static switch

--
Use a contactor.

A three-phase motor will draw 5 to 10 times the nameplate current upon starting or with a locked rotor or a missing phase.

1. Choose a triac which can handle that much current.
2. Choose a breaker which will trip with that much current.*

*Motor may be equipped with an internal self-resetting thermal breaker (Klixon, for example).

For more severe faults (phase-to-phase or phase-to-ground), you must protect with sub-cycle fuses whose I^2 * T (eye-squared-t) rating is less than the triac's I^2 * T rating.

Good luck.

John

What makes you think you have the skill level to do this?

Why did you put "static switch" in the subject line? What does this imply?

--

    Boris Mohar

Yup; the trouble is that it seems nobody has been specifying I^2*T on power semis for about 20 years. Even if ignitrons were still being made, i bet that spec would be dropped for them also. And the engineers that were around then are retired or have departed... The newbies do not have a clue when asked...

modern machines are high efficiency, and end up about 10x to 20x (10% to 5%)

a LOT less.

work it out from Ifsm.....

its spec'd for real power semis, just not toys ;)

Cheers Terry

It has been a few years since I worked in motor controls. I accede to your more recent knowledge.

In any case, the OP should be able to get the locked-rotor rating from the manufacturer.

That's a good one!

I quickly looked up a triac on Mouser, the T1620W made by ST. The I2t is given. I know this does not mean that the I2t is given for all (toy) triacs and thyristors, but it seems likely that the info is available for most. I designed motor control power stages from 1980 through 2000 and I never ran across a thyristor data sheet which did not have I2t given. However, my work was with Westinghouse, Westcode, and, later, Powerex thyristors. We never used triacs.

OP: what triacs are you using? What are the motor nameplate ratings?

John

There are two ways that I have seen this problem addressed, when I worked as an engineer at a company that made static switches. Both of these answers have been mentioned in the replies to this thread: use fuses, or oversize the semiconductors to handle whatever you are going to throw at them.

You mentioned triacs in your post, but you may also want to consider SCRs and IGBTs.

I'm doing two projects with toy thyristors at the moment.

TIC106 - no. and a ratshit datasheet

BTA/B10 - yes.

BTW67/69 - yes.

I think the conclusion we can draw here is that lousy thyristor datasheets dont include IIt.

However its a different story when we look at FETs, IGBTs & diodes. I havent seen a toy diode, FET or IGBT spec IIt yet, but its pretty common with the big stuff. interestingly enough, none of the Semikron miniSKiiP datasheets have IIt.

but everyone has Ifsm, and there is a relationship between the two; an IIt value based on Ifsm is lower than the actual fusing IIt, so it gives a reasonable safety margin.

If only manufacturers would tell us the die dimensions!

Cheers Terry

Google is your friend....

:-)

DNA