Temperature circuit help

Discussion in 'Electronic Design' started by kmillar, Feb 9, 2005.

1. kmillarGuest

I'm looking for some help designing a crcuit to measure temperature in
a central heating system.

Sepcifically, I have a ZWorld OP7200 programmable display which has

would like some help deisgning the external circuitry. What type of
temperature sensors should I use? Temperature Dependant Resistors?
Something else? There will be 2 of them in total, one recording
temperature of flow from the boiler, and one reading temperature of
water returning to the boiler, by being clamped onto the copper pipes,
near the boiler.

Any help would be gratefully accepted, schematics (JPG or Eagle etc)
would be fantastic. It doesn't need to be specific to the OP7200.

2. j.b. millerGuest

When I designed remote energy management systems, I used the National
LM34DZ temperature sensor and an LM358 dual opamp. Used one stage as a
buffer the other a simple x4 amplifier.
My systems also monitored boiler emps as well. Simply use a gearclamp and
lightly hold the LM34 onto the pipe.
It's not real 'rocket science' and you will have to experiment to get the
gain right for you application.Nice thing about the LM34, 70*F = .700 volts
!
Good luck , it is easy though

Jay

3. Tim WescottGuest

Thermistors work well if you can stand accuracies to a few degrees.
With calibration you can do better yet.

I've used the circuit below in the past, and it's worked well. In spite
of appearances it does a good job killing common-mode noise from the
thermistor. It takes the exponential resistance from the thermistor and
shunts it so you have an S-shaped voltage vs. temperature characteristic
that's easy to solve for temperature -- I've done it either by
calculating the temperature from first principals using floating point
arithmetic or by building up a look-up table in a spreadsheet and doing
linear interpolation on integers.

Figuring out the Vout vs. thermistor resistance and Vref is left as an
exercise to the reader. Once you've figured out the thermistor
resistance you can get thermistor temperature using equations from the
manufacturer. Note that if the board's Vref is available to you this
circuit is entirely ratiometric.

Vref
+
|
|
.-. Rf
| | Rbias ___
| | .---|___|----.
'-' Ri | |
| ___ | |
<-----------o--------|___|------o |\ |
To '---|-\ | Vout
Thermistor | >-----o------
<-----------o--------|___|------o |/
| |
.-. Ri |
| | .-.
| | Rbias | |
'-' | | Rf
| '-'
| |
=== |
GND ===
GND
created by Andy´s ASCII-Circuit v1.24.140803 Beta www.tech-chat.de

4. GuestGuest

I would side-step all the analog stuff and go directly to digital.
Maxim/Dallas has a chip that speaks 3-wire SPI and would be ideal for your
purpose,
It's their DS1626/DS1726.

Jim

5. MichaelGuest

Also DS1620 and DS1820. I use gobs of DS1820's.

No calibration with these digitals. I like that. Also, attachment is
easy; I use wirewrap wire even for long runs. 3 conductors twisted
together with the aid of a hand drill. I have yet to see how long such
a run can be and still have a DS1820 work, but I've made many 30 foot
runs without encountering troubles.

6. Robert BaerGuest

I have found the LM35DT to be linear and accurate to at least 185C.

7. tlbsGuest

There will be 2 of them in total, one recording
What is the max. exit temperature. If this is a pressurized boiler,
then temperatures could get up to 1000F and a semiconductor temp.
sensor would "fry".

Is this design part of an industrial control system, or just a
lab/school experiment? If it is for an industrial control, you will
need to use devices qualified to the local building codes and national
electric codes.

If it is just a school lab experiment, then pretty much anything will
probably be OK (within reason). The suggestions, above, are all valid
so long as the temperatures are within the range of the devices. If
the expected temperatures are higher, go to Omega.com -- they have a
very wide range of temperature sensors. They will have something to

8. Ted EdwardsGuest

Is this a hot water system for your home?
That's the one I was thinking of. Linear, accurate and no hassles
assuming the answer to the above question is yes. If much above 100C,
then thermocouple and more complicated circuitry.

Ted

9. kmillarGuest

The max output temp is 85 degrees C
It is a domestic central heating system, split into 2 zones., the CH
and the Hot water.

I wish to measure the temp of the 'flow' and 'return' close to the
boiler.
It's all part of a bigger project to autmate my home central heating so
I can monitor and control it via the web.

Thanks for all the replies. I think I have all I need now.