Potential Divider and ADS1115 inputs

Refer the sketch below. The Pins AI0 & AI1 will need to measure a +/-10V signal and hence a 2:1 divider is deployed. Like wise AI2 and AI3 will need to either measure +/-10V signal [or] 4-20mA signal. Just so that I display the proper mA values, I use a digital input on the connected NANO module. On the table in my office all of this work fine.

But in the field when they are connected to a proportional valve amplifier things go wrong and the values are no longer proper. I see the display values drop some 2 volts and also drift.

Due to the external amplifier configuration its possible that the 5V Ground gets linked with AI1 and AI3. Could this be an issue ?

Also when using differential inputs is it OK to use potential dividers as I have shown?

ADS1115_FloatingInputs.png

ADS1115_FloatingInputs.png

You may have forgotten to common the grounds with the circuit you are measuring?

Even in differential mode the inputs to the ADC must be strictly between the ADC's
rails, otherwise it will not give the right results, and may even be damaged. Differential
does not imply isolated/floating.

MarkT:
You may have forgotten to common the grounds with the circuit you are measuring?

Even in differential mode the inputs to the ADC must be strictly between the ADC's
rails, otherwise it will not give the right results, and may even be damaged. Differential
does not imply isolated/floating.

The very reason to use the differential mode is to measure negative voltages ... and the potential divider ensures the max input stays within the Vcc to Gnd which is 5V.

I am not too clear about the common ground aspect ?

Today we did a trial ... brought in a separate mains powered 5V power supply for the ADC and a three port isolation for the FB+ and FB- . And it worked. But its way too many additional components and cost.

The ADS1115 measures the voltage difference between inputs that must be positive with respect to the ADS1115 ground.

BOTH of the ADS1115 differential inputs must be POSITIVE..
EXACTLY WHAT ARE the voltages you are measuring (without the ADC) , measured on a DMM ?

4-20mA requires a 250 ohm resistor to convert the current to voltage.
The voltage is then measured across the resistor.

Typically such a configuration should have a high value resistor from the -Vin to sensor ground. Often this value is about 1 megaohm.

Its function is to insure the inputs don’t “float” out of the Common mode voltage range.

Regarding measuring negative inputs. While the inputs may go negative, the output cannot so measuring negative voltage is not in the cards.

If you need to measure negative voltages then you must bias the input in such a way that when the two inputs are connected together (i.e. 0 input voltage) the output is at 2.5 Vdc (or 1.6V if a 3.3v supply). Then when measuring negative voltages the output will remain above 0 Volts.
In this case:

  • Output voltages above 2.5 V are positive
  • Output voltage of 2.5V the inputs are at zero
  • Output voltages below 2.5 volts are negative

There are more complicated ways to do it using a
bipolar supply and an inverting op amp to convert the negative voltage to positive then using an analog input to measure the positive voltage then using code to convert it back to negative, but that
might be a bridge too far for you if you are even asking this question.

jremington:
The ADS1115 measures the voltage difference between inputs that must be positive with respect to the ADS1115 ground.

Some technical concepts hit you right on the dot when worded properly. This is one such !!

Are you by chance a teacher ?? ha ha !!

raschemmel:
BOTH of the ADS1115 differential inputs must be POSITIVE..
EXACTLY WHAT ARE the voltages you are measuring (without the ADC) , measured on a DMM ?

4-20mA requires a 250 ohm resistor to convert the current to voltage.
The voltage is then measured across the resistor.

When I measure with DMM removing the ADS115 , I read +4.7V and -4.8V on the extreme ends of the signal span. And in current mode its a clean +1 to +5V as expected.

raschemmel:
There are more complicated ways to do it using a
bipolar supply and an inverting op amp to convert the negative voltage to positive then using an analog input to measure the positive voltage then using code to convert it back to negative, but that
might be a bridge too far for you if you are even asking this question.

You sensed correctly on my electronic circuit prowess. I am a qualified mechanical engineer and MCU coding is a hobby !! For sure in grad school they did teach differential inputs and common mode voltages and I skipped it never knowing it will haunt me later !!

The solution with bipolar supplies would have been a good one to pursue but then its one more power supply and circuitry. Anyway will not rule it out.

@Mogaraghu can you answer a few q's please?

On the table in my office all of this work fine.

But in the field when they are connected to a proportional valve amplifier things go wrong and the values are no longer proper. I see the display values drop some 2 volts and also drift.

The ADS1115 - in fact (AFAIK) every ADC - measures voltages that are in dome way ground referenced.

Your schematic does not show how the grounds of the different inputs are connected. Can you use a meter and check the voltages on the input points WRT the ADS1115 0V GND?

..and list your results here?

