Help identifying a temp sensor leads

I have a Boarduino (it isn't serious, my doctor said it should get better on its own).

I'm trying to interface an infrared temperature sensor I'm salvaging from an inexpensive, Cen-Tech, hand-held remote thermometer. The original unit is inexpensive and sweetly functional. However, I need to output the sensor it has through a TTL/USB cable to a laptop. I need a graphic bar display of the temperature data.

After opening the unit I note a TO5 case device with four leads, one is attached to the case. Looking up from the bottom of the device (where the leads exit) one sees the familiar key tab in order to identify leads. I do not have the manufacturer of the part, much less the data sheet.

If I apply power to the unit and use a digital multimeter I get these voltages (without turning the unit on: from clockwise, starting closest to the tab, 1.9 v., 2.1 v., 1.94 v., and 2.36 v. respectively. I thought I'd get a null, ground voltage from the case lead, which is the fourth clockwise.

I found what I thought was the same device on a website for a Netherlands firm, but now I don't think so. I do know this, however: one lead controls ambient temperature of the device in order to cut down on spurious readings.

I am hoping someone can help me identify which lead is which so I can safely hook it to my Boarduino and get some data piping into my laptop. Any kind of notions or links would be appreciated.

I would also appreciate hints about an adaptable app for windows or linux (Kubuntu's Hardy Heron) to use as a GUI dynamic graphing display.



Note, I'm sleepy right now, so my advice with a dash of salt.

Try turning it on, and measuring the voltages while you point it at different temp areas. Specifically watch the last 2 because they seem to be irregular numbers.

If the voltage of the lead connected to ground is not zero, it makes me think you are not measuring your voltages with respect to ground. Are you sure the black lead of your multimeter was on ground while you were making your measurement? What did you choose as your reference point?

  • Ben

First, thanks for the replies. I'll go back and double-check my ground reference point, I think I was careful about that before, but maybe not.

It's hard to get the unit turned on properly b/c it's out of the case with the control buttons and the LCD readout is disconnected too. But I bought a backup unit to the one I'm cannibalizing and will try to utilize it for comparisons without, hopefully, ruining it! I need to get some of those micro clip probes so I can sample the leads w/o risk of shorts, etc. Teeny contact points.

It sure would be neat if I could get the data sheet or schematic from CenTech, but fat chance these days. If I just knew who supplied their sensors that would probably be all I need.

Any hints about an application I could adapt to run on the laptop to display serial data like I'm after? I could learn how to write my own, I guess, but it would take a lot of time, I think. I'd prefer a Linux app, especially a Java based one.

I'll get back here once I have more data from the unit--hopefully.


Thanks again, all


For the computer side of the system developing a custom interface with Processing is a good bet It's the program that the Arudino IDE was based on, so it's a familiar UI and a very similar syntax (C-ish), it also produces Java Applets. I (and probably quite a few other people) am currently using Processing to display serial data from an Arduino, it's quite simple, but occasionally requires a few work-arounds.

This sounds like a thermopile sensor. They range from the raw sensor (outputs a voltage in the range if microvolts indicating the temperature differential between the sensor and the target) to sensor plus on board temp sensor (so that the difference can be calculated) to highly integrated devices with onboard sensors, amplifiers, and I2C interfaces - I have seen all three types in the package you describe.

See if you can find a part number on the can, and start googling. Hopefully you haven't fried it by applying voltage to an input.


Hey, 73 from KE7SLX

Thanks for your reply. I couldn't decide if this was a passive/resistive device or one that generated a voltage; I now am fairly certain it's the latter.

But the TO5 is encased in a special holder that may also serve to stabilize ambient temp fluctuations, and also holds a funnel-shaped director. The cone itself is a little large than a pencil eraser. The holder overall looks vaguely like an RS235 socket, in that it's elongate, rectangular and has hold-down screw holes on the tips of the wings. The TO5 is tiny inside. It won't be easy to extricate it to see if there are any identification marks on the case, and now I may not have to.

I have not unsoldered it from the original unit yet. I did find a data sheet pdf online for a device that, if it isn't the same one, is probably (hopefully) close enough. At any rate, I cut a hole in the original plastic case so I could access the pins with my MM with power applied. I get more realistic voltages that pretty much match those in the PDF document, when using the case as ground.

