Hello there.

I am designing a circuit that will phase control 240VAC at 12A. The snubber diode is a VS-15ETH06-M3. The characteristics of this are as follows:

Forward voltage at If 15A = 2.2 (Max)
Maximum junction and storage temperature range TJ, TStg = -65 - 175 °C
Thermal resistance, junction-to-case Rjc = 1.0 (Typ) 1.3 (Max) °C/W
Thermal resistance, junction-to-ambient per leg Rja = 70 (Max)
Thermal resistance, case-to-heatsink Rcs = 0.5 (Typ)

Unfortunately, I do not have the exact specifications of the heatsink, as I'm purchasing from LCSC (heatsink category here), and the heatsinks don't appear to give any thermal specifications.

So, worst case:

Pd * (Rjc+Rcs+Rsa+Ta) = (2.2 * 12) * (1.3+0.5+20+25) = 1235 C/W

Yes the above is the worse case, it still seems excessive for such thermals. Surely if this is correct such components like this diode are all but useless as they could never thermally achieve their ratings.

I'm clearly missing some piece of the puzzle here, would you mind pointing me in the right direction.

Thank you.

 

I'm clearly missing some piece of the puzzle here, would you mind pointing me in the right direction.

Phase control circuits have a resistor-capacitor snubber and an inductor-capacitor filter. No diode.

 

Hello Ian,

Thanks for the prompt reply. I don't know if it makes a difference, but I driving it via an inverse-parallel SCR circuit as per figure 10 on the optocoupler datasheet (MOC3163M).

However, that aside, I would still appreciate correction/guidance for the thermal calculations, as something does not seem right.

Thanks.

 

Hello Ian,

Thanks for the prompt reply. I don't know if it makes a difference, but I driving it via an inverse-parallel SCR circuit as per figure 10 on the optocoupler datasheet (MOC3163M).

However, that aside, I would still appreciate correction/guidance for the thermal calculations, as something does not seem right.

Thanks.

How do you have it wired?
For triacs and SCRs I have a simple rule-of-thumb: Allow 1W dissipation for each Amp of current.
If you have inverse-parallel SCRs each one dissipates half the power, because it only conducts half the time.

 

It isn't wired in any special way, just as per the basic application of the MOC3163M datasheet:

 

The current through D1 is only the gate current, so Vpeak/360 = about half an amp. It only lasts for the time it takes the SCR to trigger (<1us), so a 1N4004 will be perfectly adequate.
Each SCR will dissipate half a Watt for every amp of output current.
By the way, it's not a snubber.

 

Ah I was under the impression that R1+D1 formed a flyback circuit, likewise for R2+D2.

Many thanks for your help...

 

Ah I was under the impression that R1+D1 formed a flyback circuit, likewise for R2+D2.

Many thanks for your help...

R1 and R2 are pulldown resistors that (hypothetically) stop stray current triggering the thyristors. If you have a standard gate thyristor (say one that needs 50mA to trigger it) then it's probably got resistors like R1/R2 built in, so you can leave them off your circuit.
D1/D2 provides a current path for the trigger current to the other thyristor. The trigger current for the RIGHT thyristor goes through D1, because it can't go through the gate-cathode junction of the LEFT thyrsitor, and vice versa.