Overclock.net › Forums › Cooling › Specialized Cooling › Peltiers / TEC › Scarlet Infidel's TEC Calculator
New Posts  All Forums:Forum Nav:

# Scarlet Infidel's TEC Calculator - Page 2

good luck ..
 My System (13 items)
 My System (13 items)
Er turns out my exam is tomorrow and not today. My mind is all over the place at the moment.

Quote:
 Originally Posted by Ultrasonic2 (muffy) Imax,Umax, dTmax and Qmax ALL increase as Th increases .
I'm not actually using any "max values" in my program. Not yet anyway. The equations I use can work out these values though.

Quote:
 Originally Posted by Ultrasonic2 (muffy) Electricity used increases as the load increase relative to Qmax
Again, I've not used Qmax but I think mine does this? If you set hot side C/W to 0.0, ambient to whatever, this fixes the hot side temperature regardless of load. The set the cold side temperature a bit below. This fixes the dT regardless of load and cooling etc. Then vary the heat load and see the power use change.

Quote:
 Originally Posted by Ultrasonic2 (muffy) As a matter of interest.... is the resistance electrical or thermal resistance ? I looks like your intending on using resistance and couples to define umax, Imax ,,Qmax ... is that correct ?
Yes, its asking for the electrical resistance of the element and the operating temperature that it is quoted for, I have clarified this now in the program.

From this I can work out the ratio between the length and cross-sectional area for each couple. As all peltier elements use the same materials (near enough) this allows me to work out all the specifications of any peltier. This is information that is easy to get for all peltiers I have seen (unlike lambda values etc).
Quote:
 Originally Posted by Ultrasonic2 (muffy) i know it's early days and so on Your electricity used relative to your Qmax is WAY WAY to low
I know, I can't figure out why yet. What equations did you use for your calculator?

I have used one document for all my maths but now have another. I will compare the maths and see if i can find the problem.
Just to make it clear how my calculator is meant to operate (though I imagine you have got it, others might not be so sure).

1. First specify the peltier elements you are using.
I'm using TEC1-12709. They have an electrical resistance of 1.3ohms at 25c. So I type in 1.3 and 25. The program now knows all the specs of my peltier (in theory).

2. Specify the heatload you are trying to cool.
I'm going for 200w so I put in 200.

3. Specify how many peltiers you are using.
I'm using 10.

4. Specify cooling for the whole hotside or for each peltier.
I have heat sinks which with thermal paste included have a C/W of 0.5 or better. So I put in 0.5 and tick the 'per peltier' box because I have 10 heatsinks.

5. Put in the ambient.
22c seems fair for me.

6. Choose the desired cold side temperature. This does not include thermal resistances on the cold side!
I'll go for a modest 17c.
Edited by Scarlet Infidel - 5/21/08 at 3:24am
Results:

The results show that this systems will consume 147w of power to achieve a cold side of 17c with a 200w load in these conditions.

This is a COP of 1.36 as shown. The effective thermal resistance of the system has gone from 0.05 to -0.02. The dT across the element itself is 22.34c (39.34c hotside - 17c cold side).

Without any peltiers you can see that the 'cold side' would be 32c for obviously 0w power used.

e.g. with higher fan speeds my heatsinks achieve 0.4C/W which brings the power down to 108w (COP = 1.84). Alternatively if I decided that 17c is too ambitious I can choose 22c(ambient in my case) as use only 85w. Or I can have both 0.4C/W and 22c set to give 65w power use (COP = 2.34) etc.

Of course, my maths is still not right but that's the idea so far.
Edited by Scarlet Infidel - 5/21/08 at 3:24am
The program can be used to find out all sorts of interesting things, at the moment it's a manual process, but in future I could automate.

Pretend for now the results are correct and accurate.

With the above example (200w, 17c cold side) if we assume my ten heatsinks are one large heatsink with total C/W of either 0.05 or 0.04.

It can be seen by putting these C/W values in and varying the number of peltiers that at 0.05 C/W 10 peltiers is optimal and at 0.04 C/W 12 is optimal. To me this is the kind of interesting information you can't guess at.

So basically, if I was to stick 2 peltiers under each of my heatsinks instead of one (each running at lower voltages/currents to achieve the same overall cold temp), id end up with a higher power use overall (at those exact operating temperatures only!).

You can also see things like sometimes its more efficient to use lower rated peltiers. I just hope these trends are true and not just a side effect of my incomplete maths.
Ultrasonic: I'd be interested in what you estimate your ideal chillers specs are?

Whats the estimated C/W of the cooling loop you'd use?
What load are you trying to cool?
What temperatures are you trying to achieve? (and the C/W of the cold side block).
How many peltiers and of what type were you going to use?
Edited by Scarlet Infidel - 5/21/08 at 4:59am
Quote:
 Originally Posted by Scarlet Infidel I know, I can't figure out why yet. What equations did you use for your calculator? I have used one document for all my maths but now have another. I will compare the maths and see if i can find the problem.
I took a working app and revers engineered it to get my numbers

but for electricity used i mapped it V input voltage
 My System (13 items)
 My System (13 items)
What are the stats on the TEC your using MAX ?
 My System (13 items)
 My System (13 items)
Quote:
 Originally Posted by Scarlet Infidel Ultrasonic: I'd be interested in what you estimate your ideal chillers specs are? Whats the estimated C/W of the cooling loop you'd use? What load are you trying to cool? What temperatures are you trying to achieve? (and the C/W of the cold side block). How many peltiers and of what type were you going to use?
um well my current cooling loop is 0.1 with that fans at 32% but most of that is coming from CPU block. im not that impressed with my fuzion. I probably should have gotten the nozzle kit. so my hot side C/w should be NO worse than 0.1 . obviously i'll have a similar cold side c/w or 1/2 that if we're talking to watertemps not cpu temps
i think these numbers are really worse case though.

my CPU is a calculated 190watts i'll have a 18 watt cold side pump and a total of 36 watts worth of hotside pumps.

I think im going to be shooting for a 0c water temps e relevant of load and ambient . because as i decrease the input V Umax the dTmax decreases making super low temps impossible.i of cuase want a low input voltage to get increase electrical efficiency.

im shooting for a hot side temp of 35-40c. which i'll try to keep it at this e relevant of ambient temps it's this high cause in summer there were quite a few 29c ambient temps

Also i've seen some rather interesting data on phase change slush boxes. in Short even using a VERY high % of anti freeze the fluids viscosity increased making it hard to pump around the system so there was a point in which low fluid temps were negated by the reduction of flow rate. also the increase in anti freeze decreased efficiency. from memory -10c is about the lowest practical temp . but it was along time ago i read this now

i think i want to run the TEC's between 15-30% of Umax
so for 20% which is what im thinking im gonna go for i want to get 1000watts of rated Qmax and run them at 20% is of umax

Though in the end im not 100% sure 20% of Umax will give me enough of a dTmax to achieve a loaded dT of 40c.

i just worked out the dTmax at 20% and it's an unloaded 30C !

You know this could actually be a good case for stacking TEC !!

i'll probably be using the 226 modules or 437's im still very skeptical about the 437 modules and would love to find the manufacturers data sheet on them as i get the feeling there not rated at 25c/300k
Edited by Ultrasonic2 (muffy) - 5/21/08 at 2:28pm
 My System (13 items)
 My System (13 items)
New Posts  All Forums:Forum Nav:
Return Home
Back to Forum: Peltiers / TEC
• Scarlet Infidel's TEC Calculator
Overclock.net › Forums › Cooling › Specialized Cooling › Peltiers / TEC › Scarlet Infidel's TEC Calculator