Home
JAQForum Ver 24.01
Log In or Join  
Active Topics
Local Time 12:12 07 Jul 2025 Privacy Policy
Jump to

Notice. New forum software under development. It's going to miss a few functions and look a bit ugly for a while, but I'm working on it full time now as the old forum was too unstable. Couple days, all good. If you notice any issues, please contact me.

Forum Index : Microcontroller and PC projects : Reading Thermistors.

Author Message
Phil23
Guru

Joined: 27/03/2016
Location: Australia
Posts: 1667
Posted: 10:22pm 02 Jul 2016
Copy link to clipboard 
Print this post

Before I do too much more "Welding" on the board to control my head pump, can I get some advice on how to connect the thermistors correctly.

I've been play with one on a breadboard for a while, so It's basically working, but know there is other considerations.

What I Have so far.





The circuit is this.



I know I should consider filtering on the inputs, but way past rusty there.

Should I do this?




Or this? With what value resistor?




The Heat Pump contains 5 thermistors, 4 of which are installed in a manner such that they are not easily replaced with digital sensors.

Their prime purpose is monitoring temperatures to protect the compressor from overloading & the evaporator icing up.

So they are buried amongst the 240V components to an extent.

Thanks

Phil.
 
matherp
Guru

Joined: 11/12/2012
Location: United Kingdom
Posts: 10215
Posted: 02:01am 03 Jul 2016
Copy link to clipboard 
Print this post

My approach to this would be in three areas:

First, make sure the voltage on AVDD is clean. Use the circuit on the colour Maximite with a resistor and capacitor. Have just a single point of connection between Analog ground and the digital ground. Don't just use the analog ground as part of a single ground plane.

Second, it is probably worth having a filter on the inputs 1K + 0.1uF should be OK

Finally, and most important is the wiring to the thermistors. I would use two core twisted pair wire with a separate shield. The shield should be grounded at the Micromite end and left open at the thermistor. The two cores then connect to AVDD and Analog ground.
 
Phil23
Guru

Joined: 27/03/2016
Location: Australia
Posts: 1667
Posted: 10:21am 03 Jul 2016
Copy link to clipboard 
Print this post

Thanks Peter,

I've already draw heavily on your other post regarding using thermistors.

Only thing I don't have a lot of option with is the wiring as the thermistors are already installed.




All they have is a light twin core attached.

Only the ambient one would be easily changed for a DS18B20, but thought I'd leave it in place & add a digital one for reference.

They do not need to return precise values, mainly comparisons to protect against failure events.





Still having a bit of trouble establishing what the correct co-efficient are for the sensors.

I have my own test data & service manual test Temp Vs Resistance values, but they don't quite agree & neither is giving a linear reading.

That said, the test bread board is a mess, so could be a contributing factor.

Cheers

Phil
 
crez

Senior Member

Joined: 24/10/2012
Location: Australia
Posts: 152
Posted: 01:26pm 03 Jul 2016
Copy link to clipboard 
Print this post

Phil,
If you have enough filtering on the input then twisted pair is not important. The lowest frequency you need to filter out is 50Hz, so set your filter cut-off around 10Hz and you'll be fine. This corresponds to 1K series resistance with 15 microfarads across the input. With that sort of filtering your reading will still settle at the right value in a few seconds.

David
 
centrex

Guru

Joined: 13/11/2011
Location: Australia
Posts: 320
Posted: 12:02am 04 Jul 2016
Copy link to clipboard 
Print this post

Hi Phil
Just a question Why would you not allow the current thermocouples to control the heat pump via its own controller and just turn the power on when heat is required?
Cliff
 
Phil23
Guru

Joined: 27/03/2016
Location: Australia
Posts: 1667
Posted: 10:49am 04 Jul 2016
Copy link to clipboard 
Print this post

  centrex said   Hi Phil
Just a question Why would you not allow the current thermocouples to control the heat pump via its own controller and just turn the power on when heat is required?


The Smoke has escaped from the original controller.










Not even worth considering repairing, as they have a bad reputation, but also, as seen at the bottom of the first picture, it also reads tank sensors connect to the ribbon cables.

I believe the circulation pump fails first, if not routinely cleaned,
(as if owners will do that), then it's MOSFETS or whatever from there.

Can't see why this should happen, as in the event of the pump seizing, the controller should see heat exchanger input/output differences that would indicate a fault.

My modified version does away with the storage tank & original circulation pump.

Cheers.



