0
\$\begingroup\$

I need to build a resistor that will serve as a heat source to test the performance of a Peltier cell.

I want the resistor to have a square plate geometry, of 40 mm × 40 mm and a height the shorter the better. The length and width can grow a bit. However, I am having trouble selecting the material I should use.

I can use power sources preferably of 12 V or 120 V, and I want the resistor to dissipate between 5 W and 10 W. The solution of the resistance estimation equation $$R = \rho \frac{L}{A}$$ yields that I need a material of resistivity between 0,14 Ωm and 0,30 Ωm if I use a source of 12 V and between 14 Ωm and 30 Ωm for a 120 V source.

How would you go about this? I was thinking of using salt water but its resistance is said to vary highly with temperature.

\$\endgroup\$
3
  • \$\begingroup\$ How hot does it need to get? Three common technologies in use for heating by the 3D printer folks are PCB track heaters, hand wound nichrome wire assemblies, and off-the-shelf power resistors- typically used with surplus switching supplies in the 12-24v range. \$\endgroup\$
    – Chris Stratton
    Commented Nov 20, 2014 at 17:46
  • \$\begingroup\$ Any reason why you're not just using a commercially available hot-plate or resistive heating coil? Like this? \$\endgroup\$
    – Dave
    Commented Nov 21, 2014 at 1:08
  • 1
    \$\begingroup\$ A heat plate indeed sounds like a good idea! Thanks for the suggestion. The reason would have been that I knew not about heat plates. I have very limited knowledge about the available hardware :) \$\endgroup\$
    – Severo Raz
    Commented Nov 21, 2014 at 4:16

3 Answers 3

Reset to default
1
\$\begingroup\$

I would definitely just buy a resistor and fabricate a copper heat spreader maybe 6mm/0.25" thick to match the Peltier, then use a bench DC supply to provide the power.

This one is rated for 10W (for example, 10 ohms, so 10V will give you 10W), and is only a few dollars.

enter image description here

It's 19.05mm x 20.63mm, so will fit easily on a 40mm x 40mm x 6mm copper heat spreader with a couple of tapped holes (the spreader can be made quickly, even with hand tools).

\$\endgroup\$
1
  • \$\begingroup\$ If I could find something like this in a local store that would be awesome! I'm going to look around and see what I can find. \$\endgroup\$
    – Severo Raz
    Commented Nov 21, 2014 at 4:15
0
\$\begingroup\$

Just make a PCB with one very long, thin, serpentine trace, zig-zagging to fill the area you need. The resistance will be relatively low, but it will be easy to produce. If you go through oshpark or similar, it should cost less than $20.

\$\endgroup\$
2
  • \$\begingroup\$ Thanks, I might use this. Seems to me it would be pretty similar to a heat plate like the one Dave suggested in his comment to my question, am I wrong? \$\endgroup\$
    – Severo Raz
    Commented Nov 21, 2014 at 4:14
  • \$\begingroup\$ More or less. That one looks like it might be made on some sort of flexible substrate. \$\endgroup\$
    – alex.forencich
    Commented Nov 21, 2014 at 4:29
0
\$\begingroup\$

Does it have to be a square plate? Might be easier to control a wire laid out across a non-conducting plate of metal (like a heatsink).

From what I understand, you're heating your peltier with said source, and I assume you're going to measure some numbers from it, meaning the 5W-10W consumption will yield different temperatures, dependent on the type of material you're using, and this will affect your peltier's behaviour?

Wouldn't it be better to control the temperature difference between the two sides of your peltier?

Grab some nichrome wire, line it up on the hot side of your peltier, measure the temperature difference, and see your voltage/current change?

You can control nichrome pretty easily, just follow these tables, it's a pretty predictable and commonly used material so it might make your life easier? You can control length of wire to obtain total resistance you need to pull specific amount of current for a specific temperature.

\$\endgroup\$
1
  • \$\begingroup\$ Yes it better be a square plate so there is as little waste heat as possible. I am modeling very coarsely a portable CPU unit, so I am interested in the waste heat as a parameter. The temperature is my variable to measure. But sure, nichrome sounds like a great idea! \$\endgroup\$
    – Severo Raz
    Commented Nov 21, 2014 at 4:14

Not the answer you're looking for? Browse other questions tagged or ask your own question.