How to add temperature measurement to Uni-t 61E?

[timfayz]

Hello all! I'm very new to electronics and I'm trying to get started with it. Recently, I've bought Uni-t 61E and realize that temperature measurement would be very useful in my little reasearches! That is why I'm a little bit disapointed
But today an idea has came to my mind. I don't know how thermocouple exactly works exactly, but I assumed: maybe it works like the thermistor?!

Maybe I can connect them to my multimeter using "resistor measure" mode and based on resistence it has calculate actual temperature? What do you think?

[3db]

It doesn't work like a thermistor.
It uses something called Seebeck effect.
The junction of two dissimilar  metals will produce a voltage proportional to temperature.
The type normally supplied with multimeters is type K.
Hope this helps.

3DB

[MrSlack]

I bought a UT61C as well as a UT61E. Problem solved and you can never have enough meters.

[tszaboo]

Thermocouple outputs voltage. Use a standard chart to determine the temperature. Usually the voltage is too low to be able to read it accurately, we are talking about less than 10mV.
Fluke has a thermocouple to millivolt "converter". It took the thermocouple input, linearised a and scaled it, so the output was 1mV=1K. It is called:
Fluke 80TK Thermocouple Module
I see it sold in the US used for 20 bucks. Probably new it is more expensive than your meter. In that case just get a second meter with Type K input. Other manufacturers might have something similar.

[timfayz]

Ok got it, big thank you for clarification! Anyway it's sad..

[timfayz]

I've read some information on this topic and there are two suggestions how to measure temperature: detect mV using thermocouple or Ohms using RTD if the range between -200 to 500 °C (which is very suitable to me).
For example, reading the datasheet of PT100 RTD I can see that if I plug it to 61E and measure resistance, using this table I could certainly calculate more or less an actual temperature (as I see 61E has good precision on measuring resistance).
The second example is ITS-90 type k thermocouple which
requires measure mV. 61E has resolution of 0.01mV (1% and +/-10) and even if we assume that resolution is 0.1mV in worst case that should be enought to detect temperature using this table with accuracy near to 3°C. Of course this is my theoretical consideration and of course I'm wrong in good extend , but I absolutely don't need hight precision (even +/-10C is acceptiable).
May be at least the use of RTD can be solution?

[daybyter]

You could just by this as a 2nd dmm:

http://www.dx.com/p/uni-t-ut136c-2-lcd-digital-multimeter-red-dark-grey-1-x-9v-battery-168367#.Vw1rv3q-eL4

But I didn't want to carry another dmm with me, so I bought a temperatur sensor like this:

http://www.reichelt.de/PLAT-PT100/3/index.html?&ACTION=3&LA=446&ARTICLE=151259&artnr=PLAT+PT100&SEARCH=temperatursensor

, added a small handle, 2 cables with banana plugs and printed my a small sheet with resistance values for 0 - 100 dg celsius.

[Gyro]

Or just pick up a standalone thermocouple thermometer, they're ridiculously cheap these days. eg:

http://www.ebay.co.uk/itm/TM-902C-Type-K-Digital-LCD-Thermometer-Thermodetector-Meter-Thermocouple-Probe-/251681029162?hash=item3a995bbc2a:g:xoUAAOSw6EhUQI1v

Edit: Note that some are 1'C resolution and others 0.1'C. Check with sellers if necessary.

[tablatronix]

I was watching this guy repair one and noticed one of his symptoms was it was showing temperature ( his functions were all messed up due to bad matrix logic ) and he worked out the matrix not written in the schematic for temperature, not sure what other supporting components would be needed but it is possible the software is there.

https://youtu.be/dKdTcXE-Kc4?t=4m22s

[Arhammon]

The simplest way to add  temperature measurement to any meter - is  LM35 IC + 9v Battery + wire

[knks]

You could just by this as a 2nd dmm:

http://www.dx.com/p/uni-t-ut136c-2-lcd-digital-multimeter-red-dark-grey-1-x-9v-battery-168367#.Vw1rv3q-eL4

But I didn't want to carry another dmm with me, so I bought a temperatur sensor like this:

http://www.reichelt.de/PLAT-PT100/3/index.html?&ACTION=3&LA=446&ARTICLE=151259&artnr=PLAT+PT100&SEARCH=temperatursensor

, added a small handle, 2 cables with banana plugs and printed my a small sheet with resistance values for 0 - 100 dg celsius.

