Best cheap simple thermocouples?

[Hepth]

I'm looking to install thermocouples or thsrmometers around my house to monitor temperatures in each room throughout the day. I'll wire these to my pc, probably with a cheap USB DAQ set and record it with labview. I'm wondering if anyone has any experience in fthis. Are there devices that give an output voltage based on the ambient temp but are powered by thedaq card, or are there ones that are battery powered?
Just looking for the cheapest and easiest solution and some ideas and suggestions. Thanks everyone!

 

[berkeman]

I'm looking to install thermocouples or thsrmometers around my house to monitor temperatures in each room throughout the day. I'll wire these to my pc, probably with a cheap USB DAQ set and record it with labview. I'm wondering if anyone has any experience in fthis. Are there devices that give an output voltage based on the ambient temp but are powered by thedaq card, or are there ones that are battery powered?
Just looking for the cheapest and easiest solution and some ideas and suggestions. Thanks everyone!

I've used the Dallas DS1820 "1-wire" powered temperature sensor before (in a TO-92 package). Dallas is now part of Maxim, and the new part is the DS18S20:

http://www.maximintegrated.com/data...tegrated.com/datasheet/index.mvp/id/3021/t/al

I don't think running long wires to thermocouples will be very practical. Thermocouple differential output voltages are low, and you will pick up a lot of common-mode (CM) noise on those wire runs. Better to convert the temperature to digital info at the sensor, IMO.

 

[Hepth]

Hmm with one or two I'd agree but if I want 25-50 to measure heat gradients, doing a A to D conversion means I need a controller at each. So then each adds the cost of two AAA's and some 2 channel analog-digital processor right? I've been out of DAQ for a while, are there cheap solutions for this?

 

[berkeman]

Hmm with one or two I'd agree but if I want 25-50 to measure heat gradients, doing a A to D conversion means I need a controller at each. So then each adds the cost of two AAA's and some 2 channel analog-digital processor right? I've been out of DAQ for a while, are there cheap solutions for this?

It has the ADC inside each unit. You read out the temperature in ASCII (IIRC) from multiple of these things with just one uC somewhere, over the "1-Wire" (actually 1 wire + ground) network. I've only used it as a single device on a PCB read by a uC, but from the info at Maxim, it looks like you should be able to network a bunch of them together:

Each DS18S20 has a unique 64-bit serial code, which allows multiple DS18S20s to function on the same 1-Wire bus. Thus, it is simple to use one microprocessor to control many DS18S20s distributed over a large area. Applications that can benefit from this feature include HVAC environmental controls, temperature monitoring systems inside buildings, equipment, or machinery, and process monitoring and control systems.

 

[AlephZero]

I don't think running long wires to thermocouples will be very practical. Thermocouple differential output voltages are low, and you will pick up a lot of common-mode (CM) noise on those wire runs. Better to convert the temperature to digital info at the sensor, IMO.

Take a step backwards from that position: a thermocouple works by generating a differential voltage between two junctions. So if you plan to measure "room temperature", you need a known reference temperature for the other junction of each thermocouple. That doesn't sound very practical or cost effective in a domestic envuronment.

IMO if would be much simpler to use a sensor chip that produces a voltage that is a function of its temperature (e.g. the voltage across a semiconductor junction), or possibly a thermistor - but semicondiuctor sensors are probably more accurate and easier to use.

 

[dlgoff]

With thermocouples, there's no problem with run lengths from my experience. And the wire isn't that expensive. http://www.omega.com/toc_asp/subsectionSC.asp?subsection=S07&book=vhle&flag=1
You can run bare T/C wire and make-up/solder it's sensing end. I would use type K. Extension wire is cheaper and just a good for your application. K Type Thermocouple Extension Wire

I don't know what you consider cheap but here's what you need IMO.

http://www.omega.com/ppt/pptsc.asp?ref=OM-DAQ-USB-2400&Nav=dasc01

Each OM-DAQ-USB-2401 module is supplied with a free Windows Software Suite that includes the DAQ Central menu driven Windows software.

 

[dlgoff]

... if you plan to measure "room temperature", you need a known reference temperature for the other junction of each thermocouple.

The industrially accepted standard for T2 is 0°C; therefore, most tables and charts make the assumption that T2 is at that level. In industrial instrumentation, the difference between the actual temperature at T2 and 0°C is usually corrected for electronically, within the instrumentation. This emf adjustment is referred to as the cold-junction, or CJ, correction.

Practical Guidelines for Temperature Measurement

 

[AlephZero]

The industrially accepted standard for T2 is 0°C; therefore, most tables and charts make the assumption that T2 is at that level. In industrial instrumentation, the difference between the actual temperature at T2 and 0°C is usually corrected for electronically, within the instrumentation. This emf adjustment is referred to as the cold-junction, or CJ, correction

I think you missed the point I was trying to make. If there is a large temperature difference between the hot and cold junctions, it's quite reasonable to have the cold junction at "room temperature" rather than in a temperature controlled environment (e.g. an ice-and-water bath). You then have to measure the cold junction temperature indepedent of the theromocouple to make the correction in your quote. Of course the T/C instrumentation system may make the measurement for you, and do the corrections aotomatically.

But the OP is talking about measuring room temperature with the thermocouple, so either you do need to keep the "cold" junction at a known temperature, or you need to measure its temperature indepedently - but you might as well use that "independent" measurement to measure the room temperature directly and forget about the thermocouple.

I agree noise pickup in T/C leads is not usually a big deal, because they are very low impedance and common noise is pickup easy to reject.

 

[dlgoff]

But the OP is talking about measuring room temperature with the thermocouple, so either you do need to keep the "cold" junction at a known temperature, or you need to measure its temperature indepedently - but you might as well use that "independent" measurement to measure the room temperature directly and forget about the thermocouple.

But there's probably not going to be that much variation from the T/C room temp and the Data Acquisition Module. Besides, how much accuracy does he need?

According the the "www.omega.com/DAS/pdf/OM-DAQ-USB-2400.pdf" , the Cold Junction Compensation Accuracy is ±1.0°C

 

[jim hardy]

check the AD590 from Omega
http://mdm.kpno.noao.edu/td/PDFs/AnalogDevices/AD590.pdf
can't get any simpler than that.
For better than 1 degree resoution, one would want to check them all at some uniform temperature and get a correction for each.

I once used some Microchip I^2C units - once one gets past the knee on handshaking learning curve they are amazing.
see appnote TB052 at
http://www.microchip.com/stellent/idcplg?IdcService=SS_GET_PAGE&nodeId=1824&appnote=en011963

 

[Devils]

Try Raspberry Pi ($35 computer running full linux with built in support for I2C and many other sensors)

http://www.adafruit.com/blog/2012/11/23/my-kitchen-temperature-using-raspberry_pi-piday-raspberrypi/

A major issue us device drivers. These newer RP linux kernels have support for numerous sensors built in.

And you can program in C or Python & do whatever you want with the results.