Peltier (TEC) + Liquid Cooling Project

myom · Jul 21, 2014

Do fans cool down heatsinks all that much? My current TEC hardly does any cooling bc the hot side gets way too hot, so I'm looking for a small fan to mount.

This is one that I could find: http://www.amazon.com/Scythe-Mini-U...=UTF8&qid=1405953179&sr=1-2&keywords=mini+fan

--

On a side note, since I won't be using a reservoir in this loop, is a T-line necessary? Or can I get by without one?

Bobnova · Jul 21, 2014

Adding a fan to a hot heatsink makes a massive difference.

Keep in mind that with only 1.5w of power going to/through the TEC you're only going to get around 0.75w of heat moved from side to side, while the hot side you're going to have to get rid of both the 1.5w and the 0.75w.
How cold the cold side gets with a given voltage and wattage being applied is based primarily on two variables:
1) How hot the hot side is.
2) How much heat transfer there is from the ambient environment into the cold side.
Once the cold side goes below ambient the air it is in contact with is heating it directly.

What voltage are you powering the TEC with?

myom · Jul 21, 2014

Bobnova said:
Adding a fan to a hot heatsink makes a massive difference.

Keep in mind that with only 1.5w of power going to/through the TEC you're only going to get around 0.75w of heat moved from side to side, while the hot side you're going to have to get rid of both the 1.5w and the 0.75w.
How cold the cold side gets with a given voltage and wattage being applied is based primarily on two variables:
1) How hot the hot side is.
2) How much heat transfer there is from the ambient environment into the cold side.
Once the cold side goes below ambient the air it is in contact with is heating it directly.

What voltage are you powering the TEC with?

It can take 12V and with some regulation, that gets shot down to .75A x 1.8V = 1.35W

Did some math and the minimum thermal conductivity I'll need to cool the hot-side is 0.24333 W/mK

Bobnova · Jul 21, 2014

That's quite possibly why it's getting so hot, the poor thing is only rated for 0.85V.
The higher the voltage the lower the efficiency. You're probably pulling 0.1w or less of heat across it.

myom · Jul 21, 2014

Bobnova said:
That's quite possibly why it's getting so hot, the poor thing is only rated for 0.85V.
The higher the voltage the lower the efficiency. You're probably pulling 0.1w or less of heat across it.

I'm testing with 4V though, which would regulate it to much less than 1.85V and probably less than 0.85V.

Anyways, I'll still be needing a small fan to mount. How can I mount a fan to a heatsink of such a small size?

RnRollie · Jul 21, 2014

you can't, you'll have to mount a bigger heatsink and bolt the whole thing to a bracket or something where it can sit in the flow from the fan.
Should be possible to do with a simple L- bracket.

And.. i'm still sure you need a bigger tec & wb (40x40mm)

Bobnova · Jul 21, 2014

You'll need to mount the fan somewhere else and aim it at the heatsink. Maybe run a duct or something to focus the air.
Air movement over the cold side will warm it up significantly.

Silver_Pharaoh · Jul 21, 2014

myom said:
I'm testing with 4V though, which would regulate it to much less than 1.85V and probably less than 0.85V.

Anyways, I'll still be needing a small fan to mount. How can I mount a fan to a heatsink of such a small size?

You could always try and find the tiny fans used to cool off old Dell Latitude's.
I have one of those old laptops, and the fan is about 1" wide. Pretty much a 30x30 fan :shock:

Then again, they don't move much air.

myom · Jul 23, 2014

I currently need to power a 12V/1.5A pump, a 12V/0.16A fan, and a TEC. Is this achievable with one 12V/1.25A power supply? Not really sure how power supplies work

Bobnova · Jul 23, 2014

No, the pump alone is too much of a load for the power supply. The power supply needs to be able to serve up more power than the devices that are attached consume.

myom · Jul 23, 2014

Bobnova said:
No, the pump alone is too much of a load for the power supply. The power supply needs to be able to serve up more power than the devices that are attached consume.

Thanks! How do power supplies take on multiple loads? Do I just add up (max voltage x current usage) for each device, and this has to be less than the power supply's wattage?

Bobnova · Jul 23, 2014

Pretty much, yeah.
Whether to add wattage or add amperage depends on the type of regulation you're using.

If you have a linear regulator doing the regulating the amperage consumed by the device is the same as the amperage of 12v used. If it's a buck regulator, the wattage is the same (roughly) and the amperage of 12v is smaller.

If all the devices are 12v and the PSU is 12v then it's easy and you can add amperage or wattage.

In your case you have a 1.5 amp pump and a 0.16 amp fan, so 1.66 amps in total. The TEC it depends on the TEC of course.
At 0.75v yours draws 1.5w if I recall correctly, so about 2 amps of 0.75v.
If fed 12v it will draw something more like 64 amps and explode.
Fed 0.75v with a linear regulator the regulator+tec combo will draw 2 amps (24w at 12v), and odds are the regulator will overheat and fail.
Fed 0.75v with a buck regulator, the regulator+tec combo will draw about 1.7w, or ~0.14 amps of 12v.

In any event, the total combined usage needs to be lower than the PSU's maximum output, or the PSU will likely fail.

myom · Jul 24, 2014

Bobnova said:
Pretty much, yeah.
Whether to add wattage or add amperage depends on the type of regulation you're using.

If you have a linear regulator doing the regulating the amperage consumed by the device is the same as the amperage of 12v used. If it's a buck regulator, the wattage is the same (roughly) and the amperage of 12v is smaller.

If all the devices are 12v and the PSU is 12v then it's easy and you can add amperage or wattage.