AFAIK the cheapest multimeter with temperature is the DT-838 - around $6 on 'bay. And yes I know it's shit and not safe.

[MosherIV]

Hi

I have just bought mine, still waiting for it to arrive. I did lots of investigations about this meter.

The chipset does support temperature measurement. The screen even has the symbols for it.

Sadly, adding it on to the meter is so much work that it is easier to buy another meter.

FYI if you are interested in the technical how to do it :
The chipset (Cyrustek ES51922) selects function by a five bit binary pattern on inputs SLADC, FC1, FC2, FC3 and FC4
00100 is needed to select the temperature function.

Problem 1 - the rotary selector of the UT61e does not provide this pattern. Changing the PCB pattern to give this pattern is impossible (without a complete PCB re-design)

Problem 2 - additional support circuitry. Even though the chipset does support temperature reading, more circuitry is needed. The circuit is not that hard (for those that can figure these things out). I have a UT70b with the schematics, quite difficult to follow but it is essentially a bridge which allows calibration and then a cold junction temperature circuit/amplifier which just produces the 1mV/C

Just thought some of you might be interrested.

[picitup]

I just took Arhammon's advice and built an external circuit using the LM35DT and a PP3 9v battery.   The DT is a TO-220 package so I can clamp it to items, such as my radiator etc.  The only slight down side, is it's reaction to temperature changes is slow, probably due to the mass of the device.  By contrast, my K-type bead probes is lightning quick.

I did consider using a PT100 or PT1000 as these are great accuracy and a linear output, but there were two problems for me;  The output would not be human readable at  0.384 ohms per degree C and secondly my 22000 count UT61E is only 22000 count on voltage so I couldn't make an accurate reading.

The LM35 supplies 10mV/ degree C and requires a negative power supply for negative temperatures, which almost put me off, but a couple of diodes in series with the ground pin sorts that out so not too painful.  Attached is a pic of the diagram from the data sheet which is here:

http://www.ti.com/lit/ds/symlink/lm35.pdf

At 10mv per degree C, all you need to do is multiply the display output by 100.  Even someone with my maths ability can do this.  If you look at the pic of the DMM, the reading is 0.2077 or 20.77 degrees C.

Accuracy and resolution
The LM35 has a better accuracy than the TMP3x range and is 0.25 degree at 25 degree C and a maximum of 1 degree C across the whole range of -50 to 150 C which is good enough for me.
So what's the point of having a resolution of 0.01 C when the accuracy is only 1 C over the whole range?  We I found it very useful as a check to see if something is heating or cooling.  On our radiator, when it cools, the temperature drops by .01 C every second so you can tell where the cycle is.

Data logging
I took a look at UltaDMM and on my Windows 7 64 bit laptop, it comes up with exceptions, particularly when you choose Gauges/Add.  However, if you just choose Continue it doesn't seem to mind and still works OK.  This is a very nice program, but I don't think it's supported any more which is a shame.

http://www.ultradmm.com/

I also played with the UNI-T data logger and it's pretty basic, but probably good enough for my use.  A slight issue is it only supports COM1-4 and my 4 port USB/RS232 device installed itself as COM6,7,8,9 but a bit of uninstalling and reinstalling offered me COM1 to try.

I did have some connection problems and suspected the Quad RS232 device and opened it up and it was rusty inside so I gave it a clean with lighter fluid and now it seems to work fine.

