How to properly provide a return current path (for an amplification circuit)

Thread Starter

Henry603

Joined Nov 19, 2018
69
I learned a lot in the past few weeks regarding basic circuit design but still have some problems understanding the following simple issue:
How do I provide a return current path for my instrumentation amplifier circuit?

I have passive sensors (pyranometers that consist of thermopiles and measure sun irradiation, sensor specs are attached at the bottom of this post) that output a low voltage signal so I need to amplify it using an instrumentation amplifier.
The output of the sensor (depending on the current irradiation) changes very very slowly.
My passive sensor has only two wires, one named HI and one named LO. I'm attaching the sensor differentially to the instrumentation amplifier.

I have chosen the instrumentation amplifier AD8237 for this task: Datasheet

The datasheet of this In-Amp says:
"When the source, such as a thermocouple, cannot provide a return current path, create one, as shown in the following figure."



This is how I currently plan to hook up the sensor to my instrumentation amplifier:



Question:
From what I have read so far I need to provide a return current path in my case. Can I just do it like shown in the circuit above (using the 10M Ohm resistors)?
I'm concerned about the impedance here as my sensors have very low impedance (10-100 Ohm, as you can see in the specs below) and the INA-Inputs are high impedance (100 MOhm as you can see in the datasheet posted above).
Will I get problems here if I use the 10 MOhm resistors on the INA-Inputs?
Would I need to add series resistors in order to increase the impedance here?
(because the current will return through the path of least impedance so do I need series resistance here in order to prevent that)?
Or am I mistaken?
How would you create the return path in this case?

Sensor specs:


Thank you very much for your help! :)
 

wayneh

Joined Sep 9, 2010
16,129
Question:
From what I have read so far I need to provide a return current path in my case. Can I just do it like shown in the circuit above (using the 10M Ohm resistors)?
@nsaspook beat me to it. One should do it, but two is OK.
I'm concerned about the impedance here as my sensors have very low impedance (10-100 Ohm, as you can see in the specs below) and the INA-Inputs are high impedance (100 MOhm as you can see in the datasheet posted above).
Will I get problems here if I use the 10 MOhm resistors on the INA-Inputs?
No, the sensor's low impedance will dominate setting the voltage between the + and - inputs.
 

Thread Starter

Henry603

Joined Nov 19, 2018
69
@nsaspook & @wayneh :
Thank you very much for your fast reply!
Also I did not find that Q&A site before you posted so thank you for guiding me into that direction, that is very useful! :)

Ok, so you both agree that I can just use one resistor to ground (on the inverting input?) to provide the return path.
Can I just take a 10 MOhm resistor (what is best practice here?) or should I better consider calculating a specific value depending on the INA?

Do I have to watch out for some impedance issues when using such a high value resistor here?
Could it harm my low voltage DC input signal?
By adding the resistor only on one input will I still be able to meassure fully differential with the INA?
 

nsaspook

Joined Aug 27, 2009
6,610
10 MOhm is a pretty standard measurement impedance so you shouldn't have problems with using that value as it looks like the chip designers designed the bias network for that value with a TC. Yes, you can still measure fully differential with one resistor.
 

Thread Starter

Henry603

Joined Nov 19, 2018
69
@nsaspook & @wayneh :

Thank you very much!
Does it make any difference whether i hook the resistor up on the inverting or non-inverting input?
What impacts would it have to use one resistor on every input (just curious).

Can you please explain to me why this works?
Because if I look at the input impedance of the instrumentation amplifier (about 100 MOhm), the sensors being very low impedance and adding the 10 MOhm to one input, I would expect the current to flow through the path of least impedance.
Wouldnt that cause issues here?
 

nsaspook

Joined Aug 27, 2009
6,610
I can only look at what the datasheet recommends as the best practice for bias resistor connections.

The bias current path is internal to the opamp input pins, it doesn't come from the sensor, it only passes through. The sensor current path is across the internal impedance of the two input pins.
 

Thread Starter

Henry603

Joined Nov 19, 2018
69
@nsaspook : Ok so I should be safe using one 10 MOhm.

