I have 32 servos running from a 7.5V battery rated at 2500mAh and I'm sure that it has enough power to handle the servos, but it doesn't look like it can handle all of the servos pushing at once.

I was thinking I could add a capacitor to the circuit to solve this issue, so bearing in mind that I'm very new with electronics and have only just found out that amperes are a direct result to the requirement of the circuit rather than what the battery puts out, please could you answer my questions below?

  1. Could I put a capacitor just before the servo controller to deal with the intermittent load requirements?
  2. Would a large capacitor damage the circuit?
  3. Does a capacitor work just like a battery and only output the required power rather than just discharging everything it has into the circuit?
  4. Does a capacitor have to be mounted across the positive and ground or does the positive have to go through it?
  • 1
    \$\begingroup\$ A capacitor holds very little energy compared to a battery, in general. It will smooth out short spikes in current, but will not help with longer bumps or continuous demand. \$\endgroup\$ Apr 7, 2014 at 14:42
  • \$\begingroup\$ That's a lot of servo's. Does the datasheet say how much current they draw? \$\endgroup\$
    – RJR
    Apr 7, 2014 at 23:33
  • \$\begingroup\$ @RJR yeah the robot has a lot of degrees of freedom, especially for a biped. I'm sure it does but the servos I've used are cheap due to how many I need, and so they look unnamed. \$\endgroup\$ Apr 8, 2014 at 8:16

2 Answers 2

  1. Yes
  2. It shouldn't. It should smooth out voltage droops.
  3. Yes, it acts more like a constant voltage source. An infinite sized capacitor would be an ideal voltage source.
  4. It should be mounted between positive and ground. Basically, it should look like it's a parallel mounted battery.

You can think of a capacitor kind of like the tank on most toilets. It stores extra water for the times that you need extra water (current) to flush. The capacitor in this analogy would be like a pressurized toilet tank. With pressure being voltage and the volume of water being current. When your servo turns on, you now have more temporary volume of water at line pressure to satisfy your needs if you install a capacitor. In short, more current at the same voltage.

  • \$\begingroup\$ Thank you for your quick answer, though it has sparked a new but very related question. Does a capacitor match the voltage of the battery or do I have to get a capacitor rated for the voltage of the battery? \$\endgroup\$ Apr 7, 2014 at 14:44
  • 1
    \$\begingroup\$ If you put the cap in parallel with the battery, then yes, it will match the voltage of the battery. Keep in mind when choosing a capacitor that you will want it rated about 25% higher than the highest voltage you expect it to see (engineering rule of thumb). If the capacitor is placed at some other point in the circuit, say after a regulator, it will match the voltage of the regulator. You still need to have the cap rated for at least 25% above this voltage though. \$\endgroup\$
    – horta
    Apr 7, 2014 at 14:49
  • \$\begingroup\$ Thank you for helping out with this, I like the toilet analogy. Very easy to understand, thank you. \$\endgroup\$ Apr 7, 2014 at 14:59
  • 1
    \$\begingroup\$ @horta Out of interest, does the 25% rule vary dependant on the kind of cap? A lecturer I had seemed to hint that electrolitics were far more sensitive to voltage than other sorts and some other kinds cared far less about voltage. \$\endgroup\$
    – Vality
    Apr 7, 2014 at 19:45
  • \$\begingroup\$ At least 25% over-sizing is just general engineering guideline for any component. Electrolytics have definitely have a bad rap due to capacitor plague: en.wikipedia.org/wiki/Capacitor_plague Since electrolytic capacitor plague was an aging issue, over-sizing them even more would have mitigated the ageing issues seen. In short though, if you trust a manufacturers spec, it shouldn't matter which kind of cap you use. \$\endgroup\$
    – horta
    Apr 7, 2014 at 19:58

In this situation, you can think of a capacitor as a fast charge and discharge battery in parallel to your normal battery.

1 - maybe (see below)

2 - A large capacitor will demand a large amount of current to charge and be capable of sourcing a large amount of current for a short time. This could damage the circuit (i.e. battery), but it depends on the circuit. It would most likely take much larger capacitance than you are going to use to cause issues, with one exception. When connecting a battery to a discharged capacitor, the current drawn can be extremely high as the cap charged up.

3 - In DC realm, both a battery and a capacitor will discharge what is required to satisfy Ohms law. This means that as your load power goes up (i.e. load resistance goes down), more current will be supplied until the voltage drops.

4 - You would mount it in parallel with the battery.

But before you jump to capacitors, you need to look more closely at your setup.

A battery has a certain voltage and a certain internal resistance. Say you have a cell with 4V and 0.1 Ohm internal resistance. If you try to get 5A out, you get 5A * 0.1 Ohm = 0.5V lost with the internal resistance of the battery. So your load only sees 4.5V. You also may have to factor in wiring resistance from your battery to the load as further voltage drop. At a certain point, the voltage left at the load is insufficient. An extreme example of this is a car starting, where the starter may only see 8V from the 12V battery. A capacitor can help by demanding less instantaneous current from the battery, and thus keeping the bus voltage higher.

It is possible that a capacitor or capacitor bank is enough to get you through the peak demands. This assumes that your motion is not continuous on most servos and the holding power required is low. A much more ideal solution is to size the battery such that the discharge rate is close to what is needed. This depends on battery chemistry and is expressed in a multiple of C (battery capacity). For example, if your battery could discharge at 3C, then you could pull 3 x 2500 mAh from it or 7.5A.

If you use a large capacitor bank to feed peeks, unless you have enough resistance between the battery and capacitor banks, you can still damage the battery. The battery will push as much current as possible into the capacitor bank to recharge it. Batteries are simple things and it is the responsibility of the designer to structure the load in a way that the battery is not damaged.

If you choose to use capacitor bank, you also have to realize that there may be situations when the power available is exhausted and you are pulling more than you should from the battery. This is going to happen VERY fast unless you have an extremely large capacitor array.

  • \$\begingroup\$ Thank you for taking the time to explain, you've definitely given me something to think about in terms of battery size etc. The only reason I didn't mark yours as the answer is because @horta got there first. \$\endgroup\$ Apr 7, 2014 at 15:01
  • \$\begingroup\$ @LukeAlderton: you should accept the answer that is best, not the one that is first. \$\endgroup\$ Apr 7, 2014 at 22:06
  • \$\begingroup\$ @insta They both answer my questions and both have different ways of explaining them, they are also both very good and equally satisfying answers. In which case, he who comes first gets accepted first. \$\endgroup\$ Apr 8, 2014 at 8:13

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.