TVS vs Zener - Unable to Decide!

[ACharnley]

Hi,

Need some fairly urgent advice!

I have a circuit protected by a 1W 36V zener. Technically the voltage shouldn't reach that as a comparator+FET switch power off just before it, but spikes could occur. Max current is only 500mA (thereabouts).

The first component that would blow if the zener weren't there would be a 40V schkotty. It doesn't leave much room but the 40v schkotty is chosen for performance reasons.

The zener is potentially being replaced with a SMAJ33A TVS (36V breakdown). The potential issue is the TVS has an avalanche decay (faster though), and I'm unable to determine if it would drop all the current by 39V. What I need is a graph which says what current is being dropped at what voltage but datasheets don't have it, only a low res decay graph.

https://www.digikey.co.uk/product-detail/en/littelfuse-inc/SMAJ33A/SMAJ33ALFCT-ND/762508

Would anyone know roughly what a non-stacked (uni) TVS typically drops current wise a few volts over it's breakdown, or would you play it safe and keep with the zener?

Regards,

Andrew

 

[tvtech]

Check 1N5333B chart from Omnisemi. All the figures are there. I can't upload it from mobile

 

[tvtech]

Edit: ON Semi not Omnisemi

 

[tvtech]

 

[kubeek]

https://www.digikey.co.uk/product-detail/en/littelfuse-inc/SMAJ33A/SMAJ33ALFCT-ND/762508

The datasheet tells me that at 7.5A clamping current the voltage on the transil is 53.3V. What happens between the breakdown voltage and clamping voltage is hard to tell, but the current generally should rise exponentially. So I guesstimate that if the voltage jumps from roughly 38.5 at 1mA to 53.3V at 7.5A, that is very roughly four orders of magnitude in current across span of 14.8V.
So that gives me about 3.7V increase in voltage for every ten-fold increase in current, i.e. 45.9V at 1A.

edit: its actually 3.875 orders of magnitude, but 4 seems close enough for the estimate

 

[ACharnley]

OK so based on that the 36V TVS wouldn't work. The 33V would but would be quite close and means a bit more loss in the 33-36v region. To me it seems the zener is the better choice.

 

[tvtech]

Zener has tighter specs.
Just today I saved another set. 5.6V zener which protects the Microprocessor on and old Panasonic went dead short. Some clown had tried to fix it before me. Where the set required a 56V zener...he replaced it with a R2KY....a zener rated at 145V.

A protection zener does its job when it goes dead short.

It kills anything going past it. The ON Semi 5W series does a brilliant job with that.

 

[tvtech]

This above is probably confusing.

Lemme explain.

The set uses a 56V zener to (apart from other things) to supply the tuner with its 33V for tuning the set. The 56 Volt zener keeps things stable. Nothing heats up...or blows up if it's working.

So the clown before me replaced it with an R2KY.... rated at 145V...

All control lost. Popped caps, 7805 Has hole a hole in. So lots of spares used but the little 5.6V zener protecting the micro saved the day. It got cross and went dead short.

)

 

[ACharnley]

tvtech, it's not really answering the question though. Both zener and TVS are shunts, it's the way in which they shunt - the zener has a tighter window but is much slower. Kubeek did a good job explaining that given the parameters the TVS wouldn't shunt all the 500mA (while the zener will).