What size capacitor needed to power solenoid?

I have an Arduino based project I’m doing for work. The project uses a 12vdc automotive starter solenoid to power a wire cutting shear. To the best of my calculations, the solenoid draws somewhere between 35 and 45 amps @ 12vdc. (coil resistance is .2 to .3 ohms). I’m using a solid state relay to energize the solenoid, but sadly, the relay is only rated for 8 amps; I’ll have to find another. Solenoid engagement time is <50ms. Power supply for the project is a dc to dc converter: input 24v, output 12v, 3.8amps. I’m wanting to supplement the power supply using a capacitor, perhaps a bass capacitor used for stereo systems. I’m not sure how to calculate how many farads of capacitance I’m going to need, any suggestions? Also, can someone suggest a source for dc to dc SSR’s (hockey puck based)? I thought Mouser would have them, but instead, they have DIN mount style units that are expensive. Right now, I have this project up and running on a test basis only, using the 8 amp SSR and a 20ah SLA battery, so far, my relay has performed well.


Approximately : t = RC

C=t/R = 0.05/0.2 = 0.25 Farad. That is pretty big value for capacitor, 250 000 uF. Electrolytic type will works, just limit charging current with resistor. Probably, current will be less than you expect as V/R, as load inductive, so less capacitors needed.

Magician, Okay, thanks for the formula. Now, can I use a motor start capacitor or maybe one used to start air conditioner compressors? I know these are rated for AC use, but are they not constructed in the same manner as bass capacitors? Their cost is much less...

Thanks, Cris

Gigundo Capacitors are widely available: http://www.sonicelectronix.com/cat_i6_-capacitors.html etc etc

You could use this kind of power FET board with 4 - IRF540 FETS ($13.50) and just parallel the channels for extra margin:

And they are optically isolated, so you could mount this as close to the solenoid / capacitor combination as possible and keep the Arduino farther away…

DISCLAIMER: I mentioned stuff from my own Shop… But I like finding solutions

I guess you mean capacitor in air conditioner that help to start three phase motor when
it plug on one phase?
Afraid its not good, 1-10 uF what I’ve seen so far.
250 000 uF should not be costly, as you need low voltage, let say 16V, and
it even better if 5-10 smaller one connected in parallel.

This concept of using a hugh capacitor to make up for an underated voltage source is bound to be problematic in practice. How often are you going to be turning the solenoid on and how long will the solenoid need to be on when you do activate it?


It might be more common than you think:


Thanks for all for answers, I’ll check out the links provided. I will be cycling the solenoid every 5 seconds, but just for a brief instant, < 50ms.


I'm with lefty on this one, with these sorts of currents you are just asking for it. That link was for a small current model railway.

Remember that the PSU will probably go into current shut down trying to charge up a capacitor of that size. Also the ripple current rating on the cap is going to be vastly exceeded.

The ripple current rating is mainly about overheating, not instantaneous overload of the plates - I doubt if manufacturers quote ripple currents for low duty cycle conditions? What's needed is electrolytics of the same sort of grade as used for flashguns, but with a lower voltage rating. It may prove difficult to get enough current without going for high ripple-current ratings and using a parallel array to keep the ESR really low. Some of the audio-rated electrolyics for car sound systems may be up to the task, if rather expensive!

Alternatively Ultracapacitors may be suitable - again probably expensive.

Redtank, take another think about this...

You only need 35 to 40 amps for 50Ms. Your power supply only supplies 3.8 amps. But the duty cycle is only .05/5 or .01 (1%). So the AVERAGE current is less than .5 Amps.

So you need some storage. Unless this whole thing has to be real portable, why not just use a small 12V GelCel battery, maybe 25 aH or so. It should not have any problem with 45 amps for short duration. Your 12V supply can charge it, and not overcharge it. No reason this has to be a capacitor.

So you have a fairly 'stiff' (well-regulated) 12 volt supply good for short term currents of probably 50 to 100 amps.

