Voltage Regulator for Peltier cooler project

[xgarb]

Hi guys,

I'm working on what I thought would be a simple project - A solar powered dehumidifier and air cooler. I have a collection of peltier coolers and have been experimenting with various small solar panels and was getting results.

So.. I bought a 100W panel (http://www.ebay.co.uk/itm/2610695128...%3AMEBIDX%3AIT) thinking I could just connect it to a couple of peltiers in series to get the voltage below the maximum but it's not working out. The peltiers get hot very quickly (I have a big heatsink) and the voltage goes up outside their recommended level.

I think to start I need to control the voltage and have been looking at voltage regulators online but they all seem to be designed for charging batteries and I just want a direct connection to the peltiers.

So what do you think I should look at doing? My friend says I'm trying to make a car without a gearbox!

Thanks,
xgarb

[Sunking]

There are a number of regulators you can use. Depends on how much current you are talking about.

As for Peltier, it is an old technology and extremely inefficient compared top today's modern cooling systems with a COP in excess of 3 and EER as high as 26. It will take your Peltier unit 100 watts to equal the cooling capacity of a 20 watt conventional cooling capacity.

[silversaver]

Yes, it takes lots electricity and generated lots heat. You have to spend money to build a proper cooling system. Completely waste of money.

[daz]

I think to start I need to control the voltage and have been looking at voltage regulators online but they all seem to be designed for charging batteries and I just want a direct connection to the peltiers.

I think here you are referring to charge controllers. Charge controllers are designed to charge batteries, and are required to be connected to a battery bank to use them.

A straight forward voltage regulator does not require being connected to a battery bank to work. A common DIY voltage regulator is the LM317, although you can buy commercially built voltage regulators. When selecting your regulator, you will need to keep in mind the regulators maximum ratings with regards to what your peltier's max voltage and current is.


What you need to look at now is how much current (and at what voltage) does your peltier draw (some of the smaller ones draw 7A!). Once you know what the current draw of the peltiers is, you can then get a suitably rated PV panel. <---- when I say "suitably rated", I mean a much much large panel than just the minimum current required. This is because solar panels are dependant on the sun....so clouds, angle, season, etc change the amount of power the panel can give.

So in essence, if you get an appropriately rated voltage regulator (and solar panel), then it should work!

[Mike90250]

The specs on that panel state it produces 18VDC at full power.

Peltier modules have a voltage rating, you may need to stack several in series to get to the 18V rating

The Peltier modules have a hot side and a cold side. The cold side gets 40F cooler than the hot side. Generally, the hot side has to have a large finned heat sink AND a cooling fan, to keep it cool enough to produce useable cold. Expect to have to blow away 200W of heat to get 10 watts of cooling. Proper torque and thermal grease is required too.

[xgarb]

Hi,

Thanks for all the replies.

I understand that the peltiers aren't efficient and heat is a big problem but I have a soft spot for silent passive systems. This is also a fun project rather than a prototype for commercial purposes so it doesn't have to be better than buying something to plug in the wall. Also I would like to use the condensation to water plants and in the winter change the system slightly to dry and heat air with some sort of air circulation internal to the device. I live on the coast in the south of Spain so I have 340 days with no clouds but a fair amount of humidity.

These are the peltiers I bought: http://www.ebay.co.uk/itm/4006111910...%3AMEBIDX%3AIT (Size: 40mm x 40mm x 3.8mm
Voltage: 12V, Umax (V): 15V, Imax (A): 4.3-4.6A, QMax (W): 72W)

I thought if I connected two in series then they would be about the right voltage but in my first experiments it seemed I couldn't wick the heat away fast enough (massive heatsink but no grease) and the peltier (I think) stopped functioning, the resistance dropped and the voltage went over the 15v Max.

Maybe I just need to try again but take more care with the heatsink... or do I need to think about controlling the output from the panel more?

Thanks,
xgarb

[inetdog]

You really want to be able to limit the current through the junctions, not just control the applied voltage.
By limiting the current you can keep the heat transfer within the capacity of your heat sink and avoid damaging the devices.

Note also that the single Peltier device you are looking at already consists of a series stack of many low voltage junctions.
Those junctions are electrically in series and thermally in parallel.

[Mike90250]

.....72w......
I thought if I connected two in series then they would be about the right voltage but in my first experiments it seemed I couldn't wick the heat away fast enough (massive heatsink but no grease) and the peltier (I think) stopped functioning, the resistance dropped and the voltage went over the 15v Max......

2 things kill Peltier junctions. Overheat and Thermal shock. Should never be directly run off a battery, the high current sets up a high thermal gradient, and the shock will destroy the junctions in a couple dozen thermal cycles. Ones we use at work, have slow rate ANALOG control circuits (PWM still causes thermal shock) A PV panel should be OK, as sunlight exposure happens gradually over the morning and evening.

You need at 2 in series, to handle your 18-22V PV panel.

Heatsink. You need a LARGE heatsink. 3x the size of one on a high power computer CPU chip. AND a fan. You only get a 40F differential from hot to cold side. You need to keep the hot side as close to ambient. If its 100F, you can only cool to 60F on the cold side. You must use heatsink grease and the proper mounting torque, or you destroy the parts.

[xgarb]

So.. over 4 years later... I 'finished' the project. I discovered after I completed this that the Peltiers work better when they aren't driven close to their rating so might have got better results with two Peltiers rather than driving one as hard as possible.

If any one is interested in my experiments I blogged it here: https://robotzero.one/solar-peltier-...investigation/ and a Youtube video: