• State Resolved
  • Locked Locked
  • Replies 3 replies
  • Subscribers 46 subscribers
  • Views 249 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.

ADS1278: High Speed mode with no input buffer?

Bruce McKenney47378
Guru 88540 points
Part Number: ADS1278
Other Parts Discussed in Thread: THS4521, THS4561, THS4551

We're looking to design an ADS1278 into our monitoring system. We haven't worked with the ADS1278 before. We have in mind to use High Speed mode.

Data sheet (SBAS367F) p. 37 (Application Information) and Figs 88-89 seem to assume that we will use an input buffer. We would like to connect the sensor directly to the differential inputs, without any amplifier in between.

The signal level is roughly 1V(p-p) so we don't need any gain. Our input's impedance will be <200 Ohms. Based on Fig 69, I estimate the ADS1278 effective impedance is about 10kOhm (Fclk=37MHz, so Fmod=~9MHz). It seems as though our sensor should be able to keep up with the current requirements.

So I'd like to ask here: Is there a reason we can't/shouldn't omit the buffer amp between our sensor and the ADS1278 inputs?

  • 0
    TI__Guru 57296 points

    Hello Bruce,

    Thank you for your post.

    The recommendation for the input drive amplifier comes from settling the dynamic transients introduced by the ADC's sample-and-hold capacitor input. The input impedance spec, which is proportional to the sampling frequency of the delta-sigma modulator, is only an average impedance, which is calculated from the average input current. The dynamic current demand at the beginning of each sample is much higher. This will produce a voltage transient at the input pins, which needs to settle below the noise floor within approximately 1/2 the modulator sampling period. Any remaining unsettled voltage error will translate directly to an ADC output error.

    We have a pretty comprehensive training series on ti.com called Precision Labs - ADCs. There is a specific section that looks at optimizing the R-C components at the ADC input (known as a "charge-bucket" filter) as well as choosing the input amplifier bandwidth requirements. The series focuses on SAR ADCs, but the fundamental principles are very similar.

    In addition, there is a series on delta-sigma ADC systems with videos on how to optimize the peripheral circuitry. There is also an input drive section that illustrates how improper drive amplifiers produce unwanted signal distortion.

    Best regards,

  • 0 in reply to Ryan Andrews
    Guru 88540 points

    Thanks. I guess when one gets into high-end ADCs these things become more important.

    In earlier versions of this design, we had trouble with signal distortion from front-end opamps, so the first thought was to avoid them entirely. The THS4521 specs look pretty good, so maybe we'll just try it with the EVM (we know what we're looking for). It looks like the EVM only uses the THS4521 on half the inputs, so I imagine we can even try them side-by-side.

    I don't suppose TI is thinking of getting (back?) into CTSDs? I think they avoid these artifacts by design.

  • 0 in reply to Bruce McKenney47378
    TI__Guru 57296 points

    Hi Bruce,

    The gain-bandwidth and output swing capabilities of the driver amplifier play a big role in minimizing distortion. On the EVM, these circuits are usually tuned to handle the fastest modulator sampling speeds with full-scale input signals. For your information, there are some newer FDA devices that were released after the ADS1278 and THS4521. I'm not a high-speed amps expert, but you might be interested in the THS4561 and THS4551.

    Unfortunately I'm not at liberty to discuss our future roadmap on a public forum.

    Best regards,