Resistance measurement Reference design solution

Hi Ilya,

I have moved this post for further assistance with your query. Kindly note that as this a holiday season, replies might get delayed.

Best Regards,

Bhushan

Hello Ilya, 

Please provide more details on the "solution" in your post.
TI cannot tell what you are asking for.

Hi Ed

I need develop a device that gets residence value from 6 sensors, each sensor provide residence value from 140Kohm to 1.3Mohm

I need measured the resistance with resolution of 0.001% .

hope now my explanation a little bit clear

Best Regards,

Ilya

How do you want the measured resistance value to be communicated to your system, Ilya?
     SPI, I2C, Parallel, Jesd204, or Other?

It seems you need some A to D converters, or a Micro Controller that has some A to D's inside.
Is there a preferred method being considered?

Hi Ed

Probably will be some controller with high resolution of AtoM, or poor control and you can suggest AtoD chip, I'm open to all options.

any communication, it doesn't matter,I accept any solution, i am existing at development stage.

Main core of device is gets resistance value from 6 sensors. all peripheral i will build around solution --> "resistance value measure"

Best Regards,

Ilya

Thank you Ilya,

I will notify the MSP430 team of your needs.
MSP430's come in a wide variety with up to 20 A-D's, and can provide the communication and central function of a sensor system.
Please allow extra time during the Holidays for a reply from the MSP430 team.

Here is the landing page to see the possibilities.
https://www.ti.com/microcontrollers/msp430-ultra-low-power-mcus/overview.html

Hi Ed

Thank you for suggestion,

But i think only MCU -MSP430, not will solve my requirements, is MSP430 can measured resistance ?

I need a solution that can read the resistance.

Best Regards,

Ilya

MSP is a heart of many sensing applications just like yours.
Let's allow the MSP team some time to reply with their thoughts.

Hi Ed

Can you please estimate the answer time from MSP team?

Thank you

Ilya.

Hi Ed

Unfortunately i not got any answer from MSP team, i am exist in major stage i need find solution to  my issue.

Thank you

Ilya.

Hi Ed

Pass 21 days from your note, but unfortunately i not got any answer.

Can you please estimate the answer time from MSP team?

Thank you

Ilya.

Hello Ilya,

One common practice for precise resistance measurement is to use accurate current source to convert resistance value into voltage signal, the use high resolution sigma-delta ADC to measure this voltage. this technique is used for RTD sensor measurement like the image below. The current is passing through both the measured resistor, and a reference resistor. the voltage measurement is done referenced to the reference resistor voltage, in technique called radiometric measurement. There are specialized ADCs dedicated to such kind of measurement.

Your case have several challenges though:

- the maximum resistance value is quite high
- the need to measure 6 sensors instead of one
- the required accuracy is extremely high, not typical for industrial applications.
I'm going to address some of these aspects in later posts for the benefit of shorter posts

The issue of maximum resistance 1.3M

- having very high resistance to measure mandates very high input impedance, in this case high impedance is required. This takes away the possibility to use potential dividers to scale down the voltage generated on this resistance. we have to measure the voltage directly on the resistance. Most of the dedicated ADCs I mentioned above are 5V ADCs, which would mean a current source in the range of 2-3uA is needed. This is much lower than the lowest value those ADCs can generate.
- a high voltage input ADC like our recent ADS125H02 can be used, which accepts inputs up to +/-15V. This relaxes our lowest current requirement to ~10uA, which is kind of low but practically possible.

- still we can't use the internal IDACs within the ADC for a couple of reasons; Its minimum setting is high 50uA, and it's compliance range is below 5V, while measuring 1.3M with 10uA source would require 13V compliance at least.

- that leaves us with the option of creating external current source. This current source needs to be high side, and can be either variable or constant. variable sources would have some benefit in noise performance in case of lower resistor, but when we apply 10uA to 140kOhm we get 1.4V which is still good to measure with ADS125H02 especially the existence of PGA. so you can go simply for fixed current source.
please refer to High-side current sources for industrial applications for more details about different options to create such current source.

PS: if you need still to use the 5V ADC, you might think of adding a resistor in parallel to the measured one in order to reduce the overall impedance. this would of course impact the accuracy and resolution of measurement, and render the measurement nonlinear and some digital calculation is needed to get the resistor value.

The issue of measuring multiple resistors

- in the case of multiple resistors, differential multiplexer at the input, and another multiplexer for the current source would allow the use of a single current source, and single ADC.

- Suitable high voltage multiplexer (like MUX36D08, or MUX507 for the differential 6 inputs), and a similar one for the current source path (excitation multiplexer) should be fine, and eliminate the need for multiple current sources. there are some newer TMUX family devices whose names I don't have out of my head now.

- we can't get rid of the current source multiplexer and use only an input multiplexer as there will be a drop over the ON resistance of that multiplexer, and the input resistor should have no drop whatsoever.

- the switching time of the current source should be factored in, some current sources take time to settle to proper accuracy level. if higher speed switching is needed may be a current steering technique is needed and make before break switching is necessary

-for ADS125H02 reference resistor would require 1V at least, while our maximum resistor voltage drop would be 13V with 10uA current. the current source should have a compliance of 14V minimum, along with any drops that might occur on the mux on resistance. a current source supply a bit higher than 15V might be necessary based on the current source selected.

Hi Ahmed

Thank you to clear explanation. I will learn the RTD Measurements topic, Because I'm not familiar with this topic.

Please advise if i am understand clear your explanation:

I need use ADS125H02 because - with high impedance of 1.3M with 10uA the required source require is 13V, and your calculation of 10uA is came from data sheet - Input current max 10mA. Please confirm.

In general, i need to build with ADS125H02 a RTD sensor measurement like the image below.

But with external current source ? I am right?

addition question: unfortunately, i am not understand your note in sentence below:

- A minimum of 0.9V is to be maintained over the differential REF input of ADS125H02. a 100 kOhm resistor should get a reasonable 1V.

Thank you

Ilya

Hello Ilya,

"I need use ADS125H02 because - with high impedance of 1.3M with 10uA the required source require is 13V, and your calculation of 10uA is came from data sheet - Input current max 10mA. Please confirm."

- The high impedance means even the little current would result in high voltage. as I said 5V ADC will not work, potential division will not work due to large impedance of the measured resistance. so you need a high voltage input ADC, ADS125H02 is high resolution high voltage ADC. The 10uA is the maximum reasonable current you can use if your ADC supply is the common 15V (leading to 13V in max). knowing the ADS125H02 can accept up to 38V supply, you might choose to raise your supply up to 30V for example and be able to use a current source of 20uA, working with larger voltage would give you better dynamic range. 

"In general, i need to build with ADS125H02 a RTD sensor measurement like the image below. But with external current source ? I am right?"

- right, you would need an external source, and your reference resistor should be connected to ground rather than to the supply as shown in the diagram above (I will try to post a circuit diagram later)

"addition question: unfortunately, i am not understand your note in sentence below: - A minimum of 0.9V is to be maintained over the differential REF input of ADS125H02. a 100 kOhm resistor should get a reasonable 1V."

- for ratiometric measurement, same excitation current has to pass through the measured resistor and the reference resistor. reference resistor voltage drop is connected to Refp and Refn pins, there is a minimum value requirement for that differential voltage found in the datasheet. this value is 0.9V. to account for tolerance in the current, let's target 1V as minimum for the design. given the current, you are able to calculate the required resistor.