Pin one shows + 0.739 v; pin 2 continually bounces around with several readings ranging between 0.23 v max, and 0. I figure that's the data (duh), but it does not change in any detectable way when I am it at cool or hot objects. Pin 3 shows + 0.69 v., consistently slightly lower than pin one. Pin 4 is electrically identical to the case; I used it as the ground in the above measurements. Except for pin 2, the other voltages seem constant on pins 1 and 3.

The PDF document has been extremely helpful because it also gives two suggested application circuits which I am really expecting will allow me to reconstruct a usable interface for my Boarduino, or for my XBee radios. I just need an op amp and a handful of resistors and capacitors.

Unfortunately, the document does not give the values of the resistors and capacitors. I think I'll regard the caps as typical bypass units, give the way they are used in the circuit, and start with 0.01 mfd on the breadboard. I'll use 10k resistors in what looks like voltage dividers connecting to the sensor itself. I'll just keep subbing in lower values until I get some readings (or smoke) I guess. I'm trying to figure out how to display the circuit here.

Meanwhile, If all this works out, I'll be more than half way home on my project.

Then I'll need a computer application to manipulate and display the data on a large screen.



I’m helping out another guy with a thermopile (aka pyroelectric or pyro) that’s in a TO5. Dexter is the manufacturer. It has 4 leads, two of which are NC. (an internal thermistor for the reference temp is an option that may be connected to the other two pins.) The development board sold by the manufacturer includes a tiny (SOT23) temp sensor (LM20? can’t recall for sure right now, something from National) that is mounted underneath the TO5 can with a helping of thermal grease in between. The eval board also has a 1000 gain amplifier on board to increase the signal to a more friendly level.

Melexis makes a really spiffy one that has the pyro, onboard temp sensor, amplifier, ADCs, and an I2C interface. Unfortunately my guy has some very specific requirements on wavelength sensitivity so we had to go with the raw sensor from Dexter.


Hello again,

I feel confident that my UV temp sensor is sufficiently similar, if not identical to the Dexter unit, and I plan to proceed as if it were (famous last words). I extracted it from the commercial unit and find no markings on the tiny TO-5 case.

BTW, I checked out the Dexter subunit and it is sweet, and provides as much as I'd need, except it's a bit pricey at I think $69 per. I still may swing for it though, but right now, I like learning as I go so I'll first try to build my own.

I have a 741 op amp to use to amp up the output before interfacing it with my Boarduino, or my XBee radio, but "Servo" mag recommends using an LM324 op amp because it only needs one voltage for power (e.g. not both neg and pos). I will probably try to get one later but I'm so anxious to start wiring I can't wait.

If I may, I have another question for you, or someone else who reads this: How would you interface the thermopile to the op amp input safely, given no other data to go by? I think I don't have much to worry about, but I'm not sure and don't want to fry my only available sensor. Or can you point me to some good tutorial for an Op Amp user newbie to read up on?

Thanks Again


I'm fairly weak on op-amps myself, but I think that you won't damage the pyro by attaching it to an op-amp input. The op-amp should be fairly high impedance, so there's much less of a chance of damaging anything. Pyros behave like thermocouples, and some (thermopiles) are effectively a bunch of thermocouples in series.

The 741 should be fine, it's just less convenient than a single supply model such as the 324. Since a pyro will give a temperature relative to its own temperature, the negative voltage capability of the amplifier can be a good thing. In the long run you'll probably want to build in an appropriate offset so that the voltage is positive.

If your to-5 is tiny we may be talking about different packages - my specs say the to-5 is about 10mm in diameter...


FYI, I found an excellent tutorial site for Op Amps at:

Since I'll be applying some part of 5 volts, after it goes through a voltage divider, across the thermopile, I wasn't sure how much current it could tolerate without burning up. But maybe the internal resistance of the thermopile is high enough to self-limit this. At any rate, I'm not sure what combination of values to start with for the formula: Vout = Vin [1 + (Ra/Rb) ]

Is there any way to post a schematic in this forum? It would be easier to point to the places I could use advice about component values. The "insert image" button at the top of the post menu does not seem to work (meaning, probably, I don't yet know how to do it!).

Please try this: