I am trying to measure temperature of a solar panel using TMP006EVM. This testing is done on a roof where panel is planted (air flow is prime concern), mostly during the time when sun is high.
Since the environment conditions are varying continuously how effectively this device can measure temperatures for a solar panel?
In our measurements there is a difference in object temperature, it is fluctuating randomly for both Local and Object Temperature.
Is this device suitable for measuring temperature of a solar panel?
What is the role of calibration factor in the TMP006 EVM GUI software?
Do we need to change calibration factor while measuring the temperature of a solar panel?? (right now all the settings i.e calibration factor, conversion rate are set according to the data specified in the data sheet provided by TI)
What kind of material is the solar panel made of? I imagine it's some kind of shiny metallic material, which could be problematic for an accurate measurement with the TMP006. The reason for this is that the emissivity of such materials is very low, so it is not an effective radiator of the infrared energy that the TMP006 "sees" when making a temperature measurement. See the TMP006 User's Guide section 3.1 on emissivity: http://www.ti.com/lit/ug/sbou107/sbou107.pdf. If it's possible in your application, an easy way to improve the emissivity of a material is to paint a small area of it (the area within the TMP006 field of view) with dull black paint.
What kind of variations are you seeing on the local temperature? If the conditions are windy on the roof where you're measuring the panels, then these fluctuations might be just due to the cooling effect of the wind. If the local temperature of the TMP006 changes quickly enough, then this could also be causing errors in the object temperature reading.
The calibration factor S0 is a gain error term which can be adjusted to account for various signal losses in a system. It is unique to each system, so to determine the appropriate value of S0 in your application you should perform a two-point calibration at the minimum and maximum object temperature points. This is also described in the TMP006 User's Guide in section 6.
Ian WilliamsLinear Applications EngineerPrecision Linear and Sensing Products
Thanks for your quick response.
The solar panel is made of monocrystalline silicon cells. The panel is protected by a layer of white tempered glass and EVA resin with a weather proof film.
We can't paint a small area with dull black paint. Also I am not clear about the S0
Further I am measuring the solar panel temperature with TMP006 EVM and one more temperature sensor as a reference.
The differences are huge as the other device measurements are almost stable whereas the measurements of TMP006 is very fluctuating ( you can see in the data attached)
I have taken temperature of solar panel with TMP006 and a reference IR sensor with emissivity of 0.96
Also if you can help me Ian in determining what is the distance beyond which TMP006 do not give proper results or what is the distance at which i can get maximum accurate temperature of solar panel. As of now I am placing the sensor as close as it can be to the solar panel, is it making any difference while measuring the temperature. Width, height & thickness of solar panel is 1318, 994 and 46 mm.
All the data are measured with calibration factor of 6.4E-14.
I am not able to find what I am lacking, hope you can help me as soon as possible.
If your solar panels are behind a layer of glass, then the glass temperature is all the TMP006 will be able to detect for two reasons.
I can not comment on the EVA resin without more information, such as a link to the manufacturer's data sheet.
Another potential issue with the measurement is that shiny surfaces in general will cause IR energy to "scatter" that can add significant fluctuation to a measurement with the TMP006. Your attachment did not come through successfully so I am not currently able to see the data you mentioned in your most recent post. What type of IR sensor are you using as your reference? Is it some kind of handheld IR gun-type sensor?
As for the S0, per my previous message it is simply a gain term which can be increased or decreased depending on the amount of signal losses in your system. The default value of 6.4E-14 is a good general-usage value, but it's highly possible that it would need to be calibrated for your system. This is done by performing a two-point calibration as described in the TMP006 User's Guide.
The optimum distance is ≤ (1/2)*radius of a circular area of the object to be measured. For example, if you are interested in measuring a circular area of your solar panel which is 2 cm in radius, then the optimum distance of the TMP006 is 1cm away or less. There is no issue with placing the TMP006 closer. I've attached a screenshot from the TMP006 User's Guide which shows this.
Yes, the reference temperature sensor is a handheld gun device. The data i have attached it again if still you are not able to see then i can mail it to you. Please give me your email address.
