Contact Thermometer for Component Temperature

for things that small I'd either use non-contact sensing or a very tiny thermocouple bead and silver paint such as in those pcb repair pens. a the rectal probe would act as a heatsink and defeat the purpose.

al

Tektronix had a temp probe for their early DMM's I think fluke did too...80T-150U. There are surely newer versions.

There are two big problems with contact thermometers on small components.

Look for thermocouple probe. Some multimeters take one directly (mine will work with Type K only).

ie, newport 48F5805 is a 48 inch type K with multimeter type plug installed, for $13. Also says no longer available, but 7 in stock... Stick it in a tube and it's a probe. Hot glue is your friend.

A rigid probe prebuilt would be (for example) 93F9285, at $30.

Sometimes you can find them in surplus.

You could also stick any of the TO92 devices (LM34, etc) on the end of a stick/tube. Angle the mount on the stick so you can get the flat on the flat of the case to be measured.

If you like you can simply buy the wire and weld it yourself to make thermocouples. You can even get retro and use a bucket of ice and a chart, instead of a meter to figure it all out for you and fake the bucket of ice.

k..Contact may suck but what's available in non-contact that can focus on the heat from a 1206 resistor on a high density pcb?

For contact.. I did find a bead. P60 1.5mm diameter

RTD in 0603 is interesting.

Thermistors go as small as 0201.

D from BC British Columbia Canada.

Many moons ago, I used temperature sensitive paint to read the temps of components. You had to guess at a temp, select a paint, put a dab on the component, fire it up and see if the paint changed color. If not, then use a lower temp. paint, and repeat. Yes, it was time consuming.

Check this out:

Goggle this for more: temperature sensitive paint

Al

Perhaps the surface probe, 80PK-3A

or the SST-2

or the 50012 Bell Surface Probe

I tried that with a SOT-23 TMP36 sensor.

(1)Suspect a number of causes. One is: Thermal shock. The sensor went in microwaved water for calibration and testing. Turns out, this is a method to get superheating.

D from BC British Columbia Canada.

Haven't found tip dimensions yet.

Ewwww... :P They have mice and rat rectal probes. Insect probes too.. I think my grasshoppers are running a little hot... :)

D from BC British Columbia Canada.

That's a simple matter of optics. I think the more difficult problem is knowing where it's pointed without using a heat-generating device.

Might be worth experimenting with a "funnel" Poke the small end down on the part and see if you can make some calibration curves for a bigger IR probe. You've already got calibration problems setting emissivity for IR probes. Back in the day, I heard that you could use spray-on foot powder to normalize emissivity across a board. Tried it. WORKED GREAT!! Problem was that I couldn't wash the stuff off.

You really have to glue contact probes on if you expect anything resembling repeatability.

Or just bite the bullet and call up Flir. I'm sure they can fix you right up.

For thermocouples, there are some attached to flat pieces of steel. You migtht want to customize one or make one.

For thermistors. I have a probe in a small hypodermic needle, and I have a probe on a long stainless steel bendable rod about .04 inch thick. Nether one would be good for making contact. Your best bet is to do something like me. I ordered some chip thermistors. I didn't buy it beacause it was flat, but that would work in your case. Still need some grease to get some conduction.

greg

Something like that should do the job. Not sure what their "extremely affordable" might mean, did not bother to check if the resoultion is good enough either, just searching for "IR video imaging" is what brought me there. I could use something like that in the context you (most of us?) need it, though.

Dimiter

------------------------------------------------------ Dimiter Popoff Transgalactic Instruments

------------------------------------------------------

D from BC wrote:

The exposed junction version looks interesting. Wow! A thermocouple with a 0.254mm sheath diameter for about $48.00 If there's a huge amount of thermal resistance between the junction and everything else, it might be ok for rough surface measurement. However I suspect the probe is designed to be poked into something..

D from BC British Columbia Canada.

Yup... I'm thinkn chip thermistor too. It's the cheapest solution. A blob of thermal epoxy is more expensive. :P

So I glue a bunch a thermistors on a few parts and it looks ugly. It's just for a scruffy prototype. I might try grease instead but seems messy.

I plan to connect small diameter wire to thermistors reduce heat piping.

D from BC British Columbia Canada.

Interesting. But this might be resolution limited at the range/size most SMT electronics need. According to the spec, it has a min focusing range of 0.3m and resolution of 1.3 mrad. That would make a 1206 resistor (3.2 x 1.6 mm IIRC) about 8 x 4 pixels.

Such a device with a macro focus might work quite nicely, but I'm unfamiliar with that end of the technology. Most of my work involves things that are meters across and many meters away. None of this wussy little electronics. If it doesn't need a crane to put it in place, its not a real component. ;-)

I have not tried this myself... but I gather you can get the junction temperature of a part by measuring the forward voltage drop of the internal input ESD protection diodes. I guess this would amount to sticking a few negative microamps in and measuring the voltage. You could calibrate first with an external thermometer and an unpowered device. Etc.

Well not so bad even so, sounds like being (just about) usable.

Would be perfect, but I am unfamiliar with IR optics either. Slaughtering the device and moving its lens somewhat further from the CCD might well work, too.

Ah, reminds me of a friend of mine who did - well, by far not that large - power invertors, their SCRs were the size of a teacup perhaps, and they used plenty of them per system. Once they goofed up and the bang could be heard hunreds of metres away, the blackout lasted quite some time... Me, I remember times when I first experimented with mains (220 AC) SMPS designs (about 25 years ago...), I had have my share of exploding

Dimiter

------------------------------------------------------ Dimiter Popoff Transgalactic Instruments

------------------------------------------------------

Paul Hovnanian P.E. wrote:

Yep. I do it all the time.

...Jim Thompson

Neato... Is this done with the chip running or is it a warm up, shut down, inject and measure routine?

An unpowered chip can be calibrated with boiling water or the known boiling point of perhaps a propylene glycol+water mix. No thermometer needed.

D from BC British Columbia Canada.