Pt100 sensor precision / accuracy question , batch sample selection .

[Overspeed]

Hello

What do think of that ? that an ultra low noise current source circuit

Regards
OS

[EC8010]

Batteries are quieter than Zeners, but the rest of the circuit seems unnecessarily complex for what it does. Ultimately, a constant current source compares the voltage developed across a sensing resistance with a reference voltage. All that is needed to do that is a quiet op-amp. You could make a quiet op-amp with JFETs at the input, but there's no need for them to pass the entire output current. I think you need to define what it is you are trying to do before diving into the realms of awkward constant current sources. What temperatures do you need, how stable do you need them to be, and over what time scale?

[Overspeed]

To EC

Hello

I try to reach the best possible result for the costs to match and sort Pt100 and avoid to pay 1700 USD for a Hart probe or 3400 USD for a pair of probe as I need a differential measure in the 0.05 Deg error range .

If I refer to Dieter post regarding Isotech graph value :

This log shows a sigma of about 10 ppm (0.001 Ohm) which translates to about 1/385 K = 2.5 mK.
In order to get there the resistance measurements need to be in 4 wire mode. In 2 wire mode there will be much larger errors due to connector and wire resistances.

Regards, Dieter

Measuring resistance in the range of 10 ppm (0.001 Ohm) request a real stable + low noise current source far over the basic ohmmeter source so a real proven circuit I strongly doubt than 20000 Euro multimeter as the Fluke 8558A use a simple solution .

My post with the FET circuit was just a ''possible solution ''

my question 1 is : What circuit is used on Fluke 8558 A or other even Agilent 3458A ?

my question 2 : If a Voltage ref and a resistor is enough , perhaps but what exactly ? could be interesting to have a BOM / schematic

all ohmmeter I know ( AOIP 22 and 27 . Valhalla 4300 even HIOKI are in the 0.03 % so .....a bit far from 0.001 Ohm on a 100 Ohm resistor

Regards
OS

[EC8010]

What you're paying for in those expensive probes is calibration (and very small quantity production). But you're capable of doing the calibration, so no need to pay. The issue of noise in the CCS comes down to the sense resistor. Make the voltage across that large enough and noise from the voltage reference barely matters. Obviously, as you make the sense resistor larger, it dissipates more power, and in heating, changes its value. But you only need to drive 1mA, and if you used a 1k resistor, it would only dissipate 1mW. A close tolerance wirewound resistor would be perfectly adequate. Perhaps not a metal foil because they're so small and the temperature rise would be greater. After that, an ADR1399 would produce a stable 6-7V that could be attenuated to 1V and compared by an OPA1641 op-amp to the 1V developed by the 1k sense resistor. That would be a handful of components that would fit in a matchbox. Lead-acid batteries are big, heavy, gently drift down in voltage, and need recharging. The 3458A probably derives its CCS reference from its main LTZ1000 reference.

[mzzj]

To EC

Hello

I try to reach the best possible result for the costs to match and sort Pt100 and avoid to pay 1700 USD for a Hart probe or 3400 USD for a pair of probe as I need a differential measure in the 0.05 Deg error range .

0,05 degree differential accuracy in 0-100cel range is fairly easy task. Most of the meter/readout errors cancel out as they are same for both probes. Old 34401 will do fine.
Calibration is also fairly easy if you are mostly concerned in the differential temperature, again most errors cancel out and you don't need to worry for example about reference uncertainty.
4-point calibration at 0, 33, 66, 100  to match the sensors with each other should get you within few mK over the entire 0-100cel range.

Film-type Pt-100 sensors have possibly too much hysteresis, especially if you can't be sure that both sensors experience similar history of thermal cycling. ( one sensor approaching 50Cel from 100Cel and another going from 0 to 50Cel would be bad for hysteresis)
Whatever sensor you select you better first "anneal" the sensor at somewhat higher temperature than expected in use and also test for hysteresis by taking one sensor from 0-100-50 and another 0 to 50 and compare the readings at 50Cel.
1 hour 150C to 200C anneal repeated 3 times in this case sounds about right.

[dietert1]

In another thread i posted some results of a high resolution temperature measurement using a NTC, see https://www.eevblog.com/forum/metrology/scannermultiplexers-for-voltage-references/msg5533611/#msg5533611.
The schematic shows that circuit in a revision for measuring a PT100. The ADC and its reference are on a little module that is cheap at amazon. With the NTC the reference resistor is a metal foil 10K resistor mounted next to the NTC. This time one needs a 100R precision resistor. Also shown is the 5 pole half bridge connection that eliminates connector contact resistance from the measurement. Thermal EMF is eliminated using the ADG736 MUX for polarity reversal. One of the ADS1256 DIO pins can be used for polarity control. The two 1 KOhm resistors give us about 1 mA test current at 2.5 V reference voltage. They don't need to be precision resistors, as the resistance measurement is a ratio of two voltage measurements:
Rx = Ux / Urref * Rref
For best accuracy the two voltages should be measured with the same PGA gain setting. The ADS1256 incudes high impedance input buffers. They should be turned on for best linearity. This circuit will do the 10 ppm measurement. With the NTC one gets a resolution of 4 uK in one second. Accuracy depends on the sensor element and its calibration.

Regards, Dieter

[Overspeed]

Dieter
Thanks

I will order some components to make a test and try to install them in the sensor enclosure in the aim to enter in the triple point bath

Regards
OS