myRIO ADC problem: lateral channel affacted by its previous channel

[Mengqiu]

Hi

I am using multiple ADC channels of myRIO now and facing an issue about it.

I am using 3 ADC channels of Port A: AI0, AI1 and AI2.

I connected AI0 to the midpoint of a potentiometer to get the voltage accoss it. And Connect nothing to AI1 and AI2.

The ADC works very well. I can get the valtage of the potentiometer perfectly through AI0.

However, when I changed the value of potentiometer, the ADC results of AI1 and AI2 were changed too, from 0.1V to 0.8V roughly. Please noted that I connected nothing to those two ports.

Is that true? the ADC port AI1 and AI2 or even AI3 (PORT A) are affacted by AI0?

When I connected AI1 to a Anolog output senser, the result was also affacted by the potentiometer which was connected to the AI0.

When I connected AI1 to power, it was out of influence. 

Have anyone ever had such problem? Can anyone help me to solve it?

Thanks,

Mengqiu

[altenbach]

This is normal and not a problem. If nothing is connected, the reading is random because there is residual charge left that has nowhere to go because the internal circuits are high impedance. Start reading about bias resistors.

Why are you worried about signals that are not connected? Just ignore their reading! Else terminate these unused connectors.

[Mengqiu]

Hi Knight of NI,

The main problem is when I connect the AI1 to my analog sensor, the ADC result of AI1 is affacted by the circuits connected with AI0. I got an unstable digital result (AI1) when I changing the potentiometer connected with AI0.

It is weird.

Then I disconnect the sensor, still find the digital result of AI1 is changed a lot by the potentiometer connected with AI0. It is not random, I can obviously find there is sort of coupling issue between those channels.

Is that because of some setup issues? or it is a normal issue that lateral ADC channel affacted by its previous channel?

[DigDoug]

If you connect your potentiometer to AI1 rather than AI0 does AI0 then show the behavior you described?

To the values of AI1 and AI2 increase or decrease as the resistance of the potentiometer increases? Does it seem random?

I found a KnowledgeBase article for you about troubleshooting this type of thing. It is for DAQ, but the concepts apply to your myRIO as well.

Troubleshooting Unexpected Voltages, Floating, or Crosstalk on Analog Input Channels

http://digital.ni.com/public.nsf/allkb/B9BCDFD960C06B9186256A37007490CD

Also, if you have not yet reviewed the white paper linked by altenbach above, there is a lot of good information about noise, crosstalk, etc. in analog signals. Please review these two documents.

[Mengqiu]

---

@DigDoug wrote:

If you connect your potentiometer to AI1 rather than AI0 does AI0 then show the behavior you described?

To the values of AI1 and AI2 increase or decrease as the resistance of the potentiometer increases? Does it seem random?

I found a KnowledgeBase article for you about troubleshooting this type of thing. It is for DAQ, but the concepts apply to your myRIO as well.

Troubleshooting Unexpected Voltages, Floating, or Crosstalk on Analog Input Channels

http://digital.ni.com/public.nsf/allkb/B9BCDFD960C06B9186256A37007490CD

Also, if you have not yet reviewed the white paper linked by altenbach above, there is a lot of good information about noise, crosstalk, etc. in analog signals. Please review these two documents.

---

Hi DigDoug,

Thank you so much for your reply.

When I connect potentiometer to AI1 rather than AI0, AI0 is not affected by potentiometer. It is quite interesting that lateral AIs is influenced by its previous AI channel.

To the values of AI1 and AI2 increase or decrease, it is not random, it increases when AI0 increases, decreases when AI0 decreases, I can even get a equation for that.

I will read those materials, and try to find the problem.

Thanks for help.

[Mengqiu]

Hi guys,

I am consulting this problem with NI now, see more detail description:

I have connected both to the same ground: AGND. Please note that there is an interesting thing which I mentioned,

The lateral channels of AI (Analog Input) are affected by the previous AI channel (same Port e.g. PORT A). However the previous channel is not affected by the lateral channel.

For example, I connect the potentiometer to AI1, connect my sensor to AI2. The result of AI2 is changed by tuning the potentiometer.

See below, when AI1 changes from 4.96V to 0.004V, the AI2 is changed from 3.39V to 3.35V (I do not change the sensor connected with AI2), such change is stable.

When there is no connection, the change of AI2 will be more than 1V (stable), where does it come from?

Then, I switch those two channels, connect potentiometer to AI2, and my sensor to AI1, then the result is quite good, see below (with some noise less than 0.01)

AI1 is not changed by AI2, when AI1 is no connection.

So I am quite confused about this, and now I am using different PORTs (PORTA and PORTB) which works fine.

The changes are shown in attached pictures.

Best regards,

Mengqiu

[DigDoug]

This behavior certainly seems odd. I am not very familiar with myRIO, but I will do some digging and see if I can find an answer for you. 

[Mengqiu · Accepted solution]

This is the answer from NI support:

Please note that the small variation you notice when both of your channels are connected can be due to several factors such as poor wire shielding, EMC (electromagnetic coupling) and etc.  

In regards to a floating channel terminal, random or correlated variation is normal and due to internal circuitry of myRIO and should not be a cause of concern.

Although, I am still very confused.

[jjlee128]

I noticed that too, there is only 1 ADC being shared among all the analog input channels, so it uses a multiplexer to cycle between the various channels; I think its probably switching noise.

[charlesy]

Hello jjlee128,

You are correct that this is an artifact of the multiplexed input to the ADC. The myRIO API scans AI channels in your scanlist in numerical order (AI0, AI1, AI2, ...), at a maximum rate. As the multiplexer switches, it injects charge from the ADC input on to the next channel. If you have a high impedance on the channel (like an open, or a resistor network larger than about 3 kOhm -- see "Recommended Source Impedance" in the specs), there is not enough time to drain away the charge and the victim channel measurement is contaminated from the previous channel. 

You can mitigate this effect by driving the inputs with lower source impedances that can absorb the charge kick. That's why measurements of power rails or ground are immune. You can also buffer high impedance networks or sensors. That's effectively what happens on the two MSP differential analog inputs on a myRIO-1900. While they drive into the same multiplexed ADC, each of those channels has a dedicated instrumentation amplifier that can overdrive the charge injection.

I hope this is helpful.

Charles Y.

National Instruments