Thermal sensor in a heatsink.

[thepaan]

I recently purchased a heatsink, model: U81C-60NMC, manufacturer: Alpha Novatech, from this web site:

**broken link removed**

The heatsink has what I assume to be a thermal sensor or switch embedded in the base which I would like to use for the LED lamp I am building but I cannot find any specifications for it. I e-mailed the web site from which I purchased the product but they told me they didn't have a data sheet for it. The heatsink is intended for an AMD Opteron 2U rack mount chassis so I thought there might be a standard connector for it on some motherboards but I can't find any such thing either.

I am baffled. Is it a sensor or a switch? If it is a switch does it open on rise or close on rise? Why does the sensor have only two wires? how am I to interpret data from such a device? I need an understanding of it only so far as is required to integrate it into the circuit for the LED lamp so a full data sheet is not necessary. For example, if anyone knows of a standard temperature sensor which connects to a motherboard, what the behavior of such a thing is expected to be, then I can reasonably assume that to work.

Any help would be appreciated.

 

[Chippie]

Eliminate the fact that it could be a switch by measuring with a dvm...It will be closed until the temp rises to its setting...Heat the heatsink with a hot air gun to raise its temperature...

If its not a swtich, then you should read something quite different than zero ohms..If the front-back and b-f resistances are the same, it could well be a thermistor..If the readings are different then it may be some sort of semiconductor device

 

[thepaan]

I just measured it with a multimeter from work and a AAA battery (used some twisted pair copper to connect everything ). The multimeter reads 1.8V and 2.2 ohms. When I connect it in reverse on the ohm setting the multimeter reads 1 and on voltage reads -1.8.

I don't have a hot air gun. Do you think it would be OK to put it in a pan on the stove?

 

[thepaan]

OK, so I thought I had some funky readings because I reread your post and then I measured again and thought I misread the numbers the first time but now I've measured again and the forward voltage past the battery is going up. Originally I was getting 1.8V in both directions, 1.8 ohms forward and in reverse the mulimeter flipped out and turned off. Now, after sitting on the floor for a few minutes (it was upside down on my desk earlier), it reads ohms as my previous post but the forward and reverse voltages are .002 and -1.2 respectively.


Soooo. It's some sort of semiconductor device?

 

[thepaan]

I just did some more testing - this time without the battery. I get no readings when the device is connected backwards which means it isn't a resistor. I get 5.18MΩ when I select the ohms display and .660 on another setting where it shows at the bottom what looks like a sound wave and next to that a schematic diagram for a diode.

I think this means it is a diode. Anyone have some pointers on how to use a thermal diode in a circuit? I'm googling right now but so far no luck....

 

[microtexan]

If you tried to measure the resistance with the battery connected then you may have blown a fuse in your meter. This would explain the odd readings you are now getting.

 

[thepaan]

I'm pretty sure the multimeter is fine. I've tested the device about a dozen times now (still sitting on the floor) and it gave me consistent results all evening. I believe the initial wonky results are caused by the temperature difference between the air around my desk and the surface of the floor. And, even this site, which I was using for reference, has a battery connected - though on the voltage setting. Furthermore, I have a hard time believing I can burn out a multimeter with a AAA battery.



Now, my multimeter has the continuity, diode, and Ω test on the same dial setting. A selector button allows the user to toggle between Ω and continuity/diode.

According to this site, to test for polarity, one can connect a diode to an ohmmeter and when it is connected backwards it will read OL, which mine does.

If an ohmmeter shows a value of “1.73 ohms” while forward-biasing a diode, that figure of 1.73 Ω doesn't represent any real-world quantity useful to us as technicians or circuit designers. It neither represents the forward voltage drop nor any “bulk” resistance in the semiconductor material of the diode itself, but rather is a figure dependent upon both quantities and will vary substantially with the particular ohmmeter used to take the reading.

When connected properly I get 5.18MΩ which tells me that is the correct orientation.

The forward voltage reading obtained with such a meter will typically be less than the “normal” drop of 0.7 volts for silicon and 0.3 volts for germanium, because the current provided by the meter is of trivial proportions.

When I select diode test it reads .660 in this orientation which leads me to believe it is made of silicon. All the tests so far have been consistent with that of a diode.

When using a multimeter which outputs a very low test voltage to the probes in the “resistance” function mode, the diode's PN junction will not have enough voltage impressed across it to become forward-biased, and will only pass negligible current. ... If such an ohmmeter were used to test a diode, it would indicate a very high resistance (many mega-ohms) even if connected to the diode in the “correct” (forward-biased) direction.

Again, my reading is consistent with a diode which doesn't have enough forward voltage applied to it. If I connect the battery the proper way, to provide additional forward voltage, am I correct in assuming that the value of Ω will go down? How does temperature have an effect on the functionality of a diode?

 

[blueroomelectronics]

It's a thermal diode, you need to amplify the output voltage as it's very tiny.

 

[thepaan]

from Designing with Semiconductor Temperature Sensors | Sensor Tips - Pressure, Temperture, Photoelectric, Vision, Wireless Sensors, Safety Sensors, and more.

A silicon bipolar diode (a.) is the temperature-sensing element. When biased with constant current, its voltage drop varies with temperature.

Is the voltage drop increased or decreased with a higher temperature?

 

[blueroomelectronics]

This may help.
Simple & Cheap Thermometer

 

[thepaan]

Thanks. I think that will do.