 
Phil23
Guru

Joined: 27/03/2016
Location: Australia
Posts: 1667
Posted: 11:33am 04 Jul 2016
Copy link to clipboard 
Print this post

Thanks Peter,

  matherp said   My approach to this would be in three areas:

First, make sure the voltage on AVDD is clean. Use the circuit on the colour Maximite with a resistor and capacitor.[/quote]

Ok,
Have done that. Correct?

Gnd => 0.1µF => Pin 28 => 10k => 3.3V Bus.

[Quote]Have just a single point of connection between Analog ground and the digital ground. Don't just use the analog ground as part of a single ground plane.
[/quote]

Ground side of thermistors is already wired to a 0V bus, I can isolate that strip, then take it to pin 27, but where else would you recommend?

[Quote]
Second, it is probably worth having a filter on the inputs 1K + 0.1uF should be OK
[/quote]
Had already added the 0.1µF caps in Drawing B when I got back to this post.
So at the moment it's essentially across the thermistor & earth.

Was going to guess on 5 to 10k for the series resistors considering the thing will be in near the compressor & 240V 50Hz.

Is that not a good thought?

I can still change the wire links seen in the pic with series resistors, but would like a bit more advice first.

[Quote]
Finally, and most important is the wiring to the thermistors.
[/quote]
Unfortunately I can't do much with that. Have to stick with what I've got.

The purpose of this Micromite is primarily to protect the heat pump from damage in the event of loss of circulation, which would be indicated by input/output temps;
Compressor overheating and Evaporator icing up,
so super accurate values aren't that crucial.

I've already had a couple of scares where circulation has been lost & the compressor has begun to over heat due to lack of water flow.

It will most likely have a 20x4 line LCD to display values.

Cheers

Phil.

 
TassyJim

Guru

Joined: 07/08/2011
Location: Australia
Posts: 6269
Posted: 11:57am 04 Jul 2016
Copy link to clipboard 
Print this post

  Phil23 said   Thanks Peter,

  matherp said   My approach to this would be in three areas:

First, make sure the voltage on AVDD is clean. Use the circuit on the colour Maximite with a resistor and capacitor.[/quote]

Ok,
Have done that. Correct?

Gnd => 0.1µF => Pin 28 => 10k => 3.3V Bus.


Cheers

Phil.



That's 10 ohms, not 10000 ohms!

Jim
VK7JH
MMedit
 
Phil23
Guru

Joined: 27/03/2016
Location: Australia
Posts: 1667
Posted: 12:22pm 04 Jul 2016
Copy link to clipboard 
Print this post

Ugh,

Thanks Jim.

Not reading.....

Also, realised 10k in series with the thermistor divider circuit would give a 159 Hz cut off.

47k would cut off at about 34Hz.

Is that Ok, or am I on the completely wrong track here?
I've already added the 0.1µF from the analogue pins to ground.

Thanks

Phil.
Edited by Phil23 2016-07-05
 
matherp
Guru

Joined: 11/12/2012
Location: United Kingdom
Posts: 10215
Posted: 08:55pm 04 Jul 2016
Copy link to clipboard 
Print this post

  Quote  Also, realised 10k in series with the thermistor divider circuit would give a 159 Hz cut off.


Don't go with high value resistors. This increases the input impedance beyond the 10K limit at which the PIC ADC works properly. If you really need a lower filter increase the capacitance instead.
 
Phil23
Guru

Joined: 27/03/2016
Location: Australia
Posts: 1667
Posted: 10:55pm 05 Jul 2016
Copy link to clipboard 
Print this post

  matherp said  
Don't go with high value resistors. This increases the input impedance beyond the 10K limit at which the PIC ADC works properly.


Thanks Peter,

Had no idea about that.

Can you recommend a good reference that contains general introductory & helpful information in these areas.

Either in print or electronic form, (prefer the latter).

Something covering a broad base in basic areas.
(Edit:- Not sure if above qualifies as tautology or not).

I have little interest in other languages at this stage, mostly interested in circuitry etc.

Cheers

Phil.Edited by Phil23 2016-07-07
 
matherp
Guru

Joined: 11/12/2012
Location: United Kingdom
Posts: 10215
Posted: 11:24pm 05 Jul 2016
Copy link to clipboard 
Print this post

Sorry there are no easy guides to this. In fact the recommendation is 5K for the MX170. See AD17 on page 275 of the MX170 device manual

The 10K comes from AD15 on the same page

However see note 1
 
Print this page


To reply to this topic, you need to log in.

The Back Shed's forum code is written, and hosted, in Australia.
© JAQ Software 2025