Thermocouple vs IR for electronics troubleshooting

[TimNJ]

Hi everyone,

As usual, I'm trying to make excuses to buy a new piece of test equipment. I've seen Dave use a thermal imaging camera to find trouble areas on a board. However, I am no where in the market for a thermal camera, as great as they may be.

I have a multimeter which can use a K-type thermocouple. It seems pretty accurate. However, I often see people using infrared thermometer guns and I'm curious as to what the benefits of using one are. I suppose it may allow more rapid measurements. (ie. just point and shoot).

Any pro and cons for either?

Thanks.

[electr_peter]

Few aspects to this issue: accuracy, measurement speed, accessibility to tight places.

Thermal IR guns excel at measurement speed, but can be too big for some cases. Also, they take average temperature of significant size spot (not a single point).
K thermocouples are good for direct contact and tight places. But these require some time to warm up.
Thermo-camera is the best you can get, but it costs significant money.

Accuracy is only ~1-5% for such methods, but it should be good enough.

Generally you would like to quickly check temperature of whole site and identify too hot or cold places. Then you can use thermocouple for more accurate measurement.

[TimNJ]

Few aspects to this issue: accuracy, measurement speed, accessibility to tight places.

Thermal IR guns excel at measurement speed, but can be too big for some cases. Also, they take average temperature of significant size spot (not a single point).
K thermocouples are good for direct contact and tight places. But these require some time to warm up.
Thermo-camera is the best you can get, but it costs significant money.

Accuracy is only ~1-5% for such methods, but it should be good enough.

Generally you would like to quickly check temperature of whole site and identify too hot or cold places. Then you can use thermocouple for more accurate measurement.

Thanks Peter. I suppose an IR gun would be more applicable for electronics troubleshooting because of the speed of measurement. I can't think of too many (any?) scenarios where I would need measurement accuracy of <1%. Though some thermocouple thermometers do have cool features like datalogging and multiple channels.

[Yago]

IR is non contact, so safer with high Volts etc.

The parallax error (between the sensor "spot and target laser)is a git as the spot size is smaller than the error at close range.
This can make it difficult to be sure you are measuring what you want.
Often involves a bit of "zoning" in on the area and watching the temp rise and fall.
That's with my el' cheapo Ebay special.

[electr_peter]

Thanks Peter. I suppose an IR gun would be more applicable for electronics troubleshooting because of the speed of measurement. I can't think of too many (any?) scenarios where I would need measurement accuracy of <1%. Though some thermocouple thermometers do have cool features like datalogging and multiple channels.

I would think similarly.

IR gun is much quicker method than thermocouple. You also can use your fingers for rough temperature measurement depending on situation if you are careful (do not touch at first, just feel at small distance; then touch and release quickly; if not burned by know, touch for longer to get temperature estimate). Fingers are cheaper and do not require batteries last time I checked.

Be careful that IR rays reflect from mirror like surface. Temp of shinyobjects are not easured accurately.

Although it is best to have IR gun and thermocouple. Thermocouple is very useful when you already know that some part is hot and the question is: "Is max allowed temperature reached?" Then you attach thermocouple and check if temp is not too high. From this you can estimate internal die temp, heatsink thermal conductance, etc.

As for accuracy, I really do not care that much because I rarely do need an accurate absolute measurement. Accuracy is limited with cheap methods anyway.
If you need to monitor some equipment temperature on the case and have low expectation of accuracy, then stickers/thermometers with changing colors can be used.