In your case you have a 1.5 amp pump and a 0.16 amp fan, so 1.66 amps in total. The TEC it depends on the TEC of course.
At 0.75v yours draws 1.5w if I recall correctly, so about 2 amps of 0.75v.
If fed 12v it will draw something more like 64 amps and explode.
Fed 0.75v with a linear regulator the regulator+tec combo will draw 2 amps (24w at 12v), and odds are the regulator will overheat and fail.
Fed 0.75v with a buck regulator, the regulator+tec combo will draw about 1.7w, or ~0.14 amps of 12v.

In any event, the total combined usage needs to be lower than the PSU's maximum output, or the PSU will likely fail.

Where are you getting 0.75V? I'm currently testing without any regulation - just trying to power all 3 devices (pump+fan+tec) with a single power supply. Did you mean the total combined power usage needs to be lower than the PSU's max output?

I've got my hands on a 12V x 5.42A = 65.04V power supply now.
Pump: 12V x 1.5A = 18W
Fan: 12V x 0.16A = 1.92W
TEC: regulated by diodes, so not quite sure what its power usage is, but the sum of pump+fan+tec is definitely less than 65.04V.

Could I power all of these in series without damaging anything (PSU or the 3 devices)?

Silver_Pharaoh · Jul 24, 2014

myom said:
Where are you getting 0.75V? I'm currently testing without any regulation - just trying to power all 3 devices (pump+fan+tec) with a single power supply. Did you mean the total combined power usage needs to be lower than the PSU's max output?

I've got my hands on a 12V x 5.42A = 65.04V power supply now.
Pump: 12V x 1.5A = 18W
Fan: 12V x 0.16A = 1.92W
TEC: regulated by diodes, so not quite sure what its power usage is, but the sum of pump+fan+tec is definitely less than 65.04V.

Could I power all of these in series without damaging anything (PSU or the 3 devices)?

0.75 volts is what the TEC is rated for if I'm reading correctly.

myom · Jul 24, 2014

http://www.customthermoelectric.com/tecs/pdf/00711-5L31-03CA_spec_sht.pdf

Rated for 0.85V. Also, my TEC is linked with diodes in series on each lead (http://www.farnell.com/datasheets/78621.pdf). Now I'm wondering how the TEC isn't frying when I'm supplying more than its Vmax of 0.85V.

I did some testing by driving a dc power supply straight into the TEC:
(input voltage -> voltmeter reading): 4V -> 2.55V, 3.5V -> 2.4V, 3.0V -> 2.1V, 2.5V -> 1.8V

myststix · Jul 26, 2014

freakdiablo said:
I still think TEC is the better way to go. As you said in your OP, you can use it to both heat and cool the liquid. With a traditional watercooling set up, you're still going to need some sort of heating element, which I feel is only going to complicate it more than simply slapping a large heatsink on it.

Your going to have a hard time removing that much heat fast enough from such a small body. You will need to know the specific heat capacity of the material, that is, how much of a heat sink it is going to be. Then you need to calculate how fast you can remove heat. After you run these numbers, you will know if it is possible to remove the heat faster than it absorbs it.

Just roughly from these numbers, I don't think such a small mass can absorb and then dissipate that much heat. Have you considered making the scale larger? Like 2x or 3x scale?

myom · Jul 29, 2014

I've got a little test set-up going:

Water block + TEC + heatsink:

Water doesn't seem to be cooling up all that well. I'm using a 12V TEC but only driving 4-5V through it bc the hot-side overheats the cold side. Fan blows away from heatsink (sucks).

These are a few things I'm thinking might be the problem:
- Water blocks are too small (15x15x5mm)
- Need better thermal compound (i.e., arctic silver) b/w TEC and water block/heat sink
- Need better heatsink on hot-side of TEC (stuck on the remaining 5 little copper heatsinks I had left)
- Need stronger fan (is this ever the case?)
- Need better coolant than tap-water (worth buying?)

RnRollie · Jul 29, 2014

yes, you need a bigger waterblock.. one that covers the whole tec
likewise you need bigger cooling of the hot side of the tec (cover the tec) something like this http://www.performance-pcs.com/cata...oduct_info&cPath=46_185_539&products_id=32261
or better ... goto local junkyard, recover a PC

and if you want some longer life for the pump & blocks, use distilled water , its cheap

myom · Jul 30, 2014

RnRollie said:
yes, you need a bigger waterblock.. one that covers the whole tec
likewise you need bigger cooling of the hot side of the tec (cover the tec) something like this http://www.performance-pcs.com/cata...oduct_info&cPath=46_185_539&products_id=32261
or better ... goto local junkyard, recover a PC

and if you want some longer life for the pump & blocks, use distilled water , its cheap

How does that heatsink compare to this or any of these?

(Trying to take advantage of Amazon Prime)

Also... are there water blocks without fittings that are pre-attached? I can't find any that are at least 40x40mm and have barbs that 1/8'' tubing could fit over...

Connector dimensions aren't listed, but this looks like it might work? (the block is 40x32mm)

myom · Aug 7, 2014

Sigh.. cannot seem to get this to cool! I've got a waterblock, TEC, heatsink, and fan that are all about 40x40mm and are coupled together as shown in the image below. The water can heat up quite fast, but it can't seem to cool down (AT ALL!). Any insights?

Peltier (TEC) + Liquid Cooling Project

Registered

Senior Member

Registered

Senior Member

Registered

Member

Senior Member

Likes the big ones n00b Member

Registered

Senior Member

Registered

Senior Member

Registered

Likes the big ones n00b Member

Registered

Disabled

Registered

Member

Registered

Registered

Similar threads