Because of the Thanksgiving holiday in the U.S., TI E2E™ design support forum responses may be delayed from November 25 through December 2. Thank you for your patience.

  • State TI Thinks Resolved
  • Locked Locked
  • Replies 9 replies
  • Answers 3 answers
  • Subscribers 45 subscribers
  • Views 671 views
  • Users 0 members are here
Support feedback
Options
Options
Similar topics

This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

ADS1220EVM: Underdstand ADC PRo software data using ADS1220EVM kit

Minesh Darji
Prodigy 140 points
Part Number: ADS1220EVM
Other Parts Discussed in Thread: ADS1220, ADCPRO,

HI,

I'm using ADS1220 EV kit and ADC Pro Software for data colletion of my Weights scale measurement - analog input.I'm using load cells for my weight scale measurement application. I have tried to capture data, after putting 2KG weights on my load cells. Load cell's rated output is 1mV/V and it will give output form 0 to 3.3mV for 0 to 10KG weights.!(For Full bridge configuration). I have PDF data which I have shared it's reading by using half bridge configuration, whose output is 0 to 1.6mV for 0 to 12KG weights. How I can know that data is correct or not as per my weights?

Please advice about how to read and understand data which is coming out form ADC pro software.(by export data).

Waiting for your quick reply.

Thanks,

Minesh

mineshd@mpconsultants.org

M: +91 909990 76878(Whatsapp available).

ADS1220_ADCPro_Reading.pdf

  • 0
    TI__Guru** 113765 points

    Hi Minesh,

    Welcome to the E2E forum!  The ADS1220 has a bridge application circuit and discussion in the datasheet in section 9.2.3. 

    • Can you show me a diagram of how you are connecting the load cell to ADS1220EVM? 
    • Do you have a datasheet for the load cell you are using? 
    • Does the load cell happen to be a 3-wire load cell?
    • Can you send me the configuration register settings you are using?

    As the output voltage from the load cell is very small, it is difficult to get a measurable noise-free measurement without using the PGA gain of 128.  It appears that you are making the measurement as a single-ended measurement by the description in the data.  You should follow the design in the datasheet where the input is a differential input measurement.

    The ADS1220 uses a reference voltage and converts the output code relative to the voltage reference.  To maintain the lowest noise and drift, the same voltage used for bridge excitation should also be used as the reference voltage.  As the output code from the ADC is a reflection of a voltage, the value of a code in weight requires a conversion.  You can use the following conversion to determine the value of weight applied:

    This requires a calibration process to get a linear scale response.  You can approximate the Cfs and Czs from calculation based on the sensitivity of the load cell with respect to the excitation/reference voltage used and the gain applied.  For example, a 3.3V excitation on a 1mV/V load cell is a maximum deflection of 3.3mV.  Using a PGA gain of 128, the ADC full-scale range +/- 3.3V/128 or approximately +/- 25.8mV (or 51.6mV for 24-bits).  By using the repeatability formula in the ADS1220 datasheet (formula 42) you can determine the repeatability as a weight with respect to the ADC noise.

    Best regards,

    Bob B

  • 0 in reply to Bob Benjamin
    Prodigy 140 points

    Hi Bob B,

    Good Day.

    Thanks for your such nice explanation in reply. 

    • Can you show me a diagram of how you are connecting the load cell to ADS1220EVM?
    • ANSWER: See above DIagram

     

    • Do you have a datasheet for the load cell you are using? 

    Yes have please see attachemnt.(Load cell MPN : GML623A/GML624A - From GALOCE)GML623A Kitchen Scales Micro Load Cell 3kg.pdfGML624 Half Bridge Micro Load Cell 5kg_.pdf

    • Does the load cell happen to be a 3-wire load cell?

    Yes We are using Load cells using - which has 3 wire.

    • Can you send me the configuration register settings you are using?

    We are not doing anything for configuration register. Can you explain how to set it?

    FYI, We are setting PGA - 128 as Load cell's output is in mV. I'm also sahring data collected from ADC Pro software. Refer attachment.7536.ADS1220_ADCPro_Reading.pdf

  • 0 in reply to Minesh Darji
    TI__Guru** 113765 points

    Hi Minesh,

    I was hoping to see how you are actually connecting the load cells to the EVM and not just the block of your load cells.  Would it be possible for you to take a photo and attach to the thread? 

    Also, the configuration I was referring to was the EVM configuration in ADCPro.  Perhaps you can also send me some screen shots of ADCPro for the various tabs.  For me to try to understand the data you are sending I need a little more information on how the data is being captured.

