protection for static relay

[basileul]

Hello,
Please help me to solve the next problem.
I want to heat a bread oven controlled by a static relay, a triac. When one of power resistances make short circuit. current growth and triac is destroyed.
Even if a protection circuit stops the command to static relay, it still remains open until the first pass of tension through 0.
If the current exceeds in 10 milliseconds more the surge current ,I MM, static relay is destroyed.
So I believe that the introduction of inductances in series as the best solution to retard the growth of current for 10 milliseconds.
I do not know to calculate the inductance section and number of turns for conductor, if using E + I transformer
Solid state relays has IMM 1200A surge current 10 ms
Working current is 50A

 

[(*steve*)]

I recommend you do something to prevent the elements from shorting out.

 

[TedA]

I have been down this road, but it's been a few years.

An inductor large enough to protect your triac may reduce the load voltage too much at the normal current.

There may be problems getting the triac to turn on and turn off if the overall load is too inductive. Some sort of snubber may be required to make the triac behave.

You can add a special very fast semiconductor fuse, but it may cost more than your triac, and the triac may still blow before the fuse does.

A pair of really big SCRs connected in reverse parallel will be more robust than most triacs. More trouble to gate than a triac, but it can be done.

SCRs are made in larger sizes than triacs, so you can achieve more margin.

It's desirable to keep the power device as cool as possible in normal operation; this will allow it to withstand a larger overload.

Some combination of big SCRs, series inductance, snubbers, and fuses should be able to do the job.

It's difficult to predict what combination will actually work, so you might destroy several sets of parts finding out.

Ted

 

[basileul]

I agree with what you say.
Static relay is made with two SCRs connected in reverse parallel and they have
1200A surge current for 10 ms.
And yes, it is difficult to predict what inductance must be used, for that I posted this question.