I am prototyping a battery-operated toy containing a couple of motors - these motors are driven by a pair of h-bridges which are controlled by a micro controller.

I'm running this whole thing from a lithium ion battery, and I happen also to need 5v and 3.3v logic rails.

So I have a boost circuit giving me the 5v from the lithium ion (it's for some of those dinky little ultrasonic sensors), and I'm linearly regulating that down to 3.3v for the micro controller.

So far, so good.

Of course, when the motor runs and/or changes direction, we see voltage spikes on the battery of anything up to about 8v. I have lowpass filters on the logic supply rails, and they survive (for now...), but will that voltage do any damage to the lithium ion battery, and if so how do I mitigate that?

Also - when the voltage spikes happen on the battery, and thus the regulated supplies, I see a voltage drop after the lowpass filters. Is this something to do with phase? It just seems so counter-intuitive.

  • \$\begingroup\$ I have a doubt as to the 8V spike - these sort of things are notoriously difficult to measure with an o-scope. \$\endgroup\$
    – Andy aka
    Mar 2 '15 at 8:41
  • \$\begingroup\$ Good point - let's settle on at least an 8v spike. \$\endgroup\$ Mar 2 '15 at 20:01

Assuming you are using transistor H-bridges, place opposing diodes across the CE points of your H-Bridge drivers to reduce the voltage spikes from the motors. See example at: http://www.robotroom.com/BipolarHBridge.html

It is unlikely that large spikes are reaching all the way back to the Li battery terminals, but spikes could cause uncontrolled behavior of rest of your circuit. A voltage spike into the reference pin of a voltage regulator could cause a temporary unregulated condition. If the spikes are being averaged into the regulator control circuit they might cause an offset to the output voltage.

A large bypass capacitor from the top of the H-bridge (+V) to the bottom (GND) will also help reduce voltage spikes.

  • \$\begingroup\$ I'm using a fet h-bridge from TI - a DRV8837 or something. It's got built-in flyback diodes, would external ones be advisable too? \$\endgroup\$ Mar 2 '15 at 20:00
  • \$\begingroup\$ If the internal diodes are actually for flyback protection then there should not be large measurable spikes. (Any chance you were actually measuring 0.8v spikes rather the 8v spikes?) You could try the adding extra diodes (low Vf type) just in case the internal circuit is defective. Are you using the two bypass capacitors as require by the data sheet? \$\endgroup\$
    – Nedd
    Mar 2 '15 at 23:07
  • \$\begingroup\$ " Are you using the two bypass capacitors as require by the data sheet?" - not exactly, I neglected to include them in my layout and I'm now trying to figure out how to add them as rework :). I think I'd better check back once I've sorted this and some other issues out. \$\endgroup\$ Mar 2 '15 at 23:11
  • \$\begingroup\$ See the Layout Guidelines and Layout Example near the end of the data sheet. \$\endgroup\$
    – Nedd
    Mar 2 '15 at 23:17

A capacitor across the battery terminals should do more than sorting out those spikes, I guess.

If your regulator response is slow/has large lag, it could cause a momentary dip to the peripherals being supplied by the regulator. But this should all be mitigated if you have a low pass preceding the circuit.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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