ADS1282: data conversion for ADS1282

sreelatha p18

Part Number: ADS1282

Hello,

I am using ADS1282 for converting milliVolt signals. It is used in the default config, with REFs as +2.5V and -2.5V and is in the continuous mode, at 1000Sps data rate.

I need help with two issues--

(i) I gave a 20 mV input signal and got a dataset which has EBB34817.. as output and with zeroes in between.

(ii) When I gave 200 mV signal, in addition to the zeroes coming in between, I got two sets of different values at the output. (9636E191, 1B458580 )

I tried giving other signal values also, but one of the above patterns result.

As I am new to this, any help would be greatly appreciated.

The sample data files are attached

Request help

Thanks in advance

Regards

over 7 years ago

0 Christopher Hall over 7 years ago

TI__Mastermind 44760 points

Hi Sreelatha,

Welcome to the TI E2E forums!

Unfortunately, I don't see your data files. Would you mind trying to send them again?
(You might need to go into the advanced editor and click the "Insert File" button to attach the files - I haven't had the best luck with dragging and dropping files into the E2E textbox).

You're results indeed don't seem to correlative to the input voltages that you are providing. There are many reasons why that might be the case, including:
- If you read data before the /DRDY signal goes low, the output result maybe invalid.
- If the reference voltage is not enabled or is not provided the results won't make any sense.
- If the input voltages violate the PGA's common-mode voltage requirements, you may see a large gain error.
- SPI communication errors, or post-processing errors when combining data bytes into a single signed integer result.

To help me determine what might be the problem, which PGA gain are you using? Also, are you correctly reading from and writing to the device registers?

Even if you're SPI communication appears to be working, I might still recommend using an oscilloscope or logic analyzer just to double check that the results you're seeing from your processor are the same conversion results that are being transmitted via SPI.

Best regards,
Chris

0 sreelatha p18 over 7 years ago in reply to Christopher Hall

Prodigy 50 points

Hello Christopher,
Thanks for the fast reply.
Let me restate the issues and give clarifications wrt the circuit.

-Data is read on /DRDY going low only.
-Reference voltages of +2.5 and -2.5 are also provided with power on itself.
-As the input voltage is low, I expect no common-mode violation.
-Regarding the SPI communication errors/post-processing errors when combining data bytes into a single signed integer result, which you have indicated-- I am not sure how to look for it. Request some more help in this regard.
-The device register read/write also seem to be ok. Is there anything specific that I can look for?

The circuit has a front end sensor followed by an opamp, the output of which is connected to the ADS1282. Only DC signals are given for the tests. The AINN1 and AINN2 inputs are grounded. The incoming signal is fed to AINP1. The ADS1282 is configured for PGA gain of 1.

Will try to attach the data files once again.
Thanks in advance
Warm regards
sreelatha

0 sreelatha p18 over 7 years ago in reply to sreelatha p18

Prodigy 50 points

Fullscreen 20mV_ADC.txt Download

Fullscreen 200mV_ADC.txt Download

0 Christopher Hall over 7 years ago in reply to sreelatha p18

TI__Mastermind 44760 points

Hi Sreelatha,

Do you know what your command sequence is for reading data?...Do you use RDATAC mode when reading data or do you send the RDATA command? Also what is you SCLK frequency?

Please see the data conversions attached...

Are the 200mV and 20mV signals voltages applied to the ADC, or are these the input voltages prior to the op-amp?
According to the voltages calculated in the attached Excel file, the ADS1282 is measuring approximately 0.5 and -0.39V in most cases; however, I did observe some other behaviors...

There are several conversion results that appear to be outliers. Perhaps the SPI data is getting out of alignment. Make sure when you send command bytes you ignore the returned data (read it but throw it away) since the ADS1282 doesn't utilize the full-duplex capabilities of SPI
In the 20mV data file, there are sequences of 62 conversion results that return all zeros. This behavior is typical for the ADS1282 when ADC conversions are resynchronized and the FIR filter is settling (new data is available on the 63rd conversion cycle).
Finally, in the 20mV data file, there were also a few conversion results that show duplicated data outputs from conversion to conversion. Statistically, this is unlikely and might indicate that you are reading the same conversion result multiple times. In those cases you might want to check to see if perhaps the /DRDY interrupt is incorrectly triggering.

My best recommendation would be to get an oscilloscope or logic analyzer capture of the SPI communication. If you able to do this, please feel free to share it with us and we can help you interpret it and troubleshoot any issues we see.

Best regards,
Chris

0 sreelatha p18 over 7 years ago in reply to Christopher Hall

Prodigy 50 points

Hello Christopher,
Thanks for the details.
I will probe and send the results soon.
I am sorry that I could not open the attached excel file, which I understand gives the calculation. Request to please resend the same.
Thanks in advance
Warm regards
sreelatha

0 sreelatha p18 over 7 years ago in reply to Christopher Hall

Prodigy 50 points

Dear Christopher,
Further to my earlier reply, I would like to state the following.
1. The ADS1282 is used in power ON default mode. 1000SPS, Channel1, PGA gain 1 and continuous conversion mode.
2. The fclock to ADS1282 is a 2 MHz CMOS crystal oscillator, instead of 4MHz. So we get 500SPS rate on Power ON instead of 1000SPS.
3. The ADS1282 DOUT is read when the DRDY goes low. This is done by giving the DRDY to a low going trigger interrupt. As soon as the DSP SMV320C6701 gets the interrupt, it sends 32 clocks at 320KHz rate. (SCLK rate), in SPI mode of the McBSP0 port. So we read the DRR register input for the 32 c;locks sent. We have correlated the oscilloscope wave form and the DRR value several times. They are same.
4. But the ADS1282 input voltage and the output code do not match. The 20mV and the 200mV in the files that I have sent are measured wrt to GND at OP Amp output. From the OP AMP out put the signal is fed to the ADS1282 input AINP1 and AINP2 through a 100Ohm resistor. The AINN1 and AINN2 of the ADS1282 are grounded. So we expect the same voltage at the input of ADS1282 also.
We are still clue less about the DOUT code mismatch.
Thanks and Regards,

0 Christopher Hall over 7 years ago in reply to sreelatha p18

TI__Mastermind 44760 points

Hi Sreelatha,

Try this link to see if this works better.../cfs-file/__key/communityserver-discussions-components-files/73/ADS1282-data-conversions.xlsx

Since you are directly clocking out the ADC data, have you made sure to send the RDATAC command before attempting to read data? If you haven't, then you will need to send the RDATA command each time you want to read data.

Best regards,
Chris

Data converters

Data converters forum

ADS1282: data conversion for ADS1282