• State Resolved
  • Locked Locked
  • Replies 8 replies
  • Subscribers 48 subscribers
  • Views 1402 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.

Two questions about AMC7812 EVM

Seung Joon Lee
Prodigy 80 points
Other Parts Discussed in Thread: AMC7812

Hi,

I have two questions about operation of AMC7812 EVM.

1.

I didn't provide any signal including GND to ADC channel but the ADC-n-data registers have these values.

You can see above figure. Is it normal operation of AMC7812 EVM?

 

2.

And, 0x666 value of ADC-15-channel is supplied by DAC-1-channel with loopback configuration of AMC7812 EVM.

But the 0x666 value are fluctuated when the ADC-14-channel is enable. ADC-14-channel connected with GND.

And, ADC-13-channel is odd value as you can see above figure.

How can I fix it?

 

Best Regars,


Seung Joon Lee    

  • 0
    TI__Genius 11750 points

    Howdy Seung,

    1. These voltage values are typical upon start up of the AMC7812 device.  The inputs are high impedance, but do float to an internally fixed voltage that differs across all analog inputs.

    2. The AMC7812's ADC core shares one sampling capacitor that samples the voltage of every input channel that is enabled through the ADC MUX.  Using the picture you have provided, it seems you are in auto-mode.  Because of the fast conversion rate, the capacitor may not have enough time to charge or discharge the sampling capacitor, therefore not reaching it's transient voltage.  One way to obtain a more accurate read is by decreasing the conversion rate of the ADC block, this gives the sampling capacitor more time to charge or discharge and reach a final value.  The best solution is to sample the ADC inputs using direct mode, as the conversion rate is fully adjustable through the user's software.

    To recap the second bullet of your post:

    -ADC 13, ADC 14, and ADC 15 are enabled

    -ADC 13 is floating, ADC14 = GND, and ADC 15 = 0x666, ADC 

    As previously mentioned, the internal ADC core shares one internal capacitor, which samples the ADC voltages in sequential order.  Therefore, the channels are repeatedly sampled in the following sequence: ADC13>ADC14>ADC15. Since ADC15 is value 0x666, the charge from that connection gets injected into ADC13 when the switching occurs.  As a result, channel 13 displays a nonzero value that is very similar to the previous (ADC 15) channel's value.  The sampling cap then discharges to GND once it connects to channel 14.  Channel 15 fluctuates when Channel 14 is active because the sampling capacitor hasn't had enough time to fully charge and reach a transient value.  Again, the best way to increase accuracy is to decrease the conversion rate (ADC Rate) or sample the input in direct mode.

    Hope this information helps.

    Best Regards,

    Matt

  • 0 in reply to Matthew Sauceda
    Prodigy 80 points

    Howdy Mattew,

    Thank you for yuor kidly reply.

    It is really helpful to understand of AMC7812.

    Best Regards,

    Seung joon Lee

  • 0 in reply to Matthew Sauceda
    Prodigy 80 points

    Howdy, Matthew

    I had more questions.

    1. In direct mode, when the ADC channel 13~15 activated, Ikonw that these channels are sampled by just one trigger signal. Then, I think channel 14 also affect to channel 15. So, I considered that only one channel activates during one conversion time. Is it right?

    2. Setting the AMC-0-configuration register for direct mode and measuring the CNVT and DAV pin  of AMC7812, I can see unknown signals on CNVT and DAV even though external or internal trigger is not applied. And AMC7812 convert ADC input signal. I checked it in watch window with debugger. I know that the CNVT of AMC7812 is input for external trigger signal. Can CNVT oprates output pin?

    Best Regards,

    Seung Joon Lee.  

  • 0 in reply to Seung Joon Lee
    TI__Genius 11750 points

    Hi Seung,

    Sorry for the late reply, it seems like I may have overlooked this question.

    1) When a conversion is triggered in Direct mode the activated channels, in this case ADC 13-15, sequentially get sampled and update the register contents after the last channel gets converted.  This completes one full conversion sequence.  

    In Auto mode, it repeatedly performs the above.

    2) In Direct mode or Auto mode, when a conversion completes the DAV pin creates a low logic pulse:

    The CNVT pin is an input for the external trigger, and should be pulled up to IOVDD through a pull up resistor, typically a 20K value.  Would you happen to have any scope captures of the observed behavior of CNVT?

    Best Regards,

    Matt

  • 0 in reply to Matthew Sauceda
    Prodigy 80 points

    Howdy Matt,

    Here is scope capture in Direct mode.

    The AMC-configuration-0 register is sat 0x0400.

    And, any internal or external trigger signals are not applied.

    But, you can see that the CNVT and DAV is generated by AMC7812. 

    How can I control the signal?

     

    Best Regards,

    Lee

  • 0 in reply to Seung Joon Lee
    TI__Genius 11750 points

    Hi Seung,

    From the Oscilloscope capture, it almost seems like the device is in Auto mode.  Can you verify the mode by reading the AMC Configuration Register 0, 0x4C.  

    Also, you mention that there are no internal or external trigger signals. What is the state of the device in the above picture? Meaning, can you let me know what registers were set prior to this, or if using the EVM software, what settings were selected.

    Thanks,

    Matt

  • 0 in reply to Matthew Sauceda
    Intellectual 450 points

    Hi Matt

    I would like to add a question. 

    If I use the "auto mode" of GUI,  Can't I use up to 500ksps the convertion rate?

    I kown that it used to 500ksps of conversion rate in datasheets.  What is problem? Datasheets faults?

    Please recommed resources to conversion rate!  (below conditions)

    I set it to 500ksps of conversion rate in GUI and  press "auto read" . 

    In this case, problems in using ADC CH13 ~ 14 (about 20mV variation: 0.259mV ~ 0.280mV)

    (used ADchannal: 9~14  (Ch: 9~12 [2V], Ch13~14[0.259mV] <=normal value)

  • 0 in reply to Brady Kim
    TI__Genius 11750 points

    Hello Brandy,

    Thank you for the post.  The conversion rate is selectable in the AMC7812EVM GUI.  As stated in the posts above, the AMC7812's ADC inputs are sequentially multiplexed to an internal sampling capacitor.  If the channel is not driven with a low impedance source there is a chance the capacitor is not charging or discharging to the correct voltage value.  Neighboring channel voltages will also introduce some additional charge injection.  To obtain a more accurate readout the conversion rate can be lowered or Direct mode can be used, as this will increase the time between the sampled channels and allow the sampling cap to charge or discharge.

    The AMC7812EVM Rev B EVM board incorporates two buffers for ADC2/3.  These buffers provide a low impedance source that is capable of driving the ADC inputs. 

    Best Regards,

    Matt