Proper way of connecting thin film PT1000 sensors

[miro123]

I'm strugling to get decent& long term stable  electrical connection of PT100/1000 sensors.

https://nl.mouser.com/datasheet/2/619/Heraeus_Ni_LeadExtension_EN-2936545.pdf

Is seems that their end in Nickel wire
- Soldering with aggressive flux seems to me the wrong way - I've achieved  some result - but the solder connection lock bad - I have doubt about long term stability, What will be the connection resistance  after several years?
 - I did not tried any press-fit connection
My question is - what is the proper way to make a wire connection with those sensors? Is there any application note on this subject

[thm_w]

The ones I've used are silver or gold plated: https://www.te.com/commerce/DocumentDelivery/DDEController?Action=showdoc&DocId=Data+Sheet%7FPTF-FAMILY%7FA3%7Fpdf%7FEnglish%7FENG_DS_PTF-FAMILY_A3.pdf%7FNB-PTCO-045

If you want to solder to nickel then yes you need to use a very aggressive flux, maybe try another type or get plated leads.

[TUMEMBER]

Pt100/Pt1000 sensors are generally bonded by electric arc welding. The remelting of the weld is then the most durable and there is no problem with the "solderability" of materials.

[mzzj]

Pt100/Pt1000 sensors are generally bonded by electric arc welding. The remelting of the weld is then the most durable and there is no problem with the "solderability" of materials.

Or laser. Not very practical for home use. 

Electric arc welding against carbon(graphite) electrode is pretty easy at home shop if you have enough wire lenght on the sensor and can "live with" twisted T splice joint.
All you need is suitable power supply and piece of graphite(carbon brush from electric motor or carbon rod from welding store)
Some of the commercial thermocouple welders are just that. https://fi.farnell.com/labfacility/xf-1250-far/thermocouple-welder-5-10-min-250vac/dp/2897470

[branadic]

The answer is, it depends
First question to answer is, at what temperatures you want to operate the temperature sensor. Note that in case of soldering the melting point of your solder should be 100 K above maximum operating temperature for reliability reasons. Leaded solder is available with a melting point of around 300 °C, so your maximum continuous operating temperature is 200 °C here.
Beyond that soldering is no longer the way to go, without special solder and equipment. In such cases welding is most commonly used, since crimping/clamping tends to loose contact force over time and temperature.

-branadic-

[AeroETech]

Either welding, as other people have already mentioned, or - taken from the datasheet:

Elements with
stranded wire extensions are available in standard lengths
and can be customized to your optimum length for larger
requests.

That would be another good option.

[doktor pyta]

I use spot welding using graphite electrodes pulled out from zinc carbon batteries.
The graphite is used because it does not stick to both welded wires.

Popular spot welders used for LiPo battery packets welding are not sufficient to my experience.

Below some photos from the deep past.
Regarding reliability: approximately 30 sensors work without failures for 6 years operating at 200'C.
Kynar insulation of the wires puts a limit here.

Greetings!

[3roomlab]

I use spot welding using graphite electrodes pulled out from zinc carbon batteries.
The graphite is used because it does not stick to both welded wires.

Hi doktor, how many farads did you have on those aishi caps to make a good weld?
thanks

[doktor pyta]

1. 4700uF/25V each.
As I remember the supply voltage for optimal weld was 16V.

2. I think that graphite electrodes have too high resistance to make a good weld when connected to typical Chinese lipo spot welder powered from ~4V accu.
It was tested by me as i mentioned.

3. the P-MOS was used to interrupt the current flowing through the SCR. If You wisely use SPDT switch you can probably omit it.

4. one more picture attached.

[mzzj]

I use spot welding using graphite electrodes pulled out from zinc carbon batteries.
The graphite is used because it does not stick to both welded wires.

Interesting! It never occurred to me to try spot welding with graphite electrodes, I have always arc welded against graphite electrode. (Last time I actually used TIG welder for thermocouples.)

[miro123]

Thanks for all answers

First question to answer is, at what temperatures you want to operate the temperature sensor.
-branadic-

I use those sensor mostly under T=200C - T&M ovens. The temperature in most application even does not go beyond 85C.

I have many of then some have Au or Ag but some are Ni  - Most of the time I use easy soldered components - unfortunately my internal stock runs emty and I have to switch to dificult to solder parts.

I use spot welding using graphite electrodes pulled out from zinc carbon batteries.
The graphite is used because it does not stick to both welded wires.

Popular spot welders used for LiPo battery packets welding are not sufficient to my experience.

Below some photos from the deep past.
Regarding reliability: approximately 30 sensors work without failures for 6 years operating at 200'C.
Kynar insulation of the wires puts a limit here.

Thanks, I've started building the similar spot welder setup.

So I will try to find use of those nickel wired PT100/1000 sensors.

By my next purchase I will careful select Au/Ag plated wires.

- they are easier to solder

- the thermocouple effect is less dominant compared to Ni -> Cu.





Another aspect is the long term stability of those sensors. I was not aware about it. I find  link below interesting

https://www.eevblog.com/forum/metrology/long-term-stability-of-temperature-sensors/

Background information  - My reference resistor project

----------------------------------------------------------------------------

for my reference resistor project I want to have temperature sensor.

Four Vishay resistors in two serial two parallel are submerged in silone oil bath.

Mechanical part is almost ready.

All ordered component arrived

I tough that temperature sesnor in silicone oil bath was a easy task. But it seems that PT sensors with good defined long term stability are rare.

Happily my requirements are not tight - something like T=0.1K @15..20years. is still ok. Inert silicone oil environment should help too.

I start leaning toward NTC for this application

https://www.google.com/url?sa=t&rct=j&q=&esrc=s&source=web&cd=&ved=2ahUKEwjI-NiY-OD7AhXQsKQKHacnCQsQFnoECAYQAQ&url=https%3A%2F%2Fwww.eevblog.com%2Fforum%2Fmetrology%2Flong-term-stability-of-temperature-sensors%2F%3Faction%3Ddlattach%3Battach%3D412153&usg=AOvVaw09Khd8-kgf81hOig-AK4hG


Summary : The SMD-type sensor from Muratamanufacturing (NCP15XH103D03RC) intriguingly showed the least drift performance of 0.492 mK/year pp.




Greetings,

Miro

[maat]

Summary : The SMD-type sensor from Muratamanufacturing (NCP15XH103D03RC) intriguingly showed the least drift performance of 0.492 mK/year pp.

If you need to go nuttier, other alternatives might be:
Yellow Springs Instruments 44016, they are claimed to be stable to 100 µK in 8 months (https://aip.scitation.org/doi/10.1063/1.1686523).
This part of YSI was sold to Measurement Specialties Inc. in 2006, which was then sold to TE connectivity in 2014, sad story, but TE still sells them: https://www.mouser.de/ProductDetail/Measurement-Specialties/11032565-00?qs=FP9LbDqJ9ln9%252B5t8q1j%2FzQ%3D%3D

[Kleinstein]

If long term stability over a limited temperature range is the target the NTC type sensors are indeed a good alternative. It is not just the sensor itself. With the higher rate of resistance change the measurement side  / referece resistor gets less crictical and one can often get away with 2 wire connection.

A Ag/Au plating has very little effect on the thermal EMF. This applies to the good and the bad.  The bulk of the conductivity is still through the bulk of the wire and thus thermal EMF also from there. The temperature difference over the thin plating itself to the base metal is very small and this not much thermal EMF effect there either.