[3roomlab]

for contact temperature measurements ... if you are holding it by hand ... bits of jerking can change the measured value.
if the contact is not good, i find a small squirt of oil can help make a good contact
worse case scenario, temporary glue the tip to the target

for non contact measurement. there is the emisitivity error (black surface / shiny surface ... http://en.wikipedia.org/wiki/Emissivity), the measured spot and pointed spot error (laser points a spot, but measured REGION is below it, not  exactly spot on), etc error margins to consider ... in the same way, for point and shoot "gun" ... if your hand moves ... you know what happens ... http://support.fluke.com/raytek-sales/Download/Asset/IR_THEORY_55514_ENG_REVB_LR.PDF

if you follow the FLIR thread, if the discount season is still valid, it could be a good buy for you.

i find that having a spot K couple measure and a FLIR a very nice combo. for constant measuring and feedback, i find that a hacked computer temp/fan module is excellent, it has 4 temp probes and 4 individual temp readouts ... http://i00.i.aliimg.com/photo/v2/1484410122/PC_12v_cpu_4_Channel_Fan_Controllers.jpg

[TimNJ]

Thanks all. I'm giving the Raytek primer a read. I took a material physics class and it (sort of) makes sense. I understand that many IR thermometers have emissivity adjustment, if you have a table of emissivities on hand. You can also put a piece of black tape on the items you think you will be measuring. All these things take more time, of course. Again, I figure if you're using an IR thermometer, you're using it for general purpose ballpark measurements. But then again, if all you're doing is making ballpark measurements, might as well just use the back of your hand and then a contact thermocouple to make accurate measurements. I'm still a bit split on the usefulness of an IR.

[DanielS]

Personally, if I was buying something to do heat-related troubleshooting, I would give thermal imaging cameras a second look.

I have two semi-working PC power supplies in my spare parts bin. Their outputs look fine but I have a hard time to get them to turn on until they have had time to warm up. I have yet to identify which components need to warm up for them to work normally. I have tried reflowing all the solder joints suspecting a cold or broken joint but that had no effect. If I had a thermal imaging camera, I could record a video along with audio to see which areas seem to warm up in sync with the changing pitch coming off the PSU during that unreliable operation warm-up time.

[nctnico]

@Daniels: those kind of problems are usually found by using freeze spray. Heat the whole thing a little bit with a heat gun and cool one component down with a small blast of freeze spray.

[Wim_L]

Thermal IR cameras are fantastic for locating likely trouble spots. You can quickly scan a large area with them, see hot spots, and usually get enough detail that you more or less know what you're looking at. Other methods take far more time and/or are not capable of measuring small components (most IR spot thermometers have a spot that is enormous, compared to the size of small electronic components).

For precise measurements, though, contact probes are more trustworthy, especially if you're going to be measuring on metal, provided you can hold or mount the probe to make good contact with the component. Less so on plastic, where it might take a while before the plastic heats up the probe enough. Low thermal mass probes are preferred for that reason.

Emissivity correction for extra precision... Usually, I'd say, don't bother, just add patches of high emissivity black tape to any shiny surfaces (for really hot things, heat-resistant matte black paint). There are two reasons for this.
- Emissivity tables are available, but they're rough guidelines only. The emissivity will not only depend on what kind of material it is, but also its surface state (oxidation) and roughness. This isn't always easy to estimate.
- Emissivity is coupled with reflectance. The lower the emissivity, the higher the reflectance, so the part of the heat the camera sees, that is not emitted, is reflected heat from the environment. This means emissivity correction requires two parameters: the emissivity of the material (which you don't know exactly), and the temperature of the reflected environment. That's fine if you are in an environment with a perfectly homogeneous temperature everywhere. Not so good when the environment has all kinds of hot spots nearby, e.g. light fixtures, other heat sinks and equipment, humans,...

[DanielS]

@Daniels: those kind of problems are usually found by using freeze spray. Heat the whole thing a little bit with a heat gun and cool one component down with a small blast of freeze spray.

That only works if you can safely heat and cool only one (or close to) component at a time. I would prefer not sticking my hands or other equipment near live metal parts and large caps charged at 200-500V if I can avoid it and a thermal camera enables exactly that: take a look from a relatively safe distance to identify the most likely trouble area and possibly unexpected hazards before going hands-on.