ADS9120: Data corruption issue while using ADS9120 in the daisy chain mode

Shinu.V.Mathew

Part Number: ADS9120

Hi Team,

My customer is working on one of the medical equipment and facing data corruption issue while using ADS9120 in the daisy chain mode (two ADC’s), below is the snapshot of the schematics for your reference:

And below is the data which they get out of the the device.

As you can observe they are able to plot the I-Data correctly but Q-data is corrupted. They have tapped the Q-data with the logic analyzer and plotted the data and see the same pattern as shown above for 3-Q.

This confirms that the Q channel is not giving the correct data output, and from above picture the pattern which is repetitive (peak are folding back). They have probed the input of the ADC and that looks perfectly fine.

Can you please help address this issue.

Regards, S Mathew.

over 3 years ago

0 Keith Nicholas over 3 years ago

TI__Mastermind 46360 points

Hello Shinu,

I am going to close this request since it is a duplicate. Please refer to this post and response:

https://e2e.ti.com/support/data-converters-group/data-converters---internal/f/data-converters---internal-forum/1106860/ads9120-2nd-adc-doesn-t-give-correct-output

I think this may be a timing issue with the data transfer. Since the signal at the inputs looks good, and the other ADC correctly converts data using the same reference, the SPI timing with daisy-chain is the most likely suspect.

It would be good if you could provide scope pictures for a full SPI frame, including CONVST, /CS, SDO-1 for each ADC, and SCLK. Specifically, what is the conversion rate and the SCLK frequency used?

When using daisy-chain, the maximum SCLK frequency should be around 60MHz to account for setup and hold times between SDO-1 on the Q-channel and SDI on the I-channel. This will limit the maximum data rate to about 1.25MSPS for two ADCs in daisy-chain.

Another possibility is that the customer is monitoring the RVS of the I-channel, and when it transitions low-to-high, transferring the data in Zone 1 mode. Since both data converters will take a different amount of time to complete the conversion, this could also result in corrupted readings for the Q-channel. Do you know which data transfer zone the customer is using? Providing scope plots for a full SPI transfer frame will answer these questions.

Regards,
Keith Nicholas
Precision ADC Applications

0 Shinu.V.Mathew over 3 years ago in reply to Keith Nicholas

TI__Genius 10300 points

Thanks Keith.

0 Keith Nicholas over 3 years ago in reply to Shinu.V.Mathew

TI__Mastermind 46360 points

You are welcome.
Keith

Data converters

Data converters forum

ADS9120: Data corruption issue while using ADS9120 in the daisy chain mode