In that case, I think the best approach is the one suggested by @Kjeldgaard in post #9, which is the circuit posted here.

I have looked at that design several times. I think my ignorance becomes a bit apparent as it is unclear to me from that diagram how exactly I would control it with a Microcontroller DAC or PWM.

 

I have looked at that design several times. I think my ignorance becomes a bit apparent as it is unclear to me from that diagram how exactly I would control it with a Microcontroller DAC or PWM.

Replace the pot shown in that diagram with a digital pot.

 

In some cases our customers only have one port into there tanks that is currently filled with a PT100 temp sensor. What I would like to do is remove the PT100 and replace it with our sensor.

What is the rationale for that?
Why replace an accurate PT100 sensor with a different sensor, and then try to emulate the PT100?
Or are you putting something in addition to the sensor inside the tank?

 

What is the rationale for that?
Why replace an accurate PT100 sensor with a different sensor, and then try to emulate the PT100?
Or are you putting something in addition to the sensor inside the tank?

Yes that is exactly the rational, We have sensor package that tracks 4 things including temp. But most customers only have one port in there tanks. So with a pt100 output we can kill 2 birds with 1 stone.(getting our sensor into the tank and having a pt100 output.)

 

I get it now, you want to place your sensor (not an RTD) in the tank but have the data output as if it was an RTD.

 

I get it now, you want to place your sensor (not an RTD) in the tank but have the data output as if it was an RTD.

Yes, precisely!

 

So what temperature accuracy/resolution do you want for the RTD simulated resistance?
What is the accuracy of your sensors?

 

So what temperature accuracy/resolution do you want for the RTD simulated resistance?
What is the accuracy of your sensors?

We are using a digital temp sensor that is accurate to .1C. I would like to be able to simulate 0-100 degrees C if possible.

 

I have been testing a voltage divider with a 3.3V 12 bit DAC and a 100ohm x 33ohm voltage divider. I can get the range I need 0-100C (actually a lot larger then I need from -171 to 165) with around .2C change per 1 count of change on the DAC output. Is there any downside to using the DAC and voltage divider?

 

Can you diagram your voltage divider and show how you are connecting the DAC?

 

I have been testing a voltage divider with a 3.3V 12 bit DAC and a 100ohm x 33ohm voltage divider. I can get the range I need 0-100C (actually a lot larger then I need from -171 to 165) with around .2C change per 1 count of change on the DAC output. Is there any downside to using the DAC and voltage divider?

How is that simulating a resistance as you wanted?

 

Can you diagram your voltage divider and show how you are connecting the DAC?

 

How is that simulating a resistance as you wanted?

Well it's not. Someone suggested that the resistor controls the amount of voltage to the RTD amplifier "sees" and that by pumping more current into the circuit I could control the temp like I want to. by tricking the amplifier into thinking the resistance went down by increasing the voltage.

 

Well it's not. Someone suggested that the resistor controls the amount of voltage to the RTD amplifier "sees" and that by pumping more current into the circuit I could control the temp like I want to. by tricking the amplifier into thinking the resistance went down by increasing the voltage.

Okay.
But the usual way of measuring the resistance of a PT100 sensor is to put a constant current through it and measuring the voltage.
So, unless you know that current, and that it doesn't vary between different setups, then you don't know what voltage corresponds to that current through the sensor.

Here's a (possibly Eureka) thought.
Measure the current (high-side) through a small resistor in series with the line, calculate the voltage you need to simulate the PT100 voltage at that temperature and current, and then apply it to the line minus the drop across R1 (conceptual diagram below).
You could initially start the system with the amp at zero volts to measure the current and then adjust the voltage to the desired value (but the current can be measured at any time).
The amp has to sink the current, but that shouldn't be a problem.

This has the advantage of having fewer error sources and being potentially more accurate than trying to simulate a resistance with an analog circuit or digipot.
For example, with a differential A/D to measure the voltage across R1, the only other error source, besides the A/D and DAC accuracy, is the resistor tolerance (and that could be compensated for during the initial system calibration, if needed).

 

Without going back into my notes which is not very reliable with my poor memory... There are at least 3 distinct types of digipots. All seem to be broken into 100 segments with some means of switching them in or out of the ckt to vary the total resistance between the in or out and the wiper pins. A couple of different ranges but 10K seems popular. Some, or most? with external power across the resistor for more than the 5V logic power the chip runs on. The 2 distinct variations are microprocessor w/ rotary encoder controlled using 2 wire output communications connection or switch/pulse controlled using a single pushbutton or a circuit that can produce repeating pulses to the UP and DWN input pins of the digipot.

 

Here's a (possibly Eureka) thought.
Measure the current (high-side) through a small resistor in series with the line, calculate the voltage you need to simulate the PT100 voltage at that temperature and current, and then apply it to the line minus the drop across R1 (conceptual diagram below).
You could initially start the system with the amp at zero volts to measure the current and then adjust the voltage to the desired value (but the current can be measured at any time).
The amp has to sink the current, but that shouldn't be a problem.

I agree that this would be possible except for one thing; the Maxim IC may not be in continuous measure mode. It can have the bias turned off when not making a measurement and it also has a 1-shot measurement mode.
I think the best chance to do this is to use the operational transconductance amplifier circuit in post #6. This would have to be tested for response to the 1-shot or bias removal.

 

the Maxim IC may not be in continuous measure mode

Isn't that Maxim IC just for test purposes?

This would have to be tested for response to the 1-shot or bias removal.

Yes, if the customers system operates in a pulse mode, then my proposed system would have to be fast enough to respond to that, but that response time should only be a fraction of a second.
And it could easily detect no bias current, and not put out a voltage under those conditions.