    Best regards,

    Bob B

  • 0 in reply to Bob Benjamin
    Prodigy 140 points

    Hi Bob B,

    We are connecting four wires (using 4 half bridge load cells - as I hae shared Wiring diagram earlier.) to J1 of ADS1220EVm kit.(Refer below pic for yoru ref).

    EXC (+)  ----> VCC

    EXC (-)  ----> GND

    Sig (+)   ----> ANAlog Positive terminal(IN+)

    Sig (-)   ----> ANAlog Negative  terminal(IN-)

    We are using 3.3V Supply from ADS1220 EV kit for Load cell. so as per Load cell datasheet (GML624A) - Rated output is 0.8mV/V then 2*0.8mv/V(=1.6mv/V). 

    According to tihs, for 3.3V supply, Load cell's max output will be 5.28mV.

    Refer below pic for ADC PRo setting for 0.2mV analog input to AFE IN+& IN-. (We use 128 Gain as our output is mV) 

    later on I will share anohter new sheet for your more understanding.

    1. I just simply want to understnad -AFE kit's Accuracy for 1G and how I can test it and verify it?Guid me.

    2. I want to know/understnad relation between ADC pro datagraph to particluar analog input.

    eg. suppose I'm giving externally analog input from 0.1mV to 1mV(at IN+ & IN- pin of J1 connector) and getting data from ADC pro then I can justify Export data is really correct or not for 0.1mV to 1mV?

    How I know is there any noise/drift/AFE saturatuon is present or not.

    Please help to reply as early as possible.

    Waiting for your quick response.

    Regards,

    Minesh

  • 0 in reply to Minesh Darji
    TI__Guru** 113765 points

    Hi Minesh,

    Unfortunately none of the pictures are showing.  The forum does not accept drag and drop pictures or pictures pasted into the editor.  You must use the Insert/Edit Media icon to attach the picture(s).

    You want to 'simply' get an understanding of what you are doing, but unfortunately this is not as simple as you might think.  I have seen these types of load cells used as body weight scales.  Due to the wiring involved, noise is a particular issue as it can be easily picked up in the wiring.

    In your data I see a lot of negative conversion values which seems very odd unless you have connected the inputs incorrectly or are using the incorrect inputs in the mux selection.  You should connect the jumpers JP1 and JP2.  The bridge connection should then follow the silkscreen labels next to the J1 connector.

    In the ADS1220 ADCPro plugin you should select the proper input channels and reference voltage inputs along with the proper gain setting.  The data you have shown so far shows AIN0, but the bridge output is actually connected to AIN1 and AIN2.  AIN0 and AIN3 are connected to the excitation.  So the reference should be selected appropriately as well.

    Below I show the screen shot of the settings you should be using:

    BridgeRef

    Once you have these settings, you should see a more correct response to the load cell output. But we still don' know if there are other issues.  One code should be representative of +/- 3.3V/128/2^24. or 1 code is approximately equal to 3.07nV.  You would then need to convert the voltage to weight to verify the weight applied.

    Best regards,

    Bob B

  • 0 in reply to Bob Benjamin
    Prodigy 140 points

    Hi BOB,

    Thanks for your mail.

    I want to know difference between Low power mode and Normal power mode selection in ADC Pro software.

    ADC Pro software has facility to select either Low power mode and Normal power mode.

    Both power mode , has also different range for SPS - selection.

    Can you explanin about it? Why it is? or any spicific reason for it?

    I want to make sure that if I select PGA :128 then, I need to divide this PGA gain value (128) or need to multiply it (128 value) with ADC outcomes value? I believe it should be divide. Can you explain and justify about it?

    I want to get 1g accuracy using my load cell and TI's IC. You can recommended me any things that I need to take care for it eg temp control, Wifi, other paramaters which impact/affect to IC's Performance.

    Is there any recommended SNR for this EVM Kit. How I can check it and calculate theoritically and compare it with practical measurement using wheatstone bridge connection(all resistor - 10KOhm) or external 0.5mV analog input ?

    Waiting for your reply,

    Regards

    Minesh

  • 0 in reply to Minesh Darji
    TI__Guru** 113765 points

    Hi Minesh,

    The ADS1220 datasheet describes the various modes of operation more clearly.  As shown in the ADS1220EVM the modes are called Normal, Low Power and Turbo.  In the datasheet the EVM Low Power mode is actually duty-cycle mode.  For the EVM GUI it was decided to describe the duty-cycle mode as Low Power to show the result of using this mode.  Duty-cycle mode only makes sense when using continuous conversions.  The result for duty-cycle mode is to take one conversion and then wait 3 conversion periods before starting another conversion.

