Peltier

[F Murtz]

Anyone know about peltier devices? Can you drive them at half voltage? I wish to make a container that is limited to ideally between 18 and 20 centigrade and apparently they do not like to be cycled much for temp control,I would rather have it on constantly but at less than full capabilities. On further musing I would probably still have problems as the system is dependent on ambient temp so I would probably have to have a temp sw to switch on only when ambient is higher. Anyone have ideas on how to keep between those temps using peltier? It is for storing chocolate so it could probably be alright at lower temps but not higher.

[Fredxxx]

Without looking anything up, from memory the cooling effect is dependent on the current. There is a potential barrier to overcome sort of akin to a diode. However they are rather lossy, with a high series resistance where I2R losses can be significant.

Most decent peltier devices come with a series of graphs to give you cooling capacity for a specific temperature difference across the device for a specific I/V drive.

Is there one you have in mind?

[F Murtz]

Thinking of this, only a few $AU on ebay

12V 60W TEC1-12706 12v 60w

[whisky-dave]

Yes I've messed with one of those, I used an arduino PWM output which runs at about 500HZ I changed the pulse width to change the amount of cooling or heating. Using a TP31 transitor. I used a CPU cooler on the other side as I found that I couldn;t get much cooling or heating for any lengh of time without finding a way to 'disipated' the heat or cold from the other side of the device.

[Syd Rumpo]

You shouldn't PWM a Peltier device directly, at least not if you're using it for cooling. Cooling is proportional to current, but the unwanted ohmic heating is in proportion to current squared (I*I*R). Think about it - a 50% PWM at 1A peak in a 2ohm device will produce 1W ohmic heating (I*I*R for half the time) whereas a continuous (100%) 0.5A would produce 0.5W ohmic heating.

You can of course use PWM indirectly if you filter it to give a fairly steady current.

Cheers

[whisky-dave]

Well I haven't seen any info that says I can't run it from PWM. There's sources that say don;t run it in the 10s of Hz .

[Syd Rumpo]

You just did, see above. Apart from that, you'd need to look. If only there were some way of searching the Web for key phrases such "PWM Peltier".

Cheers

[whisky-dave]

runs at about 500HZ I changed the pulse width to change the amount of cool ing or heating. Using a TP31 transitor.

much cooling or heating for any lengh of time without finding a way to 'dis ipated' the heat or cold from the other side of the device.

I did I have done.

-element

Most of the microcontrollers I am familiar with do PWM in the hundreds of c ycles per second range. No thermoelectric module is going to be able to dis tinguish that from a steady voltage.

Also, (google it) there is a paper out there where they tested PWM cycling with rates of 1/10s all the way up to 1000/1s rates and the peltiers did no t exhibit any decline in performance over thousands of hours. The one that cycled every 10 seconds did exhibit temperature fluctuation due to the slow response time.

In any case, PWM is utterly safe for controlling a peltier.

[Syd Rumpo]

You do know that you're meant to read this stuff? Your own link says, "Peltier devices are one of the few things you do not want to run with pulses, particularly in cooling applications...", and then goes on to explain why.

Rubbish. What magical property does the Peltier module have which allows it to integrate PWM? Peltier modules have no reactance other than strays.

Link?

It's safe, of course, but quite inefficient when cooling, for reasons explained ad nauseum. Best stop digging.

Cheers

[Fredxxx]

Wrong.

It is utterly safe. However it's not a smart way to run a peltier as already explained to you.

First hit will assist you:

Constant current is best. It you really need to drive it from a PWM, make the frequency as high as you can and stick a big inductor in series to make the current as constant as possible.

If you don't understand the different between average current and rms current in terms of resistive heating may I suggest you go away and find out. There is a very big difference.

[whisky-dave]

ch runs at about 500HZ I changed the pulse width to change the amount of co oling or heating. Using a TP31 transitor.

t much cooling or heating for any lengh of time without finding a way to 'd isipated' the heat or cold from the other side of the device.

Yes I did. It gives two sides to it, those that thibnk you can and those th at can't see why you can't use PWM.

Your own link says,

But doesn;t explain it.

of cycles per second range. No thermoelectric module is going to be able to distinguish that from a steady voltage.

it doesn't need a magical property any more than a light dimmer does.

ing with rates of 1/10s all the way up to 1000/1s rates and the peltiers di d not exhibit any decline in performance over thousands of hours. The one t hat cycled every 10 seconds did exhibit temperature fluctuation due to the slow response time.

yes I want a limk that explain why PWM can't be used.

there was NO explanation shown that made sense.

[Martin Brown]

There are two scenarios when driving the Peltier at say 1/4 max power.

The device has to pump both the heat from the cold side kI and any resistive losses generated inside the device I^2R to the hot side.

So if you want to run it at I for 1/4 of the time and 0 for 3/4

Heat output = kI/4 + I^2R/4

But if you run it at I/4 continuously (ie filter the PWM drive to DC)

Heat output = kI/4 + (I/4)^2R = kI/4 + I^2R/16

In other words the pulse mode driven PEC is more inefficient turning a greater proportion of the higher drive current into waste heat.

Driving it with DC you can either get the cold side colder for the same input power or use less input power to obtain the same cooling.

[Syd Rumpo]

Riiight.

Cheers

[Fredxxx]

You do know that you're meant to read this stuff?

That is so, so true.

It would to someone versed in basic electronics.

[whisky-dave]

it was explained that it was not as efficinet where they advertised their c ontrollers, yes, which are also way outside the student budget.

The "ON" and "OFF" pulses occur so rapidly that the module does not have en ough time to change temperature in response to each electrical pulse. Inste ad, the module assumes a temperature difference relative to Vaverage. When the controller is properly tuned thermal cycling is eliminated. Thus, these controllers do not degrade the reliability of a module from thermal cyclin g in the same way that a thermostatic or slow "ON-OFF" controller would.

as said the diferce is in teh speedd of switching.

One thing our students get confused by is that a square wave pulse that goe s from 0V to a higher voltage is NOT AC. it's still DC as current only flow s in one direction.

0.707 compared to 0.636 isn't it for PWM square wave it would vary.

The RMS voltage is the voltage that would give the same heating effect as A C in a resistive load. That's basiclly where RMS comes in as meaingful in power.

[whisky-dave]

But you can't explain I see.

[Fredxxx]

I provided a link that did. I'm sure it gave a better explanation than I could have done, or wish to have done.

Especially after you seemingly have difficult in basis comprehension understanding Syd's reason why PWM is not an ideal way to drive a peltier.

[Fredxxx]

So when you say "No thermoelectric module is going to be able to distinguish that from a steady voltage." I'm pleased that you seem to have learned something new.

Alternating does give the game away. However, there are many instances where the voltage is not "alternating" in your eyes, where the current does.

Your quoting numbers to 3 decimal places for an arbitrary PWM waveform speaks volumes of your engineering skills.

I find it quite disturbing you might be teaching students anything like this.

[brass monkey]

I dunno what the prob is, my bipolar PWM drive (circa 1980) worked fine for years and may still be working ;) I never researched losses, inefficiencies etc etc, someone wanted something cooling fairly accurately, thats what they got :D

[Fredxxx]

No one here is doubting that a cheap and quick setup of PWM drive does not work. It is a simple form of control ideal where efficiency is not important.

What this thread has demonstrated is that someone who teaches students doesn't seem to understand that PWM drive might be less efficient than with a constant current.