SCR gate triggering

bmandl said:

I used a DIAC based optocoupler

Click to expand...

are you sure? What part number is it?

bmandl said:

or should I use VDD and transistor based optocoupler for gate firing? In latter case, wouldn't microcontroller be exposed if SCR fails and conducts between anode or cathode and gate?

Click to expand...

1) if the SCR fails ... then there is something wrong with the design. In my designs I don´t consider parts to fail. SCRs are rather robust .. and if they fail, there surely is a reason.
2) you say you use an optocoupler ... this is an isolating part and thus you should not get a destroying feedback to the microcontroller

***
First big picture:
I don´t recommend the setup of Q2 as emitter follower. Emitter followeR is a circuit mainly used for linear circuits .. but here you clearly want a digital operation.
This does not mean that it does not work ... it just is rather unusual. Maybe you did this intentionally .. for a reason I don´t see.
(the same is true for all other pictures)

***
May I ask why you switched from TRIAC to SCR ... it makes things more difficult.

***

bmandl said:

If so, what additional protection should I use for microcontroller in that case?

Click to expand...

As already said: I design them to go not defective. And why should the SCR go defective, why not the microcontroller, the power suppy ... or anything else.

******
I´ve designed TRIAC control and SCR control. SCRs are used especially because of high power (250A/400V as well as 3000A RMS continously on a 1MW application 24/7)
There are 10 to 16 SCRs independently controlled. If the SCR get killed then there is a major problem else where. The microcontoller is the smaller problem.
In case of a bigger explosion you should consdier to replace the whole electronics including power supply, maybe even wiring.

Klaus

Thank you @KlausST for your answer.

KlausST said:

are you sure? What part number is it?

Click to expand...

Yes, I'm pretty sure. The part is MOC3021.

KlausST said:

1) if the SCR fails ... then there is something wrong with the design. In my designs I don´t consider parts to fail. SCRs are rather robust .. and if they fail, there surely is a reason.
2) you say you use an optocoupler ... this is an isolating part and thus you should not get a destroying feedback to the microcontroller

Click to expand...

I'm sorry, in the meantime I figured, that I indeed need an AC type of driving triac, better yet pulsing type - for reducing gate power dissipation. Not fully familiar with that concepts yet, but I'll investigate further.

KlausST said:

***
First big picture:
I don´t recommend the setup of Q2 as emitter follower. Emitter followeR is a circuit mainly used for linear circuits .. but here you clearly want a digital operation.
This does not mean that it does not work ... it just is rather unusual. Maybe you did this intentionally .. for a reason I don´t see.
(the same is true for all other pictures)

Click to expand...

Can you please explain why this circuit is not ideal in your opinion? Is the transistor I chose not ideal, or connection itself? What should I change? Oh, yes and it works in my other project, but I agree with you, if it works it does not mean it is the correct way to do it.

KlausST said:

***
May I ask why you switched from TRIAC to SCR ... it makes things more difficult.

***

Click to expand...

Somehow I got convinced by researching, that SCR is more appropriate for application like this, than TRIAC. By my understanding, TRIACs are more appropriate for applications like phase angle control for dimmers, heaters, etc, but for bigger loads like motor (in my case 5.5kW 3 phase motor), SCR is better option, especially because heat dissipation is divided among 2 devices instead of just on TRIAC. Maybe I'm wrong and I'd also be happier if that's true


But anyway, if I still insist by using SCRs after you explain to me, why this is unreasonable (just being hypothetical here), Is the circuit I draw (latter one) correct (ignoring emmiter follower part for now)? I should probably add some diodes for gate protection and perhaps some snubber circuit right?

bmandl said:

Yes, I'm pretty sure. The part is MOC3021.

Click to expand...

I can´t find any evidence for "DIAC"

bmandl said:

AC type of driving triac, better yet pulsing type - for reducing gate power dissipation

Click to expand...

Don´t know what this means. Please explain ... maybe draw a sketch...

bmandl said:

Can you please explain why this circuit is not ideal in your opinion? Is the transistor I chose not ideal, or connection itself? What should I change? Oh, yes and it works in my other project, but I agree with you, if it works it does not mean it is the correct way to do it.

Click to expand...

I clearly focussed on "emitter follower setup". If you are not familiar ... you are free to do an internet search on it.
The MOCxxxx datasheet shows the transistor connection. DS-70-99-0019, page #8 ... don´t forget the base resistor.

Your schematic neither says the voltage of VDD, nor the signal levels.
So I guess both are 3.3V (supplied)
If you get a base current of about 120uA ... Fairchild datasheet uses a ratio of 20 for switching applications...
So 120uA x 20 = 2.4mA only... while the datasheet talks about 8mA typ.
I would not be surprised if it causes some unexpected behaviour.
Did you CALCULATE your drive circuit? ... or where is it from?

bmandl said:

but for bigger loads like motor (in my case 5.5kW 3 phase motor), SCR is better option, especially because heat dissipation is divided among 2 devices instead of just on TRIAC. Maybe I'm wrong and I'd also be happier if that's true

Click to expand...

Again your information is missing mains voltage.
But for a 230V motor the line current may be about 10A.
Farnell lists about 400 TRIACS for >10A RMS. The package can handle the power dissipation and for the heatsink it does not matter at all.

bmandl said:

But anyway, if I still insist by using SCRs after you explain to me, why this is unreasonable (just being hypothetical here), Is the circuit I draw (latter one) correct (ignoring emmiter follower part for now)? I should probably add some diodes for gate protection and perhaps some snubber circuit right?

Click to expand...

The MOCxxxx datasheet tells you how to drive TRIACs. I would not be surprised if the MOC manufacturer provides documents how to drive SCRs.
Basically I´d add a diode in the gate drive path ... because the MOCxxxx device relies on the fact that the TRIAC switches ON when you drive the gate.
But an SCR only swithces on at one half wave. Now if you activate the MOC at the wrong half wave .. it may cause gate current, but the SCR does not switch ON ... causing overheat on the MOC.

Klaus