Dimming inductive loads using SSR causes SSR to break

[SeBsZ]

Dear All,

I will try to post as much information as possible so that hopefully this problem may be solved without too much extra question-asking.

I have a microcontroller driving an SSR. There is a zero-crossing detection circuit connected to the uC. There used to be no protection across the outputs of the SSR (loads are connected directly to the SSR). I have written the software to correctly fire the SSR at the correct time in relation to the zerocrossings. I want to stress the fact that I am able to dim resistive lighting loads 100% perfectly. I've tried 220V halogens, incandescents, and even the new modern dimmable LED bulbs work perfectly fine.

Problems came in - as expected - when I tried to dim some inductive loads. In this case I'm talking about those 12V halogen light spots which obviously have a transformer transforming the 220V to 12VAC. Various problems would occur, mostly the SSR would either physically break in an always on or always off position.

The SSR in question is a S202S01.

I decided a snubber circuit was necessary so I found out an RC network is all that's needed. I found a capacitor (630V, 0.1uF, Ceramic) and a resistor (120ohm carbon 0.5W) which I connected in series across the SSR ~ terminals. I decided to fade my load on and off. This worked well albeit the snubber made a lot of 50hz noise (until fully on). I then decided to just dim the lamp at about 50% brightness. This worked for about 5 seconds until the lamp jumped to 100% brightness. Yes... the SSR has broken again. Measuring across the SSR terminals shows a resistance of about 2Kohms. Measuring across a working new SSR shows a resistance of about 20Mohms! Replacing the inductive load with a resistive load shows the load always staying on (it can no longer be controlled by the microcontroller). The resistor is still measuring 120ohms and the capacitor is measuring infinite resistance. Yes the SSR has physically shorted together more or less. Also removing the snubber circuit does not change anything. Again the SSR has physically broken.

Now I hope I've described the problem clearly. We're not talking about huge loads, or motors. I want to dim inductive lighting, as well as resistive lighting. Dimming circuit works fine with resistive loads, but SSR seems to "short closed" when driving an inductive load for a few seconds. Tried with snubber circuit, still not working - only more noisy.

My question: is a snubber the right tool for the job? Are the snubber values wrong perhaps? Who can give me some ideas?

Many many thanks for reading this and attempting to help me!

Best regards,

SeBsZ

 

[duke37]

A transformer needs to be driven with AC without a DC component or it can go into saturation and pass a very high current.

I supplied my neigbour with a transformer to heat a wire to cut polystyrene. I put a 100W light bulb in series with the dimmer switch and transformer, this worked at the expense of some power loss.

Without the bulb the transformer would buzz at intermediate settings and I was concerned that the excessive current would kill the lamp dimmer.

I do not know how your problem could be solved.

 

[cjdelphi]

Problems came in - as expected - when I tried to dim some inductive loads. In this case I'm talking about those 12V halogen light spots which obviously have a transformer transforming the 220V to 12VAC. Various problems would occur, mostly the SSR would either physically break in an always on or always off position.

If you're powering a 12vac downlight, would it not be easier to PWM a 12v supply directly to the bulb? an SSR would not even be needed then...

 

[(*steve*)]

The problem with a SSR is that you turn it both on and off.

Ignoring the effect of DC into the load, if you're turning the SSR on at zero volts, that's good, but if you're turning it off again at zero volts, that is very bad (the current will not have fallen to zero and you'll get an inductive spike that is probably destroying the SSR.

What you need to do is turn it on at zero volts (actually, that's only really important for capacitive loads), and not turn it off until the current falls to zero.

If you used a triac, that would do this for you automatically.

 

[SeBsZ]

The problem with a SSR is that you turn it both on and off.

Ignoring the effect of DC into the load, if you're turning the SSR on at zero volts, that's good, but if you're turning it off again at zero volts, that is very bad (the current will not have fallen to zero and you'll get an inductive spike that is probably destroying the SSR.

What you need to do is turn it on at zero volts (actually, that's only really important for capacitive loads), and not turn it off until the current falls to zero.

If you used a triac, that would do this for you automatically.

The SSR I am using, S202S01 is called a "Triac output SSR" so I assume it has a triac at the output. Additionally, the datasheet mentions one of the applications as "Switching motors, fans, heaters, solenoids, and
valves."

The datasheet also writes:

"When making the transition from On to Off state, a snubber circuit should be used ensure that sudden drops
in current are not accompanied by large instantaneous changes in voltage across the Triac.
This fast change in voltage is brought about by the phase difference between current and voltage.
Primarily, this is experienced in driving loads which are inductive such as motors and solenoids.
Following the procedure outlined above should provide sufficient results."

This is why I added a snubber circuit. They write this about the snubber circuit:

"When the input current (IF) is below 0.1mA, the output Triac will be in the open circuit mode. However, if the
voltage across the Triac, VD, increases faster than rated dV/dt, the Triac may turn on. To avoid this situation,
please incorporate a snubber circuit. Due to the many different types of load that can be driven, we can
merely recommend some circuit vales to start with : Cs=0.022μF and Rs=47Ω. The operation of the SSR
and snubber circuit should be tested and if unintentional switching occurs, please adjust the snubber circuit
component values accordingly."

My snubber values are different though. I used 0.1uF and 120ohms. Could this be a problem? Can someone tell me whether my different snubber values are the reason for my problems?