor balancing, I could add in an extra cap. This way the 50V will be spread across the 11 caps in series, 50/11 = 4.54 which is less than the 5.5V capacity. Hope I am reading this correctly. So it would go to 82% capacity across each cap.

 

[phinds]

I have been reading, instead of using resistor balancing, I could add in an extra cap. This way the 50V will be spread across the 11 caps in series, 50/11 = 4.54 which is less than the 5.5V capacity. Hope I am reading this correctly. So it would go to 82% capacity across each cap.

You are right that this would be better for the caps, but it would in no way change the unhappiness of the power supply, which is going to be annoyed w/ you if you do not use a resistor to decrease the initial current surge.

 

[leftcliq]

Thanks for that, yes I will put a resisitor on series to decrease the initial current surge. I'm planning to use a 5watt or 10watt resistor to disapate any heat created.

 

[phinds]

Thanks for that, yes I will put a resisitor on series to decrease the initial current surge. I'm planning to use a 5watt or 10watt resistor to disapate any heat created.

That level of wattage won't matter much if the resistance is too small ... it will likely explode. If you use a really large resistor, the wattage won't matter much but the charge time will be very long. You need to do the math and come up with a reasonable compromise that won't take to long to charge but also won't blow a power supply fuse and won't overheat the resistor.

 

[zoki85]

You are right that this would be better for the caps, but it would in no way change the unhappiness of the power supply, which is going to be annoyed w/ you if you do not use a resistor to decrease the initial current surge.

With current surges supercaps will not be happy either. Can kill them easily.

Thanks for that, yes I will put a resisitor on series to decrease the initial current surge. I'm planning to use a 5watt or 10watt resistor to disapate any heat created.

Wasting a lot of power in a form of heat ain't good. Why not to use inductor+diode in series to charge the caps? You can find smoothing chokes at reasonable price ($10<) on ebay.

 

[phinds]

With current surges supercaps will not be happy either. Can kill them easily.

I made that same point here regarding regular caps in a thread quite some time back and was roundly poo-pooed. Are supercaps different in that regard?

 

[zoki85]

They store much more energy and have higher energy density than classical caps. In conjunction with their low ESR, too fast charging can damage them electrothermally.

 

[phinds]

They store much more energy and have higher energy density than classical caps. In conjunction with their low ESR, too fast charging can damage them electrothermally.

Thanks for that clarification.

 

[leftcliq]

Hi,
Is there a circuit diagram that shows the use of inductor+diode to charge the caps? Is this instead of a balancing resistor? The idea is to charge the caps up quickly (but not quickly for them to explode) and to discharge quickly. Would Zener diodes be more useful in this instance?

 

[zoki85]

Hi,
Is there a circuit diagram that shows the use of inductor+diode to charge the caps? Is this instead of a balancing resistor? The idea is to charge the caps up quickly (but not quickly for them to explode) and to discharge quickly. Would Zener diodes be more useful in this instance?

This is basic circuit. Condition 2√(L₁/C₁) >R₁ should be met in order to save some energy. Since your supercap has huge capacity, you'll have a difficulty to find choke big enough, with small enough wire resistance R₁. One modification of this circuit is to employ two-stage resonant charging with significantly higher DC voltage source and smaller cap C₀<< C₁ on first LC stage. Then energy in C₀ is transferred to C₁ via circuit C₀-L₁-C₁-R₁. If you don't have experience/knowledge in designing such circuits, you better don't do it. Overvolting supercap means sure death it. I don't believe you have low voltage DC source capable of delivering currents like 1 kA, and charging cap too quickly won't happen. Charging them by simple series resistor that can withstand dissipation level works ok. Discharging fully charged supercap is more problematic. Too quickly (short circuit blast) isn't good for lifetime of the cap. Some SCs can withstand such abuse many times, some not.

 

[NascentOxygen]

I worked out, I will need 100F for each supercap as 100F/10 Caps = 10F.

How did you determine that your 50V power supply will need 10F capacitance? What will it be powering?

 

[zoki85]

Condition 2√(L₁/C₁) >R₁ should be met in order to save some energy.

Just noticed this isn't correct what I wrote . The correct condition should be √(L₁/C₁) > 2R₁.