As I checked the Q&A Link above you posted earlier I also saw the recommendation regarding driving an ADC with the instrumentation amplifier.
Now Im a little confused.

Could you please tell me whether you see any impedance issue in case I connect this:
instrumentation amplifier AD8237: Datasheet

to this ADC single ended:

ADC MCP3422: Datasheet

The input impedance of the ADC is on page 5 in the datasheet above.
In case I connect the amplifier output directly to the ADC input, could that overload my amplifier?
Or is there no real reason to be concerned about that?
 

nsaspook

Joined Aug 27, 2009
6,610
No real reason to be concerned about directly connecting to the ADC input if you keep the opamp output voltage in the safe range of the ADC input (VSS –0.4V to VDD+0.4V).
 

Thread Starter

Henry603

Joined Nov 19, 2018
69
Ok, thank you.

In what case could I get impedance issues between an amplifier and an ADC?
As from the datasheet of the amp I see that this amp can provide 4mA of current.
In case the input impedance of the ADC would be too low, I could damage my amp (by overloading it), is that right?
 

nsaspook

Joined Aug 27, 2009
6,610
Ok, thank you.

In what case could I get impedance issues between an amplifier and an ADC?
As from the datasheet of the amp I see that this amp can provide 4mA of current.
In case the input impedance of the ADC would be too low, I could damage my amp (by overloading it), is that right?
Even with a dead short to signal ground it's unlikely you will damage the opamp. At the speed you could sample the TC using that opamp with that ADC the impedance issues are negligible. It's time to stop worrying about non-issues.
 

Thread Starter

Henry603

Joined Nov 19, 2018
69
@nsaspook :

sorry I bother you again with this question but I just found this:
https://ez.analog.com/amplifiers/w/documents/1838/ad8237---recommendation-for-driving-an-adc
This is for the amplifier I want to use: instrumentation amplifier AD8237: Datasheet

They say due to the low power the amplifier might not be able to drive a delta sigma ADC.
I intend to use a delta sigma, the MCP3422: Datasheet

But they also say if one uses a low sample frequency it should be ok.
I want to use the ADC with 3.75SPS so that is pretty slow.
But on the other hand a delta sigma is oversampling the hell out of it so im not sure if that 'pretty slow' still applies.

Any thoughts on this?
Can I check some values in the datasheets to make sure the INA can drive the ADC?
Would really appreciate your help.
Thank you :)
 

nsaspook

Joined Aug 27, 2009
6,610
The opamp output current has to charge the ADC input sample capacitor. Higher source impedance (>~1000 ohm) increases the time needed to fully charge that sample and hold circuit. At 3.75SPS I don't think that's a problem unless you need the absolutely best sample performance and greater sample speed.
ADC datasheet:
Each input channel has a switched capacitor input structure. The internal sampling capacitor (3.2 pF for PGA = 1) is charged and discharged to process a conversion. The charging and discharging of the input sampling capacitor creates dynamic input currents at each input pin. The current is a function of the differential input voltages, and inversely proportional to the internal sampling capacitance, sampling frequency, and PGA setting
http://ww1.microchip.com/downloads/en/appnotes/00246a.pdf
Effects of Input Source Resistance A detailed model of the internal input sampling mechanism of a SAR ADC is shown in Figure 3. The critical values to pay attention to in this model are RS, CSAMPLE and RSWITCH. CSAMPLE models the summation of the capacitive array shown in Figure 2. Errors due to the pin capacitance and leakage are minimal. The internal switch resistance combines with the external source resistance and sample capacitor to form a R/C pair.
 

Thread Starter

Henry603

Joined Nov 19, 2018
69
The opamp output current has to charge the ADC input sample capacitor. Higher source impedance (>~1000 ohm) increases the time needed to fully charge that sample and hold circuit. At 3.75SPS I don't think that's a problem unless you need the absolutely best sample performance and greater sample speed.
ADC datasheet:


http://ww1.microchip.com/downloads/en/appnotes/00246a.pdf
The document you posted is focusing on SAR ADCs.
In my case, I use a delta sigma ADC (basically a 1-bit ADC attaining high resolution by heavily oversampling).
Thats why I was a bit concerned.
Or did you consider the SAR/delta sigma difference?

