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.

SN74AVCH16T245: “Notch” in waveform

Part Number: SN74AVCH16T245
Other Parts Discussed in Thread: ADS8568

Hi, my project uses SN74AVCH16T245 to connect the parallel port between ADS8568(3.3V) and C6657(1.8V). C6657 uses EMIF to drive and read ADS8568. 

But the result read has some error data. So I measured the waveform of EMIF bus.

The waveform is shown below. The yellow wave is OE, green wave is EMIF_D[x] on the ADS8568 side, and the orange wave is EMIF_D[x] on the C6657 side.

You can see some "notch" in orange waveform.

I guess that's the reason but I don't know how it happens.

I also change the C6657(1.8V) to C6748(3.3V), but only get the same result as below. The yellow wave is OE, and the orange wave is EMIF_D[x] on the C6748 side. 

  

And if leave the DSP side of the voltage transceiver floating, that is the voltage transceiver connects the ADS8568 but does not connect the DSP's EMIF_D[x] port, the wave becomes normal as well as the ADS8568 side.

So I wonder how it happens and how to solve it. Thank you!

  • This looks as if another device is trying to drive the bus to another voltage.

    Please show a schematic. Does the transceiver ever change direction? If yes, how it this controlled?

  • Thanks for your reply!

    The "OE" waveform above is the EMA_OE singnal of the EMIF port. Not the OE pin of the  SN74AVCH16T245.

    The waveform meassured above in the direction of ADS8568 to DSP.

    Below is schematic of ADS8568  :

    Below is schematic of SN74AVCH16T245: inlcuding data bus and control pin of EMIF port and ADS8568

    Below is schematic of C6748, it is actually from the LCDKC6748 EVM.

    The "DBx_A_DSP" pin connects to "EMA_Dx" pin of LCDKC6748 EVM through wire.

    CS_A_DSP ---------  EMA_CSn_2

    RD_A_DSP --------- EMA_OEn

    WR_A_DSP --------- EMA_WEn

    CON_CTRL_A_DSP and BUSY_A_DSP connect to the GPIO pin of LCDKC6748 EVM.

    The U3 transceiver's direction is DSP to ADS8568 in the initialization period. Then it changes to the direction of ADS8568 to DSP for DSP to read sample data and doesn't change any more.

    The direction of  port 1 of U4 transceiver is DSP to ADS8568, and the direction of port 2 is ADS8568 to DSP. And it doesn't change direction while running.

  • Please create a waveform capture that combines the affected signal (orange) and the DIR pin of the corresponding port at the transceiver

    I suspect that there might be too much noise on DIR (and probably other signals). As long as you have to use wires, try to shield them.

  • Below is the waveform capture required. The yellow wave is "RD_A_DSP" signal, the green wave is the "DIR" signal, and the orange wave is the affected signal.

    Maybe the "notch" is due to the data bus parking feature of the EMIF port that it will drive the data bus to the last value write when idle.

      

  • Hey Yellen,

    That might be the best assertion here. I don't really see anything that could be causing the notches other than some contention on the bus as Clemens mentioned earlier. This would also agree with the waveform being normal when the DSP is disconnected. Do you know how long of a pulse the data bus parking feature applies?

  • Thanks for your reply!

    Refer to the manual of EMIF port, the data bus parking applies when the data bus is idle. That is between the last read operation(CS pin de-assert) and a new read operation(CS pin assert). Time may vary due to the clock and configuration of the peripheral. In the waveform above, data bus parking lasts for about 400ns.

  • Hi Yellen,

    If this is something you can disable, then I believe this will solve your issue.