TDC1000: Analog always ON - RX Line always ON

Part Number: TDC1000
Other Parts Discussed in Thread: TUSS4440, TUSS4470

Tool/software:

We are working with a TDC1000 as a receiver only, in a simple point-to-point ultrasonic communication demonstrator. In that way, we would require the TDC1000 to mantain the RX line always ON. 

a) Is that possible? would anyone give us a hint, if possible, the combination of registers configuration, trigger pulse control, etc?

b) It seems that the TUSS4440 have the "Listen Mode" always ON configuration, is that correct?

c) Is there any other Texas IC that can be recommended as an AFE for ultrasound transducers with LNA, PGA, programmable bandpass filters (as TUSS4440) but without the echo processing stage? (which we do not require)

Thanks in advance

  • Hello Martin,

    Thanks for posting on the forum.  Please see the thread below regarding TDC1000 in receive only mode.  Since there is a receive window that must be set, it is not really the best implementation for a RX only application.  I would suggest the TUSS4440 or TUSS4470.    

    https://e2e.ti.com/support/sensors-group/sensors/f/sensors-forum/1205891/tdc1011-tdc-usage-as-rx-device-only 

  • Hello Eddie,

    Thanks for your answer. I've read the thread, which concludes in the TUSS4470 as a better option.

    I agree with that, but the thing is we have a couple of TDC1000 and none TUSS4440/70 by now.

    We've tried a few combinations of large echo windows, short Tx  pulses, and periodic retrigger, but always the Rx line turns off a while.

    Anyway, we will go on trying to get the best response we can, untill que can get other ICs.

    Thanks again.

  • Understood.  Yes, definately would recommend TUSS for listen only.

  • Well, we arrived to some configuration parameters and handling of the TDC1000 that seem to work well, and may be useful to other TDC1000 users that want the LNA/PGA Rx line only, and always active.
    The main idea is to make an unique TX trigger, with some Tx pulses, and predict a delay from trigger pulse that asures that the Rx line is active, in ECHO LISTEN stage. At that moment, turn-off CLKIN signal, so that the state machine remains still, in that stage. The PGA gain remains configurable (the SPI works well so the register is written), and the gain changes accordingly.

     As a reference, the process can be:

    1) Fix a number of Tx pulses greater than 0 (we used 3), some TX_FREQ_DIV (this affect the delay computation)

    e.g.: 0x00, 0b01000011 (Nºregister, bit values)

    2) Set high numbers of RX echoes expected (we set 7).  0x01, 0b00000111

    3) Set a high echo threshold (just in case, in our case we don't want any detection at all). 0x03, 0b00000111
    4) Receive mode in multi echo enabled mode (bit 6 of Reg 4, others bits doesn't care much).  0x04, 0b01011111

    5) Echo timeout disabled and long timeout config (MSB) -> Standard TOF Measurement. 0x08, 0b00000111

    6) Make the trigger pulse, and wait the amount of time required to assure ECHO LISTEN stage (we set 150 us in our case, the minimum time comes from T0 and T1).

    (Note: We used a 2 MHz CLKIN.)

    T₀ = 1/CLKIN = 0.5 us
    T₁ = TX_FREQ_DIV·T₀ = 4 us TX_FREQ_DIV=2 -> %8

    AUTOZERO_PERIOD=0

    tlisten≈NUM_Tx×T1+128×T0+2^AUTOZERO_PERIOD×64×T0= 3·(8·T₀) + 128·T₀ + 64·T₀
    = 108·T₀ = 54us

    7) Deactivate the CLKIN signal and do not retrigger (one shot).

    With that configuration, we obtained the following (yellow - output of PGA)

    Similar one-trigger configuration, but WITHOUT disabling CLKIN presents some issues.

    Periodically trigger didn't worked well.

  • Martin,

    Thanks for providing details!  This could prove useful for another user.