ADS131M08EVM: The connection of the 2 loadcells.

Hello Alex,

Thank you for your post.

There is a 5-V supply rail that is used to generate the 3.3-V LDO output for AVDD. You could solder a jumper wire to R63 and use this 5-V supply to power your load cells. Remember to also share a connection to any GND test point on the EVM.

Do you have a part number for the load cells you are using?

Regarding the input connections, you have the option to use pins 1 and 3 of the terminal blocks or pins 1 and 5 of the corresponding JPx headers. Without jumpers installed, these inputs connect to an R-C differential low-pass filter and then to the AINxP and AINxN input pins of each channel. The EVM schematic can be found towards the end of the User Guide: https://www.ti.com/tool/ADS131M08EVM#tech-docs.

Ryan many thanks for your reply, it is a big help for me to move forward.

Im more software developer and in past I did the project with my father who was a hardware developer.

Sadly to mention my father passed away and I do look for any technical help now.

It is TEDEA Huntleigh, model No. 1130 single point loadcell.

http://www.vishaypg.com/docs/12011/1130.pdf

Emax=7kg,Emin=0 kg, Vmin=0.5 g

NMI=TC5703

it is 2 mv/V, so at 7 kg its max output then will be 10 mV.

I do look for calibration to be  7 kg by 5g, what makes capacity of 1400 divisions.

This will make 1 division about 0.007mV. How realistic is to have a stable readings?

I bought the EVM for tests but if results will be well, then I will design own board and then may use say

12V power, with 24mV max output as result and 0.017 mV for 1 division for more stable readings, I think.

ADS131M08 is 24 bits but more likely last 4-6 bits (if not more) will be a heavy noise as I know, so the ADC result

then may be filtered to 16-18 bits.

Ryan, Im a bit stack to get 5V point.

My J12 is set left. I use on board ground, but cant get 5V (or 3.3V) from J12 or R63 point.

Alex

Hello, Ryan.

Thank you.

I connected all and I must admit finally that the 5V, which comes from board needs a filtering.

I sampled one loadcell connected, used gain 128 and less and compared min and max peaks difference

with 2 cases:
when I use 5V power at -Exc and +Exc
and when I use with both ground ends from board connected to -Exc and +Exc.

(in such case we do see clear noise from interference and also a "warm" noise from the loadcell).

The difference in quants between min and max peaks is about 15 times between 2 cases. (152 and 10).

I conclude that the 5V needs to be filtered.

I guess there are lot of solutions for such filter.
Say, increase up to 7V, filter and low down to 6V with some good cut of the pulses.

(I guess some terms might be wrong translated by me sorry, English is not my native).

If you had any previous experience - could you advice some hardware solutions with schemes so we may

build it here and try ? Not sure there is anything ready yet?

I need to run some physical experiments, where 2 loadcells are involved

but I cant do that as the signal is too noisy to get any proper outcome.

Alex

Hi Alex,

Thanks for the update.

Filtering the excitation voltage can certainly help to remove some of the high frequency noise from the sensor output. Ideally, a load cell would be excited by the same voltage as the ADC reference voltage, so that any noise on the excitation signal would be cancelled. This is known as a ratiometric configuration.

After discussing this topic with a colleague, it may be advisable to consider another device from our portfolio which is directly suited to support ratiometric load cell measurements. A couple examples would be the ADS1235 and ADS1261. I suggest taking a look at some of the resources below, as well as the ADS1235EVM and ADS1261EVM for example application circuits.

E2E Post illustrating a 6-wire load cell connection: https://e2e.ti.com/support/data-converters/f/73/t/875545

TI Reference Design: https://www.ti.com/tool/TIPD188

Best regards,

Hello, Ryan,

thank you for your reply, time and suggestion.

As I see ADS1235 is better.

Could we please discuss the possibility for me to return ADS131M08EVM and get ADS1235EVM board?

Let me know if the details should come via email and not here, for example.

Alex

Hello, Ryan, 

Thank you for your help again, I do much appreciate.

May I ask if additionally to ADS126X options there might be a chip exists which

supports all same ratio metric options but also have simultaneous sampling channels?

Ideally would be 4 already bipolar channels, like ADS131M08 has, which I bought first.

I need to connect  at Step 1, - 2 loadcells same time and sample these simultaneous.

ADS1235 cant do that, only 2nd after 1st. Simultaneous sampling is crucial for our physical

experiments here.

Also, at Step 2, - I need to connect single loadcell and 3 channels accelerometer (ADXL-337).

It would be nice to have same board for both experiments (Steps).

If no simultaneous chip exists with ratio metric then probably I will have to go then for ADS126XEVM.

I do apologies for my previous quick reply about ADS1235. I already got an email from 

Christopher Hall but will wait now for your reply regarding simultaneous sampling

(like ADS131M08 was) but with ratio metric support.

Alex