I just ordered an Arduino Mega and will be using it for temperature control. I will be heating a small volume of air to setpoint using NiCr80 heating wire.
I hope to control this using the PID library and having the Arduino output to one of the DO PWM ports.
The load, heater wire, can be powered using either DC or AC, I have chosen DC.
I am a little confused on SSR and PWM. Some threads have identified this can be problematic... but I think only for AC loads.
The turn off time of that SSR is almost as long as half a cycle at the PWM frequency, so you won't get very good control at low PWM values. However, as the application is a heater, you could reduce the PWM frequency to 50Hz or even 10Hz, that way the long turn off time will not be a problem.
Hopefully, it's just the opto isolator in the SSR that is turning off slowly, not the mosfet.
Hrm.... with a 50% duty cycle would the signal cut off in the middle of a PWM wave, or would it be on for the first wave and off for the next wave?
If the first scenario is true, then I agree, the SSR wouldn't be adequate for low PWM values at the stock PWM frequency.
However, if the second scenario is true, the SSR would be OK. Assuming a 10% duty cycle the signal would be on for the first wave, and off for the next nine.
SSR rated for AC service typically utilizes thyristor devices (SCRs or a Triac) to perform the electrical switching. Such devices can only start conduction randomly, but cannot stop conduction as that is done only when AC voltage source passes a 'zero crossing' every 8.333 millisecond. Also most common SSR only turn on (if commanded to) at zero crossings rather then randomly. So when you analyze that timing restrictions you should understand why most PWM control using SSRs are at a very slow switching speeds often using a 10 second PWM window, which is fine for most temperature control loops.