• State Resolved
  • Locked Locked
  • Replies 4 replies
  • Subscribers 47 subscribers
  • Views 1473 views
  • Users 0 members are here
Support feedback
Options
Options
Related

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.

ADC Problems with using ADS1298

Jonghyun Park
Prodigy 50 points

Other Parts Discussed in Thread: ADS1298, ADS1299

Hi, I'm currently using ADS1298 to precisely convert analog signal to digital signal. It is mainly designed for ECG signal, but not using for it.

I put AVSS, DGND into ground, and connect AVDD, DVDD to 3V following an unipolar configuration.

First, In this case(unipolar power and internal analog reference ), what can be the whole ADC reference range?

Second, I want to convert the signal whose center-voltage is 1.2V (I intentionally level-shifted signal to 1.2V). Then how should I connect the INP, INN?

Third, It is relevant with fist question, the datasheet mentions +Vref and -Vref in ADC data format. How should I interpret -Vref when it is working with unipolar power? so how should I understand ADC data format in unipolar power. 

Those three things have been troubling me a lot because I already made a system with ADS1298, but it isn't converting as I'm expecting.

I appreciate your help in advance. 

-jay

  • 0
    TI__Mastermind 38745 points

    Hello Jay,

    The ADS1299 requires an analog power supply range (AVDD-AVSS) of between 4.75 V and 5.25 V. That means that if connected in unipolar format, where AVSS = 0 V, then AVDD must be between 4.75 V and 5.25 V. The digital power supply DVDD may be anywhere between 1.8 V and 3.6 V. Unfortunately, if your positive analog power supply is connected to 3 V, the device may not perform to specification. If the power supplies are applied correctly and the internal reference is used, that means the ADC will measure all gain adjusted input signals compared to the 4.5 V internal reference which is provided by the ADC. That is to say that the difference between your input voltages multiplied by the PGA gain must be between -4.5 V and +4.5 V to avoid clipping. What is the input range and PGA gain you intend to use for the case you specified? As for your third question, negative digital codes will be returned as conversions when AINP – AINN < 0. When using a unipolar supply, the values of the individual inputs with respect to AVSS will not be negative. The digital codes will take a 24-bit two’s complement format where the positive full-scale code is 0x7FFFFF and the negative full scale code is 0x800000.

    Hope this helps,

    Brian

  • 0 in reply to Brian Pisani
    TI__Mastermind 38745 points
    Jay,

    After reviewing my previous post, I realize that you were using the ADS1298 not the ADS1299 and that some of my comments are incorrect. First of all, the device should be fine with a 3 V unipolar analog supply. Secondly, the internal reference will be 2.4 V when using that analog supply (ensure that the VREF_4V bit in the CONFIG3 register is cleared). My comments regarding how the inputs are measured in relation to the reference still apply, though. The difference in input voltages (AINP-AINN) multiplied by the PGA gain is measured on a scale where +/-VREF represent positive and negative full scale. Using the internal reference voltage, the inputs multiplied by the PGA gain must be maintained in between -2.4 V and +2.4 V to avoid clipping. As I said in my last post, in a unipolar configuration, the individual input voltages will not be negative with respect to AVSS, but negative output codes correspond to a scenario where AINN > AINP. Please let me know if you need any more clarification.

    Brian
  • 0 in reply to Brian Pisani
    Prodigy 50 points

    Brian,

    Though It has been long time since your reply for my question, one question has just occured to my head.

    If the difference input can be measured between +/- ref (/PGA), can I interpret the magnitude of dynamic  range of ADC as 2*Vref? (-2.4V < AINP-AINN < 2.4V ?)

    If right, It seems weird because power supply is  avss = 0 avdd= 3v(I used unipolar configuration and powered with 3V) and then the magnitude of dynamic  range(which I consider as two times Vref, 4.8V)  is bigger than the power range. 

    Adding one more question on that, as far as I know, individual input should exist between avss and avdd. What if my signal is fed into inputs which has common voltage of 1.2V? I guess then because the maximum input voltage on AINP is limited to about avdd-0.3V, voltage range that I can convert  is much less than Vref(2.4V).  From my calculation, I can just use about +1.5V of AINP-AINN ( AVDD-0.3V-1.2V ~= 1.5V) and about -1V of AINP-AINN (1.2V- (AVSS+0.3V) =0.9V).

    So It seems  I'm wasting the dynamic range of adc conversion.

    How can I correct this situation?

    One solution I'm thinking is using PGA gain bigger than 1( I was setting PGA gain as 1) , which I think could span the dynamic range of adc. 

    Jay

  • 0 in reply to Jonghyun Park
    TI__Mastermind 38745 points
    Hello Jay,

    You're right, the maximum conversion range of the device is nominally larger than the power supply range. However, that assumes that you are measuring in a "single-ended" configuration where one input is DC and the other is time-varying. Imagine instead that both inputs are time varying with magnitude equal to Vref, but with 180 degrees of phase shift between them. In this scenario you could imagine an output that spans the entire conversion range while staying within the power rails.

    You should not feel like you're "wasting" anything. That dynamic range is there for you to use as much or as little as you want! One reason to use a device with large dynamic range is because in ECG systems, there could be offsets of hundreds of mV where the ECG signal is only a few mV. You don't care about the offset voltage necessarily, but the wide dynamic range makes it so you do not need to AC couple the inputs or use a large gain stage to measure the small ECG signal riding on large offset.

    In addition, noise from the voltage reference scales linearly with input voltage. This means that the more of the conversion range you use, the more reference noise will couple into your signal chain. In summary, don't feel bad about only using a little bit of the part's dynamic range!

    Regards,
    Brian Pisani