ADS1234: Minimum circuity to get it working

Anthony Shivakumar

Prodigy 30 points

Part Number: ADS1234

Attached is a schematic using the ADS1234. Want to double check if this circuit would work.

over 2 years ago

+1 Bob Benjamin over 2 years ago

TI__Guru** 113765 points

Hi Anthony,

Welcome to the E2E forum! I don't see anything in the schematic that would be a significant issue. However I would like to make a few recommendations.

SCLK should normally dwell low, so instead of using a pullup to 3.3V you may consider using a pulldown resistor to GND instead.
To help reduce noise on the reference input, place a cap across the REFP/REFN inputs. This cap could be 100nF.
Add RC filters to the inputs for antialiasing and noise.
Make sure that all digital input pins are set to a specific logic level and that no inputs are left floating or in a float state from the micro.

Best regards,

Bob B

0 Anthony Shivakumar over 2 years ago in reply to Bob Benjamin

Prodigy 30 points

Hi Bob,

After reading through your suggestions, why should SCLK be pulled low? For the I2C bus is not SCLK pulled up?

Based on the reference design https://www.ti.com/lit/ug/sbau120b/sbau120b.pdf?ts=1632864262148 I did not see any RC filter in the input, hence I have not added it.

Attached is the updated schematic. 6082.ADS1234.pdf

0 Bob Benjamin over 2 years ago in reply to Anthony Shivakumar

TI__Guru** 113765 points

Hi Anthony,

Note the reference design has placeholders for components that would add analog input filtering. I would suggest you do the same for your design and adjust the filter values as needed for the proper filtering response for your system.

As to the communication I think there is a misunderstanding. The ADS1234 uses an SPI compatible communication protocol and not I2C. If you intend to use I2C with the ADS1234 it will not work.

Best regards,

Bob B

0 Anthony Shivakumar over 2 years ago in reply to Bob Benjamin

Prodigy 30 points

Thank you so much for your reply Bob. Now it makes sense. There seems to be an error in the datasheet then. Under 'Features' on the 1st page it says 'Two-wire serial interface'.

https://www.ti.com/lit/ds/symlink/ads1234.pdf?ts=1632933452188&ref_url=https%253A%252F%252Fwww.ti.com%252Fproduct%252FADS1234%253FkeyMatch%253DADS1234

Since the naming convention on the pin did not specify MISO, MOSI, SCLK, CS the standard SPI terminology I had mistaken the protocol. (On the home page it says SPI.)

Thank you so much for your help. Regarding the pcb design, should I make the trace length for +- equal? Is this important for the accuracy of measurement. I can consider these traces as differential signals and match their length.

0 Bob Benjamin over 2 years ago in reply to Anthony Shivakumar

TI__Guru** 113765 points

Hi Anthony,

The trace length should be similar but it is not critical that the lengths are matched. The input is considered slow moving so high speed considerations are not a concern.

What would be more beneficial is having a solid ground under the analog signal traces. Your bottom layer has a solid block for +5V, but the signal traces run over the top of the +5V block and not ground. You might consider swapping the +5V and GND on the connectors so that the signal traces have an undisturbed ground all the way to the ADC.

Best regards,

Bob B

0 Anthony Shivakumar over 2 years ago in reply to Bob Benjamin

Prodigy 30 points

Thank you for your suggestions. I have made the changes. Removed the resistors on the SPI bus, swapped 5V and GND, redefined the polygon and made the lengths match (more so for the visual appeal).

2330.ADS1234.pdf

0 Bob Benjamin over 2 years ago in reply to Anthony Shivakumar

TI__Guru** 113765 points

Hi Anthony,

The layout is much better. I still think you should at least provide device footprints for an analog filter on the inputs, but that is up to you. If you have noise issues you will have to add them later.

Best regards,

Bob B

0 Anthony Shivakumar over 2 years ago in reply to Bob Benjamin

Prodigy 30 points

Thank you Bob.I was not sure what you meant by input filter. After reading your answer on another forum I get it. https://e2e.ti.com/support/data-converters-group/data-converters/f/data-converters-forum/752994/ads1232-input-anti-aliasing-filter-problem

I have added the input RC filter. Would this work?

7484.ADS1234.pdf

0 Bob Benjamin over 2 years ago in reply to Anthony Shivakumar

TI__Guru** 113765 points

Hi Anthony,

I think the values chosen are a good starting point. As far as the implementation you are almost there. Currently your filter doesn't do anything, so move the net labels in your schematic to the other side next to the capacitor.

I would also suggest moving the capacitor closer to the analog inputs.

Also, make sure that the resistors stay within the ground area. So the resistors should be placed above the +5V power block.

Best regards,

Bob B

0 Anthony Shivakumar over 2 years ago in reply to Bob Benjamin

Prodigy 30 points

Thank you so much Mr. Bob. I have made the changes.8306.ADS1234.pdf

Would this now activate the filtering?

Regards,

Anthony

0 Bob Benjamin over 2 years ago in reply to Anthony Shivakumar

TI__Guru** 113765 points

Hi Anthony,

There is one thing to consider with capacitor placement...for the capacitor to be most effective, the signal should flow through the capacitor before the signal connects to the ADS1234 input pins. It is true that the physical connection appears to be the same, but there is always some small resistance and inductance of the connection traces that effect the performance of the filter. So I would recommend moving 3 of the caps so that the cap is between the resistor and the ADC input pins similar to the connections from the ADC1 connector.

Best regards,

Bob B

0 Anthony Shivakumar over 2 years ago in reply to Bob Benjamin

Prodigy 30 points

Thank you Bob. Got it. 0638.ADS1234.pdf

I have made the change.

0 Bob Benjamin over 2 years ago in reply to Anthony Shivakumar

TI__Guru** 113765 points

Hi Anthony,

That's the correct approach.

Best regards,

Bob B

Data converters

Data converters forum

ADS1234: Minimum circuity to get it working