• State Resolved
  • Locked Locked
  • Replies 12 replies
  • Subscribers 30 subscribers
  • Views 3451 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.

AFE5808A: data process

Sunnyrain
Prodigy 90 points
Part Number: AFE5808A

Hello, TI supporter,

Currently, I test the AFE5808AEVM, some question bother me,I need your help to clarify them.

1. If I got raw data from LVDS of ADC, how to caculate the real input amplitude?

2.When I change the PGA from 24dB to 30dB, I found the value from ADC is not change, that means the output is not amplified. why?

3. In the Table 10 which is in the page 43 of datasheet, when the Vcntlp-Vcntln = 1.5, why there is negetive gain?

I am very appreciated if get the feedback ASAP.

  • 0
    TI__Expert 4145 points

    Hi there,

    Regarding converting the raw LVDS output data of the AFE5808AEVM, it depends on your ADC output format. The ADC can output data in either 2's complement or offset binary. You would have to first convert that to decimal. Once you get the value in decimal, you need to convert it to volts using the ADC FS range of 2Vpk-pk then work your way backwards in the signal path to find the LNA's input voltage. i.e. ADC->LPF->PGA->VCAT->LNA.

    What is your exact setup with the AFE5808AEVM? Are you using the TSW1400EVM in your evaluation? What are your PGA, LNA, VCAT settings in dB?

    Is your input signal within the selected LPF bandwidth?

    I am not sure what datasheet you are referring to but this one has no Table 10 on Page 43. Pages 30 and 31 do go over how the Voltage-Controlled Attenuator works using the VCNTL signal.

    Sincerely,

    Olu

  • 0 in reply to Olu Sodipe
    Prodigy 90 points

    Hello Olu,

    I am gald to receive your feedback.

    Currently, I set the input signal is sine wave with Vpp 0.05V. I set LNA=24dB,PGA=24dB,TGC=0dB. 

    when I disable active termination, I got about 510mV from ADC. If I set active termination 50ohm, I got about 36mV from ADC. What is the function of active termination? How to select it? The two values I got from ADC are not meet the gain of LNA+PGA, so how to caculate them?

    thanks!

    Best Regards!

    Sunny Lee 

  • 0 in reply to Sunnyrain
    TI__Expert 4145 points

    Hi Sunny Lee,

    Accurate termination is needed to minimize reflections for input signals transmitted over long cables. The voltage at the LNA's input pins depends on the voltage division between the source resistance (Rs) of the input signal's generator and the impedance presented by the LNA inputs.

    You want to match the entry in Table 8 to the signal source resistance. Page 51 and table 8 go over the pre-set active termination options and the registers to set them.
    Lastly, the input signal voltage is measured at the LNA input pins and not at the signal generator. This is because most signal generators have a source resistance of usually 50ohms in the signal path before the device-under-test.

    Please go over the rest of the datasheet if you have more questions about the device.

    Sincerely,
    Olu

  • 0 in reply to Olu Sodipe
    Prodigy 90 points

    Hi Olu,

    I've done hundreds of test. But I could not get the desired result. Could you give me some detail example? Such as when input signal is ? volt, LNA=?, TGC= ? PGA=?, then you can get the result is ?.

    Hope I can deep understand from the example you give me.

    Thanks a lot.

    Best Regards

    Sunny

  • 0 in reply to Sunnyrain
    TI__Expert 4145 points
    Hi Sunny,

    Let's go with your previous example of 0.05V Vpp. For one active termination setting, Instead of measuring the voltage at the signal generator, measure it at the right side pad of C22 (from Page 34 of the user guide). This will give you an accurate reading of the voltage going into the ADC. Note that the voltage you measure here will likely vary as you change the active termination settings.

    After measuring that as a pk-pk voltage, convert the LNA+PGA+TGC gains in dB to V/V by using 10^(Gain/20). This is the peak-to-peak voltage value you expect to be represented in digital codes at the ADC output. Make sure your input signal frequency is much lower than the LPF bandwidth.

    The other half of the problem is just converting the actual digital codes from the ADC output back to volts. I am still not sure what capture card or setup you are using to evaluate the AFE5808AEVM so I don't know what to recommend here. How you are capturing the output data would determine the most direct way to convert it to volts .

    Sincerely,
    Olu
  • 0 in reply to Olu Sodipe
    Prodigy 90 points
    Hi Olu,

    My understand is same as you. According to LNA+PGA+TGC to caculate the gain. But real test is different. For example, I connected the ultrasound probe as my input signal. I set LNA=24dB,PGA=24dB,LPF=20MHz.

    I want to do a serial of test according to the different Vcntl from 1.1V to 0.1V. When Vcntl is from 1.1V to 0.9, I can get the result that the Value of peak to peak is from several mV to dozen mV. But when Vcntl = 0.8V, all code I got is 0xFFFC. Why? I set the LNA = 18dB, I can get the normal code. why? I think the input signal is less than 0.25V, I should use LNA=24dB to test, but I could get normal code. Could you tell me what is the reason. Thanks.

    Best Regards

    sunny
  • 0 in reply to Sunnyrain
    Prodigy 90 points
    Hi Olu,

    I also find something strange. When I change the Vcntl from 1.1V to 0.9V, the offset voltage of result is increasing that are 1.2V, 1.4V, 1.7V.(the code is 2's offset). So I think when the Vcntl is 0.8V，the offset voltage is more than 2V, so the result is 0xFFFC. Am I right? So why the offset voltage is change? How to reduce it or get rid of it? Thanks.

    Sunny
  • 0 in reply to Sunnyrain
    Prodigy 90 points
    Hi Olu,

    I find set DIGITAL_HPF_FILTER_K register can get rid of the offset voltage. But when Vcntl = 0.7V, all output code is 0x8000. I think the vcntl is smaller, the gain is bigger. Why the output code is 0x8000. I can read the normal value when Vcntl is over 0.7V. Thanks! By the way, how to select the K value? Currently, I set K is 10.

    Sunny
  • 0 in reply to Olu Sodipe
    Prodigy 90 points
    Hi Olu,

    Another question is about clamp current in PGA. what is the function? How to use it in real product? Thanks.

    Best Regard!
    Sunny
  • 0 in reply to Sunnyrain
    TI__Expert 4145 points
    Hi Sunny,

    You are trying to do too many things at the same time. Pick a specific value for active termination, a specific value for VCNTL, then acquire the data at the ADC's output. Convert it to decimal then post them here. Also post your gain values and the input voltage measured from C22 as described in my post above.
    The datasheet goes into a lot of detail about VCNTL on Pages 30 and 31.

    Sincerely,
    Olu
  • 0 in reply to Olu Sodipe
    Prodigy 90 points

    Hi Olu,

    May I get your email? I can share you different result according to different test condition.

  • 0 in reply to Sunnyrain
    TI__Expert 4145 points
    Hi Sunny,

    I sent you a message on e2e--you should be able to send me a message now.

    Sincerely,
    Olu