# Temperature of Solid State Relay

I would like to use a solid state relay (25A) in a 230V max. 1400W environment which will switch in 10 second intervals for a few minutes and then remain idle for hours.

Will this type of switching produce any noticeable thermal output or even require a heat sink? I am intending to place the relay close to the inside wall within a warm box (coffee machine, ~80C) with limited air circulation.

Can SSRs be used for such use cases, or are additional measures required to prevent overheating?

• Provide a link to the datasheet! You operate the SSR at 25% of its maximum at a much elevated ambient temperature, so my gut feeling is that you will need a good heatsink. – Wouter van Ooijen Nov 20 '15 at 15:47
• Which SSR are you interested in? – Marko Buršič Nov 20 '15 at 15:55
• I still havn't got a SSR. I was hoping to get a hint which SSR might be a good choise for this use case. The 25A rating was just an example. It could also be any other kind of relay which can provide galvanic isolation – Franz Kafka Nov 20 '15 at 15:59
• At such low frequency switching also consider a mechanical relay, you can get them cheaper with lower on resistance. – Matt Nov 20 '15 at 16:04
• Some will, some won't (almost certainly). Check a few data sheets and if something is confusing then this is the place to ask. – Andy aka Nov 20 '15 at 18:14

Let's do some maths:

Consider this 40 A rated triac - BTA40.

Vto = Threshold voltage = 0.85V = Voltage drop across triac.

Rd = Dynamic resistance = 10 mohm

IT(RMS) = Current through the triac = 25 A

Power dissipation = [0.9Vto X IT(RMS)] + [Rd X {IT(RMS)} X {IT(RMS)}]

Where 0.9 = 2*sqrt(2) / pi

Putting the values, you get P = 25.375 Watts

This much power can't be dissipated by a small triac. You will definitely need a beefy heat sink (maybe a cooling fan too).

In my humble opinion, if you plan to use a triac without a heat sink and in closed spaces, you should limit yourself to 1A or so (for a continuous draw, it gives P = 0.775 watts). For the application you have mentioned, I'd suggest you to go for an electro-mechanical relay which will be cheaper and smaller as compared to triac + heat sink assembly.

I actually tried something similar to control a pool pump as part of a solar water heating system. In my case the load was only 15 amps at 120V, and I chose an OMRON G3NE-220T-US, which is an SSR rated for 240VAX at 20 amps. (control voltage=12VDC). In my case the switching was not quite as regular and predictable, but was still frequent enough to raise these same questions for me. especially since the motor would surly draw more on start-up when its mechanical load was higher.

Anyway, I was somewhat dissapointed to find that the heating was a factor, but not so much from the switching as from the normal "running" load condition. Certainly the bare flat metal back of the SSR along with its screw holes on either side was an immediate clue to me that heating (and a heatsync) was to be expected, and it certainly was more heating then I originally expected. Apparently, these SSRs respond pretty much much like a TRIAC , and if you measure the voltage across it under load, you should expect to see an average somewhere between 1 and 2 VAC. Assuming my fluke was giving me a relatively close to RMS value, it became obvious that at 15 amps I was still dissipating between 20 and 30 watts here.

Now how much actual temperature rise you can expect depends of course on many factors. In my case though, the extra 20 watts of heat was a show stopper. Regardless of whether I used a heat sync, the entire control circuit was going to go into a metal box that already had to operate in the heat of a Florida summer. So there was the factor of the heat affecting the reliability of the rest of the circuit, and there was the user product perception factor of a device that would ultimately get pretty warm (if not outright hot) tot he touch.

So just FYI, I finally decided to abandon the SSR in favor of simple OMRON mechanical relay with four 10 amp contacts, (OMRON LY4-12VDC), with all contacts paralelled together and the usual capacitor across the contacts to block arcing from the inductive load. Feels like a step backward in a way, but the mechanical relay has now operated for 8 summer seasons, never gets even mildly hot, and seems a better solution for that reason.