    For the ADS1220EVM GUI when the duty-cycle (or Low Power) mode is selected the Data Rate display becomes 1/4th the normal mode rate.  For example, if you select 20sps in Normal mode, in the duty-cycle (or Low Power) mode the throughput rate lowers to 5sps because the 1st conversion period at 20sps is approximately 50ms, then there are 3 conversion periods where the ADS1220 is idle (or in the low-power state) for 150ms.  So the total length of time is 200ms.  The end result is a new conversion that becomes available every 200ms (5sps).

    The duty-cycle mode allows the ADC to operate at lower throughput rates to conserve power.  You could also run the ADC in single-shot mode and use a micro timer to START a new conversion and accomplish the same task.  The advantage of the ADC duty-cycle mode may be a slight power advantage and certainly less coding effort as compared to using the micro.

    The value of the output code (LSB) for the ADS1220 is equal to +/- (VREF/GAIN) / 2^24.  For ADCPro to show the correct voltage value, then you would need to make sure that you are using the proper VREF voltage in the Current & Ref tab in the GUI.  If you were to use 5V as the reference voltage you would need to select the proper reference input from the Reference Select drop down menu, and also enter 5  in the VREF voltage box.  The value of the LSB in this case using a gain of 128 would be +/- (or 2) times (5V/128) / 2^24, or 4.7nV per code.

    The ADS1220EVM allows for a variety of different sensor types to be connected.  The PCB layout may not be optimal for any one type of sensor connection.  External noise sources will ultimately be the biggest challenge.  Cabling will act as antenna for pickup of power line-cycle noise (50/60 Hz) as well as noise sources such as EMI and RFI.  Using shielded wiring with proper termination is the best approach.  Unfortunately the ADS1220EVM was not designed for connecting shielded wiring as sensor input.  

    Also, the analog inputs should have proper input filtering.  The ADS1220EVM has minimal filtering on the analog inputs and reference.  Changing the 0 Ohms series resistances and adding the appropriate caps would be helpful to limit some of the noise by creating low pass filters.  I would suggest following the circuit path you are using from the ADS1220EVM user's guide schematic and adding appropriate resistors/capacitors for your circuit.  1k series resistors and 100nF differential caps for the input would be a good starting point.

    As far as calculation and prediction of outcome I would suggest looking at the application circuit in section 9.2.3 in the ADS1220 datasheet.  The theoretical capability is based on noise-free voltage of the ADC itself.  At a gain of 128 and 20sps the noise-free voltage is 410nV.  The repeatability of the load cell output will depend on the maximum voltage output for the bridge configuration in use.  For example, if the output voltage is 5mV for bridge with 5V excitation, then the repeatability (or noise-free scale resolution) would be the maximum output of the load cell divided by the full-scale output voltage times the noise-free ADC voltage.  If the load cell is 5kg at 5mV output voltage, 5kg/5mV * 410nV = 0.41g.  This value is the best outcome from the ADS1220, and does not reflect external noise.

    You could create a voltage divider using precision resistors to simulate the same output response of the bridge and then measure to see what you actually get and it should be very close to the theoretical.  There are also load cell simulators that are available that will create various output voltages that you can use to check basic performance.

    Best regards,

    Bob B

  • 0 in reply to Bob Benjamin
    Prodigy 140 points

    Hi BOB B

    Good Day.

    Please see attachment of photograph which shows our Load connection to AFE EVM Kit.

    Please correct/advice us, if we did it wrongly.

    Waiting for your reply.

    Thanks,

    Minesh

  • 0 in reply to Minesh Darji
    TI__Guru** 113765 points

    Hi Minesh,

    It appears that your wiring and hardware configuration are correct.  In the ADCPro ADS1220 plugin GUI make sure that the proper voltage reference (REF1) is selected and make sure that the voltage for the reference is also set to 5V.

    As I had mentioned in an earlier post, there is no input filtering for the analog and reference inputs.  For the reference you should add a 10nF cap at position C32 and C4.  You might also consider changing the resistor values for R4-R7 from 0 Ohm to something larger like 1k.

    I believe the biggest issue you will face is noise pickup in the wiring.

    Best regards,

    Bob B