Spark suppression diodes


Well-known member
9 Feb 2002
Aberdeen Scotland
I'm currently rebuild an controller for the Lewmar Bow thruster on my boat, I've got to the bread board stage, knocking together the components to see how well it works, without worrying to much about the final form.
I've tested the 3 Lewmar systems which failed, and in every case it has been one or more of the diodes fitted in parallel with the MOSFET semiconductor switches. I wrote this up a while back when the previous one failed. They are encapsulated in a potting compound, and you can't replace them.
I'm now using, good old fashioned relays, with built-in diodes. However, I was wondering about adding extras diodes in parallel. The logic being, loosely, around the likely size of any spike from a 5kW, 12v DC motor (which can pull nearly 500amps) when it is switched on and off. The motor has significant rotational inertia, and inductance. My guess is that the voltage spikes are the root cause of the diodes failing.
Digging around on the web, I've found a device called a TVS, it looks like back to back diodes, specifically designed to supress voltage spikes from rotating machinery.
I've seen the TVS device P6KE33CA recommended for 12 volt systems, they are very cheap, so cost isn't and issue.
I'm wondering about added them across the relays, as an added precaution.
A sensible idea, or not?