Did you see on the data sheet the analog input voltages MUST lie within the bounds of the 0V - Vdd supply?

The 4 - 20mA input is often an issue as it need not be ground referenced.

"Are you by chance a teacher ?? ha ha !!"

We all are here on the forum but John is more compelled than most to reveal the inner workings of the universe. Wouldn't surprise me if he was a teacher. I'm just a technician. We both have
an affinity for op amps but he has more patience,
and probably more time . (he's probably retired
while I'm still working.)

Did you deliberately dodge this question?

"EXACTLY WHAT ARE the voltages you are measuring (without the ADC) , measured on a DMM ?"

raschemmel:
"Are you by chance a teacher ?? ha ha !!"

We all are here on the forum but John is more compelled than most to reveal the inner workings of the universe. Wouldn't surprise me if he was a teacher. I'm just a technician. We both have
an affinity for op amps but he has more patience,
and probably more time . (he's probably retired
while I'm still working.)

Did you deliberately dodge this question?

"EXACTLY WHAT ARE the voltages you are measuring (without the ADC) , measured on a DMM ?"

When I measure with DMM removing the ADS115 , I read +4.7V and -4.8V on the extreme ends of the signal span in the FB+ and FB- points . And in current mode its a clean +1 to +5V as expected.

What is generating these voltages ?

raschemmel:
What is generating these voltages ?

Its a amplifier meant to drive a solenoid valve called proportional valve. Basically it takes a command signal +/-10V or 4-20mA and delivers a current to a solenoid proportionately between 0 to about 1500mA.

A typical setup looks like the below image ...command is user generated and applied to pins D and E. In response the amplifier sends the position of the spool as a signal between F and C . The power ground and control ground are common for some brands like Eaton and for some others they are not. And for some Taiwan brands only God knows what.

Was a university lab technician, then lecturer teaching analog & digital electronics, microprocessors (6802,6809, 68000 etc) etc etc.
Maybe thats why I have the attitude "if its not right its wrong."
When you do retire raschemmel you will learn it means you have a lot LESS free time!

Are sending the control signal ir are you trying to hack it ?

"
Was a university lab technician, then lecturer teaching analog & digital electronics, microprocessors (6802,6809, 68000 etc) etc etc.
Maybe thats why I have the attitude "if its not right its wrong."
When you do retire raschemmel you will learn it means you have a lot LESS free time!"

I better not retire yet then. I barely have enough
free time now.

I assume this is a typo:
"
Basically it takes a command signal +/-10V or 4-20mA and delivers a current to a solenoid proportionately between 0 to about 1500mA"

I think you meant to say +/-5V OR 0 to 10V.
(both have 10V range but one goes negative and the other doesn't.)

** (because "+/-10V" is how engineers describe a signal**
with a range from -10V TO +10V (A SPAN OF 20V,NOT 10V)

"Refer the sketch below. The Pins AI0 & AI1 will need to measure a +/-10V signal and hence a 2:1 divider is deployed. Like wise AI2 and AI3 will need to either measure +/-10V signal [or] 4-20mA signal. Just so that I display the proper mA values, I use a digital input on the connected NANO module. On the table in my office all of this work fine.

But in the field when they are connected to a proportional valve amplifier things go wrong and the values are no longer proper. I see the display values drop some 2 volts and also drift.

Due to the external amplifier configuration its possible that the 5V Ground gets linked with AI1 and AI3. Could this be an issue ? (It won't work if they aren't 'linked' as you put it. However THEY ARE NOT LINKED
IN YOUR SCHEMATIC.

Also when using differential inputs is it OK to use potential dividers as I have shown? "

You need to be able to read -5V before you can
say you can read +/-5V.
It is easier if you can use 4-20mA instead of 0 to 10V.
In order to do that without using 4-20mA , you
have to shift the +/-5V to 0 to +10V, then you
can use your voltage divider. If I understand
your post correctly. If you can't use 4-20mA then you will have to SUM the +/-5V signal with +5V
to get 0 to 10V (not the sane as +/-5V because
0 to 10V NEVER goes negative. You can then use
/2 voltage dividers to convert 0 to 10V to 0 to +5V
which can be read differentially with the ADS1115.

Looking at your OP, it looks like all you are missing is a GND connection on FB- & E.
Connect those two to the arduino and ADS115 GND
and it should work if the signal really is 0 to 10V and NOT +/-5V.
Is the relay for selecting 4-20mA mode ?

To repeat what was stated earlier, you cannot measure a negative voltage with respect to the GND terminal of the ADS1115.

If your readings of +4.7 and -4.8 are referenced to GND, you are exceeding the maximum allowable negative input voltage by -4.5 volts and have likely blown the ESD diodes in the device, shorting the input of the PGA.