ADS1261: 2 wire excitation mode problem

Hi Elia Pellegrino,

If you slow down the ADC data rate, does the situation improve?

In the image where you are using a gain of 32, the average value you are measuring looks to be around -0.65785V (not sure? the x-axis is not labeled) and the signal does not appear to be saturating or swinging wildly from positive to negative. Would you consider this to be "good" data?

Can you provide a complete schematic so we can at least see the ADC circuitry, the input circuitry between the ADC and the terminal blocks, and the ACX circuitry?

Can you also provide the register settings?

-Bryan

Hi Bryan,

If you slow down the ADC data rate, does the situation improve?

No, I have already tried to slow it down to 14.4 ksamp/s, but the situation is the same.

In the image where you are using a gain of 32, the average value you are measuring looks to be around -0.65785V (not sure? the x-axis is not labeled) and the signal does not appear to be saturating or swinging wildly from positive to negative. Would you consider this to be "good" data?

No, I cannot consider it as good data, as the values are all the same and they cannot change if apply some load to the load-cell. Yes sorry for the missing label, y-axis is Volt, while x-axis second.

Can you also provide the register settings?

For sure, this is our register settings:

Can you provide a complete schematic so we can at least see the ADC circuitry, the input circuitry between the ADC and the terminal blocks, and the ACX circuitry?

Yes, find attached a simplified schematics of our board:
ADS1261 Report.pdf

Moreover, I do not know how, but it seems that after using the ACX mode in this way, the ADS seems to not be responding anymore to the SPI command: can it be that it has been damaged by an high voltage outside its operating condition?

Thank you very much for your help,

Elia

Hi Elia Pellegrino,

Thanks for providing the requested information

None of the supply nets have voltages, they just say e.g AVDD, AVSS, etc. Can you provide the ADC supply voltage levels as well as the excitation voltage levels (VSG+, VSG-)?

Can you also try slowing the ADC down much more, like 100SPS? I understand this might not be how you want to operate the ADC but this is just a troubleshooting step

-Bryan

Hi Bryan,

Yes you are right sorry. I cannot show you the power input stage, but:

• AVDD: 5 V
• AVSS: GND
• DVDD: 3.3V
• VSG+ / VSG- comes out from a switch and go directly for "powering" the load cell (as you can see in "Input Stage / Load cell"). When using the 2 wire ac mode, we switch it manually to EXC+ / EXC-

Can you also try slowing the ADC down much more, like 100SPS? I understand this might not be how you want to operate the ADC but this is just a troubleshooting step

Yes, I can, but I am bit afraid of breaking another ADS, as it seems it already happened with 2 of them. I will have a try on friday and I keep you posted.

In the meantime, do you have any clue of what can be the problem?

Thanks,

Elia

Hi Elia Pellegrino,

You mentioned "high voltage" in one of your previous posts. What is the voltage level you are referring to, and where does this come from in your system?

Can you tell me what is the actual voltage level for VSG+/VSG-?

Why do you have AVSS and GND nets if they are the same potential? How are these connected on the PCB?

-Bryan 

Hi Bryan,

You mentioned "high voltage" in one of your previous posts. What is the voltage level you are referring to, and where does this come from in your system?

Can you tell me what is the actual voltage level for VSG+/VSG-?

When using the 2 wire excitation mode, we switch the switch S1B, so that VSG+/VSG- is equal to EXC+/EXC-. The output of this are shown in the figure above. So, we are basically exciting our load cell with a differential voltage of VSG+ = 5 / VSG- = -5, because the difference of (VSG+ - VSG-) is given by a difference of a (square wave - (-1*square wave)), as you can see in the figure above. This is where my concern lives: given the fact that the load cell is basically (in simple term) a resistor, we are giving as analog input a voltage that is below AVSS, when the excitation voltage is -5V. Am I missing some points on how 2 wire excitation works?

Why do you have AVSS and GND nets if they are the same potential? How are these connected on the PCB?

They are connected together through a net tie, so they are at the same potential.

Thanks,

Elia

Hi Elia Pellegrino,

The applied voltage is not negative with respect to ground during AC excitation, it just has the opposite polarity. So you are applying 5V to the top of the bridge and 0V to the bottom of the bridge during the first phase, and then you are applying 5V to the bottom of the bridge and 0V to the top of the bridge during the second phase. So this should not be an issue. But when you mentioned "high voltage" it wasn't clear to me what the excitation voltage actually was

Have you tried slowing down the ADC? My originally thought is that you are seeing settling issues because the excitation polarity is switching so fast. The RC circuit you have on the ADC inputs needs to settle every time the excitation voltage switches, and 40kSPS might be too fast

-Bryan

Hi Bryan,

Yes, I tried but nothing change. And again, after some time of using with acx mode, the chip seems not to be responding any more

Hi Elia Pellegrino,

There is no obvious reason why the AC excitation circuit would cause the ADC to break, unless you are somehow overdriving the ADC pins. But what you have described so far does not indicate that that is the case

Are you able to get an ADS1261EVM so you can start with a known good board that already has a working AC excitation circuit? This will at least give you a place to start

Then you can wire in your controller to the EVM so you can check that your code is correct. You can also wire in your AC excitation circuitry (bypassing the circuitry on the EVM) to see if that works correctly. And so on, checking the individual components to see where the issue is

This sounds a bit like it could be a grounding issue, but without have the board in front of me it will be hard to diagnose remotely. So getting an EVM seems like the next best option

-Bryan

Hi Bryan,

Thanks, I will perform this tests during these days and I will let you know something!

Elia

Hi Elia Pellegrino,

Thanks for the update

-Bryan