While I was playing around fault finding, I sniffed the serial port with my favourite program TeraTerm but got no output.  The settings were correct at 19200, 7,O,1 but nothing.  Then I found a great page on reverse engineering the protocol here:

http://www.starlino.com/uni-t-ut61e-multimiter-serial-protocol-reverse-engineering.html

Here is says you must set DTR high and they use a serial program called Termite where you can click the DTR led and toggle it.  Termite is here:

http://www.compuphase.com/software_termite.htm

There is no installation program and you also need to download the plugin "Status LEDs filter, including source code" and save it in the same folder as Termite.  Then go to Settings and enable the plugin.  Next connect to the UT61 at the above baud rate etc and click the DTR led and you should see the meter communicating.

Piccies

Once built, I did a few tests and have included the graphs below:

UT61E_20.77DegC shows the meter with a temp of 20.77 C
BoilingPan shows a pan of water heating up on our induction hob.  Notice the wiggle at the start
BoilingPanCycling zooms into the graph above and shows the initial on/off cycling
LighterGasFreeze tests the sensor below 0 by spraying it with lighter gas.  Best put your cigarette out first!  You can see from the graph I had to spray it twice
RadiatorPipe shows the temperature of a radiator flow with the LM35 clamped to the copper flow pipe.  You can see the heating and cooling cycles easily.

The boiling pan test is a method used to calibrate thermometers and I was pleased to see the peak temperature measured was 99.81 or only 0.19 C error at 100 C.

All in all I'm very please with the device and the current drain measured from my UT61E was only 20uA.  I'm guessing this isn't too accurate so as soon as I can find my uCurrent I'll test it properly. The battery is rated at 550mAH so I think the shelf life will kill the battery before it runs out!

That's all for now, I hope you've enjoyed reading this.

Kind Regards

Steve

[3db]

@picitup
Will you be doing the gas plus cigarette test  ?   
3DB

[picitup]

Lol if I do, and survive, I think the graph would be interesting and it probably wouldn't take too long for my eyebrows to grow back :-)

[picitup]

Oops forgot the circuit diagram.  This is the suggested one from the datasheet.

[jeroen79]

It doesn't work like a thermistor.
It uses something called Seebeck effect.
The junction of two dissimilar  metals will produce a voltage proportional to temperature.

Actually, the thermocouple's output is proportional to the difference in temperature between the meter (cold junction) and the end where the two wires are linked.

[3db]

It doesn't work like a thermistor.
It uses something called Seebeck effect.
The junction of two dissimilar  metals will produce a voltage proportional to temperature.

Actually, the thermocouple's output is proportional to the difference in temperature between the meter (cold junction) and the end where the two wires are linked.

My objective was to aid the OP understanding.
If you buy a typical thermocouple the cold junction is integral so I figure my explanation
was sufficient for him gain understanding and also find out more for himself.   

3DB

[chemary]

I'm also interested in adding temperature readings to Uni-t 61E, found this page about an other multimeter using same IC and supporting temperature (and inductance) http://improwis.com/projects/reveng_TD2200/
The correct pattern for FC[1:4] pins can be generated with a microcontroller as other hacks to this multimeter do, Capacitance measurement seems the most easy to alter as it's like temperature measurement plus setting FC3 pin down and one of the yellow/blue keys can be used to switch the function as in Capacitance mode they have no utility.

Additionally something has to be connected to pins [ADP, VA+, VA-] but I don't know exactly what or how need to be connected there, manual states Temperature measurement mode takes input signal from ADP pin and reference voltage from VA+ - VA-.
TD2200 uses an OA OP07 an on it's datasheet there is an application example for using it as High Stability Thermocouple Amplifier, TD2200 also has a trimpot for offset adjustment of the OP07, I suppose the output of the OA goes to ADP pin, but what will I put on VA+ - VA-? maybe ground and VB- as in ADP test example?