1
\$\begingroup\$

I'm currently designing a variable frequency drive for powering an AC induction motor.

I've three 400V/680µF electrolytic caps for smoothing the 325V (nominal) DC bus.

Then the DC bus is chopped by the inverter and fed into the AC motor (which is a very inductive load, with a lot of dv/dt generated during switching).

My question is putting two TVS diodes at the end of the cap bank is useless or not? Can it provide somehow a "clamp" against these voltages spikes and protect the caps?

enter image description here

TD1 and TD2 are Littelfuse 1.5SMC510CA TVS diodes.

Thanks ;)

\$\endgroup\$
1
  • 1
    \$\begingroup\$ The capacitors should be at least 1.5x the rail voltage or about 500V. The current capacitors that you have won't be protected by 485V TVS diodes. \$\endgroup\$
    – vini_i
    Commented Feb 20, 2018 at 20:43

1 Answer 1

Reset to default
1
\$\begingroup\$

Best to install 40 mm MOV's at the motor windings, rated for 150% of the applied voltage, AC or DC. They will act as snubbers during any back-emf, and are high-impedance during operation when current is ON. During this 'switching' action they will limit voltage spikes due to overshoot and undershoot.

More or less they are protection for your VFD. The capacitors on your main power rail are your protection for that area. Better to use many small ones than one huge capacitor, so ESR is not an issue and they will keep the power rail quiet, free of noise spikes over 1 volt.

For an ultra-quiet power rail you would need to add an inductor of 1 mH, rated at twice the rail current, with capacitors on both sides of the inductor. Noise could be reduced to 100 mV this way.

\$\endgroup\$
2
  • \$\begingroup\$ except MOV's deteriorate with each usage \$\endgroup\$
    – user16222
    Commented Feb 20, 2018 at 21:06
  • \$\begingroup\$ @JonRB. Not as much as you think, especially 40 mm MOV's at 150% rating. In running a UL testing lab at my last job, UL was concerned with durability and de-rating by constant usage as well. 5,000 hits at 10 kA was the basic test they had to pass. Likely the OP's VFD has snubbers to protect the output stage. \$\endgroup\$
    – user105652
    Commented Feb 20, 2018 at 21:13

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

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