Skip to main content

Reply to "Weak DCS signals, Failed TIU Output Drivers, and Design Solutions devloped under collaboration between AGHR and MTH"

Adrian! posted:

 new_TSV_locations

 

 

Adrian! posted:

Hi there, we have 5 TIUs with the TVS and beefy clamp in parallel. In 3 years we haven't had a single channel degrade or die except, one TIU that burnt down when the building contactor failed and put 400V RMS to the single-phase outlet. It does seem pretty rugged.

The TVS turns on very fast (like 100ns-1us) so it protects you from very fast transient events. The only issue is if you have continuous voltage overshoot it will eventually go into non-reversible breakdown. The 1n4148 clamp is very slow to turn on (like 5us range) and won't protect from the like (1-1000ns) events, but can handle over-voltage pretty much indefinitely. So if a continuous overshoot occurs the TVS clamps quickly and then hands the shunting current over to the 1n4148 after a few us so it doesn't break. That's why the 1n4148 stack needs a slightly lower threshold than the TVS... so in the steady state all the current in in the 1n4148, not the TVS.

 

That sounds like as good of a review as anyone could ask for. 

That makes perfect sense. So to wrap things up, in your solution you removed the 4 small SMD TVS diodes added the SA5.0CA in the positions drawn in yellow above for the very fast transients and used the "Beefy Clamp" 1n4148 array board hooked up in parallel to take the load off after 5us thus saving the SA5.0CA indefinitely (or 3 years so far) then. Correct?

 

and sorry if I missed it but did you mention the science behind why adding a 22uH choke in series with the transformer output would help the DCS signal output of a TIU in passive mode?

Last edited by Ryaninspiron

OGR Publishing, Inc., 1310 Eastside Centre Ct, Suite 6, Mountain Home, AR 72653
800-980-OGRR (6477)
www.ogaugerr.com

×
×
×
×
×