Thanks for your input! :)
 

nsaspook

Joined Aug 27, 2009
6,610
Oops, you are correct and that changes things slightly but the input drive requirements are similar as you have switched capacitor integrator input for the typical unbuffered delta sigma ADC modulator.

https://www.analog.com/en/analog-dialogue/articles/using-sigma-delta-converters-1.html
Q: What kind of a load does the input of sigma-delta converters present to my signal conditioning circuitry?

A: It depends on the converter. Some sigma-delta converters have a buffer at the input, in which case the input impedance is very high and loading is negligible. But in many cases the input is connected directly to the modulator of the converter. A switched-capacitor sigma-delta modulator will have a simplified equivalent circuit like that shown in the figure.


These impedances, as noted earlier, represent the average current flow into or out of the converters. However, they are not the impedances to consider when determining the maximum allowable output impedance of the A/D driver circuitry. Instead, one needs to consider the charging time of the capacitor, C, when S1 is closed. For dc applications the driver circuit impedance has only to be low enough so that the capacitor, C, will be charged to a value within the required accuracy before S1 is opened. The impedance will be a function of how long S1 is closed (proportional to the sampling rate), the capacitance, C and CEXT in parallel with the input (unless CEXT >> C). The table shows allowable values of external series resistance with fCLKIN = 10 MHz which will avoid gain error of 1 LSB of 20 bits-for various values of gain and external capacitance on the AD7710.
You might want to use a RC filter between the opamp and ADC (all types) to reduce the peak drive requirements from the source.
The requirement of exponential charging means that an op amp can not drive the switched capacitor input directly. When a capacitive load is switched onto the output of an op amp, the amplitude will momentarily drop. The op amp will try to correct the situation and in the process hits its slew rate limit (non linear response), which can cause the output to ring excessively. To remedy the situation, an RC filter with a short time constant can be interposed between the amplifier and the A/D input as shown in the figure. The (low) resistance isolates the amplifier from the switched capacitor, and the capacitance between the input and ground supplies or sinks most of the charge needed to charge up the switched capacitor.

This can also be used as a antialiasing filter, so two birds with one stone.
 

Thread Starter

Henry603

Joined Nov 19, 2018
69
Oops, you are correct and that changes things slightly but the input drive requirements are similar as you have switched capacitor integrator input for the typical unbuffered delta sigma ADC modulator.

https://www.analog.com/en/analog-dialogue/articles/using-sigma-delta-converters-1.html


You might want to use a RC filter between the opamp and ADC (all types) to reduce the peak drive requirements from the source.


This can also be used as a antialiasing filter, so two birds with one stone.
Ok, this got very interesting.
Thank you for the info provided, very useful.

1.
I was afraid I had to add another active component to my circuit in case this INA can not drive this ADC directly.
But seems that in case the INA has problems driving my ADC no additional buffer (like an active unity gain omp-amp) is necessary but a simple passive RC filter can solve that issue, do I understand this correctly?

2.
In that case, would an single pole RC with C=10µF and R=1k Ohm be ok?
Or what values would you suggest?

Thank you very much!
 

Thread Starter

Henry603

Joined Nov 19, 2018
69
ok w
Your values (0.01 second time constant) should work at 3.75SPS but you might want to reduce the capacitance so it will work with higher sample rates without modification.
http://www.trance-cat.com/electrical-circuit-calculators/en/rc-low-pass-filter-calculator.php

http://ww1.microchip.com/downloads/en/appnotes/00699b.pdf
ok will do more research on filters and time constants (you have some tuts or sites that you find useful regarding this?).

So by using this single pole passive RC filter I should be able to avoid an active component (buffer stage/omp-amp) in case my INA has problems driving this ADC directly right?
 

nsaspook

Joined Aug 27, 2009
6,610
ok w

ok will do more research on filters and time constants (you have some tuts or sites that you find useful regarding this?).

So by using this single pole passive RC filter I should be able to avoid an active component (buffer stage/omp-amp) in case my INA has problems driving this ADC directly right?
Yes. Even with the buffer stage the RC filter is a good idea as an anti-aliasing filter before the ADC.
 
Top