Author Topic: Temperature sensor technology  (Read 1178 times)

0 Members and 1 Guest are viewing this topic.

Offline TeunTopic starter

  • Regular Contributor
  • *
  • Posts: 70
  • Country: nl
Temperature sensor technology
« on: May 04, 2020, 07:40:30 am »
Hi guys,

I'm looking for a temperature sensor/probe. BUT.... not just any sensor! One like this:
983852-0983856-1

It came from a 40/50 year old watermixer in a bakery. Mixes hot and cold water ;D No (service)manuals or anything available on the device. Even the company that services them now doesn't have any info on it anymore. And told me a new one could cost me only €1800! :-DD

I think it is a pt100 or pt1000 like thing. But in case I am wrong about that. How do I find out what kind of sensor it is? Then I might be able to fabricate something.

Measured the resistance as an open circuit, so it's broken.
Haven't had much luck finding a replacement because of the rim it has, to be able to screw it in the waterpipe.

Someone know of things like these? Or can someone help me to at least find out what kind of sensor it is?

Thanks!
 

Offline tszaboo

  • Super Contributor
  • ***
  • Posts: 8218
  • Country: nl
  • Current job: ATEX product design
Re: Temperature sensor technology
« Reply #1 on: May 04, 2020, 07:58:32 am »
Maybe it is a bimetallic switch. Heat it up, and see if it becomes short circuit.
 

Offline Rerouter

  • Super Contributor
  • ***
  • Posts: 4705
  • Country: au
  • Question Everything... Except This Statement
Re: Temperature sensor technology
« Reply #2 on: May 04, 2020, 08:10:00 am »
Time to take a file and crack that old one open if you really believe its not working.
 

Offline TeunTopic starter

  • Regular Contributor
  • *
  • Posts: 70
  • Country: nl
Re: Temperature sensor technology
« Reply #3 on: May 04, 2020, 10:16:17 am »
Well heating it up didn't cause a change in resistance in any case. I had already tried that.
 

Offline tszaboo

  • Super Contributor
  • ***
  • Posts: 8218
  • Country: nl
  • Current job: ATEX product design
Re: Temperature sensor technology
« Reply #4 on: May 04, 2020, 10:49:53 am »
Well heating it up didn't cause a change in resistance in any case. I had already tried that.
Maybe you can take a look at where it is connected. If it is a PT100, then it will have a bunch of electronics there, probably matched BJTs and such. Is the temperature adjustable on the dodad?
 

Offline ace1903

  • Regular Contributor
  • *
  • Posts: 240
  • Country: mk
Re: Temperature sensor technology
« Reply #5 on: May 04, 2020, 11:14:31 am »
Take a potentiometer and connect to the original controller. Start with something like 100k in case it is NTC.
Observe display or behavior. Try then 2k2 potentiometer in case it is PT1000 and last 150ohm if it is PT100.
If it outputs hot or cold water you can find if it expect NTC or PTC(PT100/Pt1000).
Once you find the correct sensor it expects then you can make on lathe proper adapter for the standard sensor you can buy.
In the past, I worked on similar equipment and often  KTY81 was used in old equipment.
 

Online jpanhalt

  • Super Contributor
  • ***
  • Posts: 4005
  • Country: us
Re: Temperature sensor technology
« Reply #6 on: May 04, 2020, 11:14:37 am »
I would not destroy the casing, as that may well be the most difficult part to replace.   The wires look that they connect to something inside the device that is potted.  I would try to dig them out -- assuming you are convinced the device is bad.
 

Offline TeunTopic starter

  • Regular Contributor
  • *
  • Posts: 70
  • Country: nl
Re: Temperature sensor technology
« Reply #7 on: May 04, 2020, 11:50:23 am »
Take a potentiometer and connect to the original controller. Start with something like 100k in case it is NTC.
Observe display or behavior. Try then 2k2 potentiometer in case it is PT1000 and last 150ohm if it is PT100.
If it outputs hot or cold water you can find if it expect NTC or PTC(PT100/Pt1000).
Once you find the correct sensor it expects then you can make on lathe proper adapter for the standard sensor you can buy.
In the past, I worked on similar equipment and often  KTY81 was used in old equipment.

Great idea. I'm going to try that.
 

Offline TeunTopic starter

  • Regular Contributor
  • *
  • Posts: 70
  • Country: nl
Re: Temperature sensor technology
« Reply #8 on: May 04, 2020, 11:56:41 am »
Maybe you can take a look at where it is connected. If it is a PT100, then it will have a bunch of electronics there, probably matched BJTs and such. Is the temperature adjustable on the dodad?

Its a Industrial installation. There is a control unit for the valves, temperature etc. (2E Vandblandere B48 for the fanboys)
 

Offline Tom45

  • Frequent Contributor
  • **
  • Posts: 556
  • Country: us
Re: Temperature sensor technology
« Reply #9 on: May 04, 2020, 05:46:06 pm »
Before even trying the potentiometer test, turn on the device and see if there is a voltage at the terminals where the sensor's wires should connect.

If there is a voltage, then the sensor might be a thermistor, RTD, or bimetallic switch.

If no voltage, then perhaps it could be a thermocouple.
 

Offline Niklas

  • Frequent Contributor
  • **
  • Posts: 404
  • Country: se
Re: Temperature sensor technology
« Reply #10 on: May 06, 2020, 07:23:57 am »
This manual in German is not from the B48, but from a similar equipment also produced by 2E:
https://pitec.ch/media/wysiwyg/technische_informationen/B11-Pitec_-_de.pdf
Page 8 and forward might be of interest.

Most of the info about 2E's water mixer equipment seems to have an adjustable temperature in the range of 0-60'C with 1'C resolution, so your mystery sensor is most likely not a bimetallic switch. If it were a thermocouple, then you would have special wires and connectors, not that kind of screw terminals marked as 762 on page 10.

The different error indications on page 8 can also give some clues if the B48 has similar indications. The two t messages at the bottom indicates high temp/sensor open circuit and low temp/short circuit. If you can get a temperature reading form the display, then you can also use that to figure out what sensor type and resistance the mystery sensor was. NTC resistance is given at room temperature and Pt100/Pt1000 at 0'C. I would have tried with standard resistance values like 47k, 22k, 10k, 4.7k, 2.2k, 1k etc and watched the output on the display. Map the indicated temperature and sensor resistance and look at the resistance at 0'C and 20'C to decide whether it is Pt100/Pt1000 or NTC.
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf