• State
  • Locked Locked
  • Replies 5 replies
  • Subscribers 31 subscribers
  • Views 331 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.

TDC1000-C2000EVM: TDC1000-C2000EVM Evaluation

Federico Gori
Prodigy 40 points
Part Number: TDC1000-C2000EVM
Other Parts Discussed in Thread: TDC1000, TDC7200,

Hi,

I'm doing some tests in order to validate some measure, I'm using a TDC1000 and TDC7200 and I'm comparing
my measure with the one taken by using TDC1000-C2000EVM. By setting the average on one cycle I measure 31,97us
and TDC1000-C2000EVM measures 31,86 us. First issue is that by using
TOFn=normLSB(TIME1-TIMEn+1)+ CLOCK _COUNTn*CLOCKperiod only CLOCK _COUNTn*CLOCKperiod results equal to 31,875us,
I measure 255 for CLOCK_COUNT(1 stop pulse set) and clock is set to 8Mhz,so I've already a value greater than
the one evaluated by TDC1000-C2000EVM.
In order to increase precision I've decided to use multi-cycle average mode, on 8 cycles. TDC1000-C2000EVM measure
is still 31,86 but my measure goes to 31,21 us. The thing I've noticed is that CLOCK _COUNTn I read is 1997 which
averaged in 8 cycles becomes 249,625 that is 5 Clock cycles less than single average measure.
If I further increase the average cycles the value drops.

What could be the problem?

Thanks,

  • 0
    TI__Expert 3450 points

    Hello,

    Were any error flags reported when taking these measurements?

    (see data sheet 8.4.8 for more details)

    Regards,

    Gabriel

  • 0 in reply to Gabriel Lake1
    Prodigy 40 points

    Hello Gabriel,

    thank you for you answer. No error flags, and when I check the interrupt source, is the correct one. 

    Do you have any other idea?

    Thank you 

  • 0 in reply to Federico Gori
    TI__Expert 3450 points

    Hi Federico,

    I am looking into this further. Could you provide more details about your testing?

    - Sensor target

    - transmission medium

    - transducer frequency

    - measurement range

    - have you made any hardware optimizations (band pass filter, tuning, etc)?

    - please save and provide your TDC1000-C2000EVM register configuration (TDC1000 tab in the GUI)

    thanks,

    Gabriel

  • 0 in reply to Gabriel Lake1
    Prodigy 10 points

    Hi Gabriel,

    I'm working with Federico, I had some problem with my TI account so he wrote the post for me. We're working with a sckrew so we're working with iron. We're working with 8Mhz oscillator as clock. We've probed with the oscilloscope the start and stop signal coming from the TDC1000 and we see that the TOF is something between 31,8 and 32 us. The problem seems to be with TDC7200, where're there's no hardware optimization.

    In term of configuration we've confugure both the Evaluation board and our TDC1000 in the same way: CONFIG_0: 0x04, COFIG_1:0x041, CONFIG_2: 0x00, CONFIG_3: 0x08, CONFIG_4: 0x5F, TOF_1: 0x40, TOF_0:0x1E, TIMEOUT: 0x00 and CLOCK_RATE= 0x00.

    For TDC7200 we've configured CONFIG_1: 0x02 and CONFIG_2: 0x40. We've changed to multicycle average mode accordingly for both the evaluation board and the TDC1000+TDC7200

    Thanks for your help,

    Michele

  • 0 in reply to Michele Tarocco
    TI__Expert 3450 points

    Michele, Federico,

    Could you please clarify the issue for me? I see that you are taking TOF measurements using the TDC1000-C2000EVM with a deviation on the order of 0.5 us or so. What is the standard deviation that you are seeing with the TDC1000+TDC7200, and what is the required accuracy and precision for your application?

    The TDC1000 or transducer frequency would be the limiting factor in measurement precision. The TDC7200 expected measurement deviation is basically negligible compared to the possible system deviation if using a 1MHz transducer for example with around 1mm of error (~0.2us TOF difference in iron). 

    Regards,

    Gabriel