Temp Sensors

[itsmoked]

So, what does one use for a clean solution to monitoring air temperature for use in a PLC application?

This is an air conditioning, (that's heating and cooling), system for a local ambient.

This is a 24VDC setup.

Keith Cress
kcress -

http://www.flaminsystems.com

 

[Skogsgurra]

NTC thermistors are dirt cheap, reliable, standard solution and with a few centigrade's tolerance - which can be calibrated out semi-automatically. You can easily pot NTCs yourself to get a custom transducer or buy them from most cow/horse/sheep/pig farm ventilation companies.

Temperature range is easily 0 to 45 centigrades. But can be extended to at least -20 to 75 C. No amplifiers needed. A best-selling application is pool water temperature control with 10 kohm@25C thermistors and a suitable series resistor (22 or 33 kohms) fed from 24 V and taken directy to the AI of a Siemens LOGO! or other low-end PLC.

Gunnar Englund

http://www.gke.org

--------------------------------------
Half full - Half empty? I don't mind. It's what in it that counts.

 

[IRstuff]

Thermistors are reasonably cheap, and possibly a good fit, particularly if your PLC already has ADC. Nonetheless, you'd still need to write some code to convert from resistance (or voltage) into temperature. Not a big deal, but still, something that must be maintained and documented AND CALIBRATED.

An alternative is a single-chip solution:

http://para.maxim-ic.com/en/results.mvp?fam=temp_sens&723=Remote&956=Single

Digital serial output, AND calibrated already. A bit pricier than just a thermistor, but much closer to a PnP solution.

TTFN
faq731-376

 

[itsmoked]

Thanks Gunnar. I see a probe in one greenhouse supplier. Of course they tell you nothing about it.

I will look further.


IR; Yeah, I could go that way but I really wanted a PnP solution that would have the wire and an assembled protected sensor. This is for very low quantity stats that get ordered intermittently. Having to build sensors every time would be a PITA.

I will be using a brick PLC that has 2 analog inputs.

I'm considering using the other analog input to read an insolation sensor so I can add in comfort offsets.

Keith Cress
kcress -

http://www.flaminsystems.com

 

[controlsdude]

what about rtd pt100 type temperature sensor?

most plc manufacturers have direct interfaces for this sensor.

 

[itsmoked]

Since this is just controlling between about 62F and 82F a 100 ohm RTD is not going to be giving much to work with. Perhaps a 1k or 2k version might work? These PLCs don't have RTD mods.

Keith Cress
kcress -

http://www.flaminsystems.com

 

[ozmosis]

I would typically use a resistance based temperature device such as a Ni1000 (

http://w1.siemens.ch/web/bt_ch/Site..._4_Fuehlertabelle_20Ni1000_1000000041142.pdf)

as this is scaled for temperatures in AC systems. Then connect into a simple HVAC controller such as this:

https://www.hqs.sbt.siemens.com/gip...le-and-energy-efficient_A6V10254842_hq-en.pdf

These are Siemens products but most companies in the HVAC world will have something similar.

 

[itsmoked]

Hi oz. You second link is an httpS so it's capoot. First link.. Not working either. Thanks for the effort though.

I've never seen a simple enough 4 stage temp controller on the market.


I'm looking at just using an LM34. It will take in the 24Vdc and give back 10mV/degree F. I don't think there's anything else out there with that much noise margin. And it's fully linearized too!

Keith Cress
kcress -

http://www.flaminsystems.com

 

[catserveng]

Keith,

How about this?

http://www.omega.com/Temperature/pdf/EWS_WALL_MNT_ALL.pdf

24VDC, different TC's available, 4-20mA or 1-5VDC output, in the temp range you described.

Mike L.

 

[itsmoked]

Hey thanks cat.

That would work perfectly in some of these. However most of these locations are in plenums with 20MPH wind and they state "not for use in fast moving air". Not sure why that would matter but..

Keith Cress
kcress -

http://www.flaminsystems.com

 

[LiteYear]

FWIW, both of ozmosis' links worked for me.

 

[itsmoked]

And so there are! Now.

Thanks.

Keith Cress
kcress -

http://www.flaminsystems.com

 

[Cbrunner]

The reason they say not for fast moving air is because of the wet bulb / dry bulb problems with thermometers. Remember hearing about "wind chill" in weather reports? That's a calculated temperature based off of wind speed and relative humidity. Most of the duct sensors out there are calibrated/designed in such a way that wind chill is not a factor.

 

[Skogsgurra]

The wind chill effect does not influence sensors that are not self-heated.

A normal NTC, that is put in still air shows, say, 10 kohms resistance at 25 centigrades room temperature. If air (still 25 centigrades) is blown across it with a fan, it still shows 10 kohms.

There is a big difference if the sensor is self-heated. Then the air velocity influences thermal coupling and the resulting resistance will change with air velocity.

Make a test! It is easily done - just hook an NTC thermistor in front of a stopped fan to an ohmmeter and note the resistance. Then start the fan and notice that there is no change in resistance.

Gunnar Englund

http://www.gke.org

--------------------------------------
Half full - Half empty? I don't mind. It's what in it that counts.