One thing which I am not getting Ian, is I am measuring temperature of solar panel with TMP006 and also with this hand held gun like device. If glass is one of the reason for false readings then why it is not affecting the measurements taken with the reference device.
For the solar panel datasheet you can go through this website:
SharpNU S0E3ENU 180E1180WpNU S5E3ENU 185E1185WpEN.pdf
[View the data from the E2E post not your email]
Thanks & Regards,
Sorry for the unclear statement. I did not mean to imply that the glass would cause these fluctuations in the reading that you are seeing; my point is that it will be difficult for you to measure the solar panel behind the glass using an IR sensor. With both the handheld gun IR sensor and the TMP006EVM you are probably measuring the temperature of the glass. However, this is probably very close to the temperature of the solar panel itself so I don't see this as an issue.
Looking at the data you've attached I believe I've found the problem. There seems to be a bug in the software which is occasionally causing an erroneous value of 0 to be reported as the sensor voltage. This causes the object temperature equations to have huge errors as you have seen. As a quick check, I threw out the rows which had the bad sensor voltage values and the behavior was significantly improved.
Are you using the most recent version of the TMP006EVM software? Please compare the GUI of your software to the screenshot below. If they do not match, then you need to update your software to the most recent version. You can download the latest software here: http://www.ti.com/litv/zip/sboc408a
If you are already running the latest software, then I will attach a new evaluation version of the software which does not have this bug.
I also recommend that you enable the "Transient Correction" feature. Since you are taking measurements outside, this feature will help correct for errors caused by quick changes in local temperature (such as wind).
Thanks a lot. You were right I was using the older version of TMP006. Now results with this new version of TMP006 which you had mentioned are pretty convincing then it were earlier.
I am sending you two data of temperature measurements which I have taken with TMP006 sensor and the other reference sensor. These data were taken in two conditions. One was in a room and other of the solar panel. The temperatures measurements of reference and TMP006 sensors taken in the room are almost same which is a good news. But thing which makes me ponder upon is the temperatures measurements taken of solar panel. The reference sensor has almost constant readings whereas the readings of TMP006 sensor are not constant at all. THE data of TMP006 matches with reference but for some time intervals only rest of the time intervals it overshoots or undershoots abruptly.
Can you help me in determining the reason behind this abrupt nature as well as the solution.
Thanks for your assistance and guidance.
Looking forward to your reply,
Was the transient correction feature enabled when you took this data? It should help suppress these sudden over/undershoots. You can also slow down the conversion rate to 0.5 conversions/second or less in order to average more samples and flatten the response.
You can also shield the TMP006EVM from the effects of wind and other temperature transients by placing it inside of a plastic enclosure. We've actually found that a standard milk jug is nearly transparent to the IR energy visible to the TMP006 and allows for good measurement while also protecting from external temperature transients.
Yes the transient correction feature was enabled. Putting the sensor in plastic and then taking the readings helped a lot in getting stable and correct readings.
Thanks for your help
I just wondered if Ian could elaborate on the transparent plastic part of this post. We are looking to put a tmp006 inside a plastic case (to be attached to the body) to measure body temp, I'm struggling to find any specs on the type of plastic we would need to use in order to get the best results i.e. as transparent to IR as possible.
There are two threads on the forum that discuss suitable cover materials for the TMP006:
All content and materials on this site are provided "as is". TI and its respective suppliers and providers of content make no representations about the suitability of these materials for any purpose and disclaim all warranties and conditions with regard to these materials, including but not limited to all implied warranties and conditions of merchantability, fitness for a particular purpose, title and non-infringement of any third party intellectual property right. TI and its respective suppliers and providers of content make no representations about the suitability of these materials for any purpose and disclaim all warranties and conditions with respect to these materials. No license, either express or implied, by estoppel or otherwise, is granted by TI. Use of the information on this site may require a license from a third party, or a license from TI.
TI is a global semiconductor design and manufacturing company. Innovate with 100,000+ analog ICs andembedded processors, along with software, tools and the industry’s largest sales/support staff.