TEC and PWM question

Dunno if I should ask here or not since it's not directly computer related, but I have to do a project for a class and was thinking of using a 50w TEC for it and was wondering if having it PWM controlled from a microcontroller would have any adverse affects. I was wondering if it would find some sort of average temperature between off and running at full power.

Really, no one knows? If I have to be a guinea pig, let me know, I'll gladly try controlling the chip by PWM when the projects over. I just want to know how to wire this thing up before getting too far into it.

I don't see why not, it moves heat when power is applied, and doesn't when it isn't.

PWM is Pulse controled .. i dont think a TEC is going to work correctly under a Pulse signal.

PWM is a common method to control TECs. The TEC actually sees an average power level of the voltage signal (Not V_RMS).

You will want to have a frequency of above 500 Hz, and anything higher is better (I typically use 1kHz in my lab). At low frequency the TEC begins to thermal cycle.

There is also a performance degradation when operating at various duty cycles. The worst performance occurs at 50% duty cycle.

Hope this helps.

I suppose its better to control a TEC using a standard Pot. Just vari the volts.

My plan was to use something like a 3way switch ( Off - 1 - 2 ) to do something like OFF - 8v - 12v using a resistance to produce the 8v state.

Boulard83 said:

I suppose its better to control a TEC using a standard Pot. Just vari the volts.

My plan was to use something like a 3way switch ( Off - 1 - 2 ) to do something like OFF - 8v - 12v using a resistance to produce the 8v state.

Click to expand...

That also works.

PWM is good in a system where you need precise temperature control. Using a TEC with PWM allows you to stabilize temperatures. With a good system you can achieve +/- 0.1C

PWM means Pulse Width Modulation, why wouldn't a TEC work with it?
It turns power on and off very quickly, if it spends 70% of the time on and 30% of the time off, it runs at 70% capacity.

freakdiablo said:

Dunno if I should ask here or not since it's not directly computer related, but I have to do a project for a class and was thinking of using a 50w TEC for it and was wondering if having it PWM controlled from a microcontroller would have any adverse affects. I was wondering if it would find some sort of average temperature between off and running at full power.

Click to expand...

Typically a TEC will draw too much current than the micro-controller can handle (unless you use a very low power TEC). One way of handling this is to have an amplifier board that takes the PWM signal from the microcontroller and translates that to a power signal to the TEC.

An example of this would be the FTC100 or FTC200 controller with FTX700 amplifier board. www.accuthermo.com

If you are interested FreakDiablo, PM me.

Bobnova said:

PWM means Pulse Width Modulation, why wouldn't a TEC work with it?
It turns power on and off very quickly, if it spends 70% of the time on and 30% of the time off, it runs at 70% capacity.

Click to expand...

The TEC actually doesnt see the on/off time if the switching frequency is high enough. The TEC will just see an average of the voltage.

IE A 24v @ 50% Duty power signal @ 1hz frequency would result in the TEC seeing around 12v.

Thanks smittys !

Im also happy to read it

Just one more thing. Im gona use a 245W peltier @ 15.6v
Running this peltier on a 12v rail will result in a 188.5w TEC (15.7amps)
My ohms law is damn far ... ill need a 0.25 ohm resistance to drop my volts to ~8volts at the TEC ?

12 volts on a 15.7 amp tec = .76 ohm
15.7amp on a .25 ohm res = 4 volts
Mean that i have 8volts left for the tec ( Res mounted in series with the tec ).

im right or my class are to far !!!!

SmittysG35 said:

PWM is a common method to control TECs. The TEC actually sees an average power level of the voltage signal (Not V_RMS).

You will want to have a frequency of above 500 Hz, and anything higher is better (I typically use 1kHz in my lab). At low frequency the TEC begins to thermal cycle.

There is also a performance degradation when operating at various duty cycles. The worst performance occurs at 50% duty cycle.

Hope this helps.

Click to expand...

Nice comment Smitty, thanks for sharing. I used to be a lot more into this stuff about 6 years ago, and would have had something useful to add.

I do remember however that there's a sweet spot for TEC's. Something around 80% of their rating is the most efficient point to run them at. At 100% rating, they'll be moving the most heat, however the heat output is proportionally greater (due to more wasted energy being converted into heat) so it may not be worth it if you don't have a cooling system which can handle the greater load. At 80% you'd be moving a bit less heat, but there would also be less wasted energy being output as heat so it could be a bit easier to keep cool. There's a lot of great resources out there to understand this better, I remember reading a lot of them back in the day. Probably a bit rusty, and these numbers are fudged and the idea is oversimplified, but essentially if you research it you'll find them to be roughly accurate.

SmittysG35 said:

PWM is a common method to control TECs. The TEC actually sees an average power level of the voltage signal (Not V_RMS).

You will want to have a frequency of above 500 Hz, and anything higher is better (I typically use 1kHz in my lab). At low frequency the TEC begins to thermal cycle.

There is also a performance degradation when operating at various duty cycles. The worst performance occurs at 50% duty cycle.

Hope this helps.

Click to expand...

Thanks, I'm actually using an Arduino Duemilanove, most of the PWM pins are ~490hz but I think one runs at ~1khz.

Bobnova said:

PWM means Pulse Width Modulation, why wouldn't a TEC work with it?
It turns power on and off very quickly, if it spends 70% of the time on and 30% of the time off, it runs at 70% capacity.

Click to expand...

Don't know, don't really know how TECs work so I thought I'd ask in case there's something that doesn't like being pulsed quickly.

SmittysG35 said:

Typically a TEC will draw too much current than the micro-controller can handle (unless you use a very low power TEC). One way of handling this is to have an amplifier board that takes the PWM signal from the microcontroller and translates that to a power signal to the TEC.

An example of this would be the FTC100 or FTC200 controller with FTX700 amplifier board. www.accuthermo.com

If you are interested FreakDiablo, PM me.

Click to expand...

Oh no, wasn't considering running the TEC off the chip, was just going to hook the TEC up to a transistor and use the PWM out to flip the transistor on and off.

Good Luck Freak!

If you have any further questions on TECs, I'd be glad to help.