TVS Suppression Diode Q’s

[SteveG]

So I’m planning to use a TVS diode to protect the motogadget when it switches the main ignition feeds and I have a few Q’s about specs. The littlefuse SLD series appears to be a good one to use.

For the reverse stand off voltage is 15V ok (just higher than normal level) or do I go a few volts higher to give room? I assume you don’t want to go too high to keep the clamp voltage down. It’s 24v for a 15V VR diode.

Also, for this type of application of protecting from switching loads, is it better to go bi-directional or un-directional for the diode? Both types are available.

Any help appreciated.

Cheers

Steve

 

 

[TSD]

It really depends on what you're trying to protect, but in general I'd choose unidirectional if the protected device cannot survive reverse polarity. Normally I would select the TVS based on the maximum safe clamp voltage - if the protected device is safe to (say) 30V, then choose the TVS guaranteed to clamp just below that. It's always tempting to set the protection just above expected voltage, but if the protected device is safe to higher voltage, then it doesn't add protection, but does increase the chance of the protection failing.

SLD seems a decent choice, I wouldn't use anything smaller. Load dump is high energy and high voltage , so TVS should be on short fat traces to the local power and ground. Typical failure mode for TVS is short circuit, so suitably sized upstream fuse is a must, and install the TVS where you can inspect it / remove it if necessary. If it fails s/c and the fuse doesn't blow, the TVS will probably eventually burn out, leaving no protection in place - visual inspection is a big plus in this case. If it doesn't burn out, and the device it's protecting is 'mission critical', you'll be glad to be able to cut the device before you run out of spare fuses.

Personally I try not to use TVS if I can help it - but there's a different set of assumptions if you design a black box for non-technical end users and have no control of the installation.

Edited February 26, 2018 by TSD
clarification

[SteveG]

Thanks for the reply TSD, the motogadget has an over voltage protection between 16-40V on it’s supply input, but I don’t know if it has the same on the outputs. I’ll drop them an email and see what max clamp voltage it can cope with. Each output is essentially a MOSFET, so it’s designed to cope with a short circuit condition, and will shut down the output.

cheers

Steve

[FridgeFreezer]

I'd really hope it's internally protected on all I/O otherwise it'd be a very fragile unit.

[TSD]

Didn't realise you were intending using them on the outputs - I don't think they make much sense there, though I'm not familiar with those units other than a quick squint at the website. Not sure if the outputs are high or low side switched?

Inductive kickback will initially be negative going and dumped by the body diode of the mosfet. If there is likely to be any positive going pulse present, then a better protection might be a big diode from the output to the power supply, so the energy is dumped into the 12V rail.

This should be more robust, as the diode is going to be dumping all the current at <1V drop, rather than 18V across the TVS, so only 5% power dissipated in the protection. (its

You can then use a single TVS across the main power supply, and you probably should if it's only safe to 40V, the standard load dump test pulse is >80V for a 12V system.

Many ways to skin a cat...

[SteveG]

I don’t think Motogadget call out needing to protect when switching inductive loads, however, it seems universal with these MOSFET based PDM’s, that the best approach is to fit a suppression diode across any inductive load.

The plan is to have the motogadget switch via an Albright isolator the ignition switched feed for the Bussmann RFRM’s (Relay and fuse power distribution unit) two 100A capacity busses. The thought was to use a TVS diode on the output to protect against any potential inductive kickback.