PWM control?

[star882]

I'm making a small, quiet portable freezer. I think I'll use 2 Peltiers in series since the heatload is very small and the application requires rather low temperatures (-20C or lower).
Can I PWM the Peltiers directly without any output filters? I can greatly simplify the circuit if that's the case. The two Peltiers (and the fans) are independently PWM controlled. The frequency will be just over 20kHz to avoid noise problems while maximizing efficiency.
Will the high frequency pulses degrade the Peltiers in any way? It's too fast for any significant thermal cycling to take place, but are there any other issues?

[Holst]

I dont think it will be any real problem with damage to the TEC's.

Why dont you just use a thermostat and turn the TEC on and off.

The Delta T of most TEC is 70*C
So at 20*C ambient you can expect -50 if your insulation is perfect.
I think -20 should be easy to acheive with just one peltier, once you start stacking the pelts then things get MUCH more complex.

Please let us know how you get on with this project.
And good luck.

[star882]

Why dont you just use a thermostat and turn the TEC on and off.
That causes thermal cycling which the PWM avoids.

[Holst]

That causes thermal cycling which the PWM avoids.

So what is it that your cooling?

Will it be expensive to build the circuit (considering the high amps?)

[star882]

It's a freezer, so only the insulation losses.
I have plenty of transistors rated at 10A or more. The Peltiers used draw a maximum of 6A each, which will decrease when run at below maximum voltage.

[jdubol]

The TEC builder Marlow, in its on-line documentation (marlow.com), says that regulation by on/off must be avoided, and a PWM above 1 kHz is right.

[littleowl]

jdubol dead thread!