PT100 Differential Amplifier

[rudika79]

Hi,

I have problem with differential amplifier which not working how should be. The circuit is not designed by me and I can speak with design engineer in this moment. The problem is the output (UT+ and UT-) I measuring all the time same voltage. Not changing when is changing resistance on PT100 probe. When I measure input voltage on AJ4 connector then I can measure changes on PT100 probe. In this moment, I use 138.51ohm for 100°C and 175.86ohm for 200°C resistor to emulate PT100 probe. I checked PCB for obvious fault like short, component not soldered properly, component on right place, power supply voltages and noise.

Here is what can I measure on circuit.

when I am using  138.51ohm for 100°C
 Between AJ4 connector pin 2 and 3 : 0.238V
 Between UT+ and UT- output measured : 2.321V (should be around 1.95V)

when I am using 175.86ohm for 200°C
 Between AJ4 connector pin 2 and 3: 0.308
 Between UT+ and UT- output measured : 2.321V (should be around 2.45V)

Vcom is 2.5V

I'm just start to running out of idea what is problem whit circuit.  I just need idea what could be wrong whit this circuit. And I will be need to repair it....

Thank you,
Rudolf

[awallin]

Here is what can I measure on circuit.

when I am using  138.51ohm for 100°C
 Between AJ4 connector pin 2 and 3 : 0.238V
 Between UT+ and UT- output measured : 2.321V (should be around 1.95V)

when I am using 175.86ohm for 200°C
 Between AJ4 connector pin 2 and 3: 0.308
 Between UT+ and UT- output measured : 2.321V (should be around 2.45V)

Vcom is 2.5V

looks like the ADA4932 is configured for a gain of 10, so you might see "clipping" because your input voltage is too high (or your gain is too high).
Try reducing the gain by lowering R99 and R102 by 2x or 5x and see what happens.

If you want to optimally fill the dynamic range of an ADC you should be comparing the PT100 resistance to some reference resistor that corresponds to the middle of your desired measurement range.

ADA4932 with a bandwidth of ~500MHz seems way overkill for a PT100 measurement. Why not use a low bandwidth DC-accurate instrumentation amplifier (AD8221 ?) or zero-drift op-amp?

[MK]

you have about a 150-160 mV voltage across the resistor that you are amplifying by the ad4932, so a max gain of 25 will clip at temps above 200 C or so, is the AD4932 specified for rail to rail input, as yours is sitting up at 5 V? You need a differential cicuit so that you see just the change in resistance, just blindly using a differentail amp is not the answer.

[rudika79]

Thank you for your replay MK and awallin! I found the problem. The gain was to high.... I reduced a gain and start to get output changes...

Thank you,
Rudolf