Just switch the solenoid circuit on for the 50 mS and then off, through some power FETs. Make sure you have a good spike suppressing diode across that solenoid inductor, probably one rated for use in switch-mode power supplies or something.

It would be "good" if you could characterize the time it takes for current to rise in the solenoid (It's an inductor so the current will start off low and then rise and then level off at the resistance-defined value after a while.) Defining the "While" would be good. Do you have an oscilloscope??

Lefty, Grumpy ?? Sanity check??

Lefty, Grumpy ?? Sanity check??

I agree that a battery would have a better chance of success then a ‘hang a hugh cap on it’ solution. At that coil current level and induction I’m not sure you could afford enough caps with the needed ESR to pull in such a hefty solenoid.


But you could have fun trying I think!

Lefty, Grumpy ?? Sanity check??

I was once sat a very short distance away from a coil being fed DC current with a large capacitor across it. The capacitor exploded, the whole room was filled with silver paper fluttering down like snow. It was only after this did the engineer in charge of it think of ripple current.

When current flows a mechanical force is generated in the conductors. In normal low currents this is insignificant and well withing the limits of a device be it battery or capacitor. However as the current increases so does the force.

Without the correct and sufficient calculations this could be a dangerous thing just to try.

However, the gung-ho attitude is very prevalent in some areas of western culture, and if you survive it is very funny.


Lefty, Grumpy ?? Sanity check??

... However, the gung-ho attitude is very prevalent in some areas of western culture, and if you survive it is very funny.

Grumpy, I assume you mean areas starting some 1000 + miles West of Manchester, UK :roll_eyes:

I'll concede that the number of capacitors intentionally exploded in school laboratories in the USA probably exceeds that of all other countries combined.

However as we mature we graduate (IF we graduate) to resistor values that simply smoke and burn in 2 seconds of less, with a satisfying "pop-whoosh". (Values supplied upon adult request). This is a cultural tradition and I am happy to say that my 40+ year old son has kept it up in teaching "Electronics 101" to employees of his company in Vermont.

Well, It is unlikely to see fireworks on a circuit that supply just 50 Amps, probably even less. I know, that Capacitive Discharge Spot Welder delivers >2000 Amps in the same time frame.

http://www.shearconnectorwelding.com/capacitor-welding-machines.html http://www.philpem.me.uk/elec/welder/ http://ultrakeet.com.au/index.php?id=article&name=cdWelder

And there is a link to capacitor selection tab:


Plus video: http://frikkieg.blogspot.com/2009/06/homemade-capacitor-spot-welder.html

I have reconsidered the capacitor vs. battery solution. I don’t want a capacitor exploding inside my control box. I do have a few SLA batteries laying around I can use…20ah capacity. I would rather use a 12v trickle charger to maintain the battery and keep this part of the project separate from other power needs (stepper motor and Arduino). Terry, I don’t have an Oscope, but will take your advice and use a suitable diode across the solenoid terminals. As for switching…do you prefer the FET board over using a SSR good for 40 or 60 amps? As you mentioned, pulling 40 amps in 50ms isn’t going to happen.


Well, It is unlikely to see fireworks on a circuit that supply just 50 Amps

Not sure what planet you are on but 50 Amps is a lot of current on this one. You can get quite spectaular with that amount of current.

I know, that Capacitive Discharge Spot Welder delivers >2000 Amps in the same time frame.

And did you see the size of those things, they weren't cobbled together by an amateur.

redtank: I As for switching...do you prefer the FET board over using a SSR good for 40 or 60 amps? As you mentioned, pulling 40 amps in 50ms isn't going to happen.


If you can get a DC type SSR like that it sounds like a good solution..

Let us know what works out so we learn too 8)

Well I ordered a Chinese build 40 amp DC to DC SSR, but I'm going to order that FET board as well...I'll try out both and report back on how things work out.

Thanks, Cris