Continue to Site

# Has anyone used this?

Status
Not open for further replies.

#### brodin

##### New Member
https://www.elfa.se/elfa/produkter/en/4898.htm

I have bought one of them(9901). But i don't know how to use it =)

It only has 3 pins: GND, DET+ and DET-

I find no info in the datasheet what the pins are for, and how the output signal works.
Is there a variable resistance between DET+ and DET-? But how is the sensor powered then?

I find no info about voltages or max.current in the datasheet either.
Du you have any ideas?

I am going to use it to meassure the temp on the rear-tyre on my motorcycle when driving.

Looks like an interesting item. I did find some info on the Smartec website - that tells a little bit more on how it works however not enough for me to understand how to use it. I like the tire sensor application.

Check the datasheet for information..... I have never used it but it seems a pretty cool device 8)

DataSheet Says said:
The relation between the resistance value and the temperature is given below:
R(T)=R0*(1+A*T+B*T2+C*T4+D*T6) tolerance class B
R0= resistance @ 0°C = 876.20 W
T = Temperature in °C
A = 5.485 * 10-3
B = 6.650 * 10-6
C = 2.805 * 10-11
D = 2.000 * 10-17

Datasheet
Manufacturer: Smartec

Ivancho

yeah i know. It is a really intresting device to a low price. But it would be good if i know how to use it =)

ivancho said:
Check the datasheet for information..... I have never used it but it seems a pretty cool device 8)

DataSheet Says said:
The relation between the resistance value and the temperature is given below:
R(T)=R0*(1+A*T+B*T2+C*T4+D*T6) tolerance class B
R0= resistance @ 0°C = 876.20 W
T = Temperature in °C
A = 5.485 * 10-3
B = 6.650 * 10-6
C = 2.805 * 10-11
D = 2.000 * 10-17

Datasheet
Manufacturer: Smartec

Ivancho

I have read the datasheet. But i find no info about how to connect it. There is no pin explanation, is it? What is DET+ and DET-?

I see what you mean.... I didn't read well the data sheet.... :shock:

Well, as I understand from the data sheet.... the sensor generates a voltage that is proportional to the amount of heat..... and a High sensitivite sensor at 100 V/W

So what I get is if you measure the temperature of a 60W bulb would create 100/60 V = 1.6V..... and I think that you are going to get that voltage between DET+ and DET- ....... But how hot is 60W?

Question

How is the measurement Watts related to temperature in terms of energy?

A Watt is a unit of power, defined as the rate of energy transferred [per second]. The energy is usually defined as Joules; therefore, one Joule per second is one Watt.
When measuring heating effects, a unit of thermal energy known as a calorie is used. One calorie is 4.184 Joules.

Materials have a property called thermal capacity or specific heat. This is a measure of how many calories are needed to raise 1.0 gram of the material one degree Centigrade. The thermal capacity of water at 15 deg. C is 1.0 calorie. That is 4.184 Joules or 4.184 Watt-seconds. The entire energy could be transferred to the water in one millisecond at a rate of 4184 watts to produce the same temperature rise of one deg. C.

The thermal capacity of materials changes slightly with temperature primarily due to changes in density, and very dramatically at phase transitions, such as ice melting and water boiling

So that is were I get lost :?

So I read some other stuff and I got this:

A radiation thermometer exploits the principle that all objects emit infrared radiation (radiant heat) at a rate directly related to the temperature of the object†. The rate of emission increases
very rapidly as the temperature increases. At room temperature every object emits energy at a rate of about 470 watts per square metre, while a hot stove element at 500°C emits about 20,000 watts per square metre. Optical components inside a radiation thermometer (similar to
those found in an everyday camera) are used to focus the infrared radiation onto a solid-state detector where it is converted into an electrical signal. This signal is then manipulated electronically and finally read out as a temperature on a digital display.

There is also emissivity, which is the ability of a surface to give off heat by radiation. The scale is 1.0 (perfect or complete emission of radiant energy) to 0.0. Most non-metallic materials have an emissivity between 0.95 and 0.8, for example: water (0.95), rubber (0.86 to 0.95), smooth glass (0.93) and paper (0.92 to 0.94). Metallic materials have a lower emissivity range, for example: aluminum (rough 0.05, oxidized 0.1 to 0.2) and steel (polished 0.13 to 0.4, oxidized 0.8 to 0.95).

But this is how far my curiosity took me... There is a law called Planck's Law, which somehow relates wavelength (in this case the one for infrared) to temperature in Kelvin, but that was more physics that I could take for a day :lol:

Maybe I am just missing the whole thing, and I am overthinking it... anyone?

Ivancho

I called to the company that creates them. He said that there is a low lovel ouput on the DET- and DET+ on a few mV. So i have to amplifie the signal to get 0-5 V

But how does that mV relate to the temperature? According to what I read is proportional... but what proportion... do you have to figuere t out.... by trial an error?

the formula is fourth power function of the temperature.
reversing it is not going to be simple and if you have to use
it in microcontroller, i'd probable have to say - forget it.
but...
it all depends on what accuracy you are looking for and how fast
you want to sample it. temperature is generaly slow function anyway.
it is not that big deal to 'guess' using interpolation till you get
close enough in probably only couple of iterations. After doing it once,
you should be able to track it with one or two iterations at most...
(just an idea)

The accurancy is not very important. +- 3 degrees would be no problem.

It's going to meassure the temperature of a tyre, which can be 80 degrees (celcius) during a race. I just want to get a aproximate value so i will amplify the output signal so it fits to about 0-100 degrees will be 0-5 volt.

Then i will plot out different temperatures with help of a thermometer, and try to make a function that is aproximate.

I beleive this type of sensor is called a Thermopile. It uses a stack of mini thermocouples that are actually heated by incomming IR. It generates a voltage just like a thermocouple. The other two leads are for a thermistor that gives the zero point for the thermocouple.

Melexis makes a similar device that is available on digikey. They also have a Prebuilt board with sensor amps and all that gives a nice digital output.

The one at digikey looks great, but it is much more expensive! It costs about $65 and the one i have costs$12.

Status
Not open for further replies.