ADS124S08: Design review

Hi Dennis,

Either share your design by attaching it to the post, or if you feel that the information is too sensitive for public viewing you can either start a private conversation with me on E2E or send to:

mailto:pa_deltasigma_apps@ti.com

Best regards,

Bob B

Bob

I've no problem showing this part of the design.  I'm hoping it will help some else to understand ADS124S08 part better.  So here's the thing:

• The input is a 0 bar to 3 bar pressure sensor using a 3.5K ohm bridge with 1mA excitation current.
• At maximum pressure, the bridge balance is 3.64K/3.34K.  This leads to +Vout=1.82V and -Vout=1.68V for a maximum differential voltage of 140mV.
• My operating temperature will be between 20 deg. C and 90 deg. C.
• Because I will be using the entire range of the sensor and noise and temperature drift must be minimized, I what to have the largest dynamic range as possible.
• To this end, I have chosen AVDD and the external reference voltage to be 5V and the ADC gain to be 32.  The internal filtering and sampling rate are not considered at this time.
• Because of the choice of the voltages and gain, the input voltages fall outside of the parameters of the ADS124S08.
• To correct this issue, I have employed buffering and voltage shifting circuitry.  Please see attached TINA9-TI file.

Any and all comments and advise are appreciated.  Thank you.

Peer Review.TSC

Hi Dennis,

For any small signal output bridge type measurement the desire is to make the measurement ratiometric.  If the measurement is ratiometric then noise and drift of the measurement will cancel out of conversion result as the reference uses the same excitation as the sensor.  If you start to add external circuitry, this will likely add noise and additional sources for drift which will affect your result.  In the end you might actually see better results using no external circuitry, even with the limited dynamic range.

I would suggest that the voltage created across the sensor be used as the source for the reference.  Or possibly you could use a voltage excited source that is the same as the reference. A 5V excitation would provide a slightly larger output.

When using a current source, you could use the IDAC on the ADS124S06 for a 1mA output.

What kind of pressure resolution are you hoping to achieve?  And what kind of data output rate are you hoping to use?

Best regards,

Bob B

Bob

I understand what you're saying about the ratiometric method.  I guess I'm just old school.  I like to get the largest dynamic range as I can.  To the point of the ratiometric method, I have attached another TINA9-TI file to illustrate what I think you mean about the bridge voltage being used for the ADC reference voltage.  Please tell me if I'm correct.  To the issue of the constant current source, do you believe the iDAC on the ADS124S08 will be as accurate, temperature stable and noise free as the one in the previous TINA9-TI file I sent you?  Insofar as the resolution I need, the sensor is in bar but we measure in torr (1bar is approximately 750 torr) and we would like to measure in mtorr.  What I'm trying to do is measure very accurately across the entire range of the sensor. 

Peer Review_1.TSC

Hi Dennis,

You will need approximately 2,250,000 noise-free counts. I have concerns about the ability to achieve this level of precision.  If you add external circuitry to maximize your dynamic range you will also add additional noise and potential drift.

If you look at the noise tables in the ADS124S08 datasheet, you can calculate resolution using the noise tables.  Use the voltage peak to peak as best the achievable performance of the converter.  You may need to consider a slightly better performing ADC such as the ADS1261.

Best regards,

Bob B