[SOLVED] AD620 RTD Bridge Circuit Help

[zeb59]

Hi all

I am using an AD620 to measure the differential voltage for a PT100 based temp measuring project, as per the attached circuit. The circuit performs correctly in a PROTEUS simulation. However in the circuit prototype the AD620 output on pin6 simply stays at 3.8 volts. After reading the literature I replaced the 7805 and the 7905 with their 9V equivalents thus raising the +/- supply voltages to 9 volts. This was done to bring the Common Mode input voltage to within the range specified in the datasheet. However now the o/p voltage simply sits at 8.6V.

What am I doing wrong?

TIA

Zeb

 

[BradtheRad]

It sounds as though the non-inverting input is always higher than the inverting input. Hence the output stays at the positive rail.

To get into a better range of operation, you need to reduce the volt level at the node between your RT1 device and R6, so that it swings above and below the node between R4 and R5. That will get your op amp output to swing between positive and negative.

Try changing R6 to 1k, and install a potentiometer next to it. Adjust until you see the output change. Watch that you do not send too much current through the RT1 device.

You may need to alter the gain as well, so that output stays within a sensible range of measurement.

 

[zeb59]

Try changing R6 to 1k, and install a potentiometer next to it. Adjust until you see the output change. Watch that you do not send too much current through the RT1 device.

Brad

Thanks for your input. Infact I did change the values but the o/p remains stuck at the same level.
Is it possible that the chip could have been damaged when the supply voltage was +/- 5 volts because VCM was 4.5 volts which is beyond the specs. I changed the chip but he result is the same.

Zeb

 

[BradtheRad]

Typical warning is to never pull the inputs higher or lower than the supply rails.

When a project doesn't behave as planned, it's time to test the project one section at a time.

Attach potentiometers to both op amp inputs so you can dial any volt level between the supply rails. Put a 10k series resistor at the inputs in case something inside is shorted. Put a meter on the output level to make sure the output swings positive or negative, according to whether the inverting input is higher than the non-inverting, and vice versa.

Measure resistance through the RT1 device (or however it indicates a change in temperature). See if it changes when heated or cooled. Not recognizing it I can't advise what is the proper way to hook it up, nor what voltage to apply, etc.

 

[Avla 0652]

zeb59, in your circuit the voltage at the AD620 inputs is too high, about +4.76V, it is out of the normal common-mode range.
Input common mode voltage should be less than +VS−1.4 according to the datasheet,
i.e. less than 3.6V (in case of +5/-5V supply).
Try to make R5 and R6 lower, and connect lower ends of R5 and R6 to -5V rail, not to the ground.

 

[zeb59]

Avla,

Thanks for the response.

As I said in my earlier post, after realizing the problem of high voltages on both inputs, I changed the +/- 5 volts to +/- 9 volts. Theoretically that creates an upper limit of 9-1.4= 7.6 volts. With 4.76 volts on the input this is well within spec. However the AD620 behave the same way with higher rail voltage, except now the o/p is stuck at 8.6V and does not budge.

Zeb

 

[Avla 0652]

zeb59, sorry, I meant that with +/-9V you still have 8.57 V at the inputs
(9V / (2k + 100)) *2k = 8.5714 V
It is still too close to V+ rail.

 

[zeb59]

Avla

Sorry I did not provide full details of the circuit.

In fact the prototype has a dedicated 7805 to power the bridge. So the voltage at the inputs remains 4.7 with 9 VDC rails.

Zeb59

 

[BradtheRad]

Same as Avla, I'm looking at R4 (100 ohms). That is what pulls the inverting input very close to the supply rail.

Then looking at your RT1 device. I can't be sure but is it a 25 ohm impedance? Because it looks like it contains a variable resistor (or potentiometer?) with the label '25.00' next to it.

So if RT1 has an approximate 25 ohm operating resistance, then it is strongly pulling the non-inverting input close to the supply rail. So it would make sense for R4 to be 25 ohms too.

This is just guessing on my part.

 

[Avla 0652]

BradtheRad
No, RTD-PT100 means that it is 100 Ohm,
'25.00' probably means 25 degrees C

 

[FvM]

You have demonstrated sufficiently the problem of not posting the real circuit. :smile:

But instead of adding more supply voltages to the circuit in post #1, it would be very easy to flip the positions of 100 ohm and 2k resistors in the bridge and make it work with +/- 5V only. If it still doesn't work as expected, you would want to check the bridge voltages first.

 

[zeb59]

You have demonstrated sufficiently the problem of not posting the real circuit. :smile:

FvM,

Thanks for the input. I attach the full circuit from PROTEUS. My intent in posting the simplified circuit was none other than to reduce the possible confusion that too much detail could cause, and detract others from the main issue.

I also confirm Avla's contention that the PT1 device is indeed a PT100. I tried to attach the file models the circuit in EXCEL but the system does not let me. I am therefore emailing it through a private message (if I can do it).


Regards

Zeb

- - - Updated - - -

Hi all,

I zipped the EXCEL model and attach the same herewith.

Zeb

- - - Updated - - -

Hi all,

I zipped the EXCEL model and attach the same herewith.

Zeb

 

[zeb59]

.... it would be very easy to flip the positions of 100 ohm and 2k resistors in the bridge and make it work with +/- 5V only.......

FvM,

We flipped the inputs as you suggested AND IT WORKS. The software has to be tweaked as the displayed temp is off by 2 degC from the actual temp.

The moral of the story is that V(Common Mode) must be kept as far away as possible from the Rails voltages.

Thanks everyone. The working circuit (PROTEUS Files) with microcontroller and associated code (PROTON BASIC), the display are available to anyone interested.

Zeb