DRV8312EVM serial-port example

I found  the demo: sci_echoback

And  found that GPIO28 and GPIO29 are both on J12.  (yes, I also connected ground)

 I see the welcome text in the terminal program. But when i try to echo back. No result, high to half-level signal on scope. It seems to be that GPIO29 is active high and an output.

 I see the welcome text so RX and TX are not crossed.  I checked with the scope that my TX (pc) signal is correct.

 So is GPIO-29, the GPio29 I think it is. DIMM Module pin 93.

Schematic and board are both revD. So why isn't it working.

It is not GPIO29 but GPIO28. But if I just run the demo this is the result:

Send 'a' to the DSP.

I did verify that this is GPio28 by writing a program that toggles GPIO28. and that works correct.

 So why is this not working.

 I use this demo: C:\ti\controlSUITE\device_support\f2806x\v136\F2806x_examples_ccsv5\sci_echoback

and this is the port initialization:

void InitSciaGpio()
{
   EALLOW;

/* Enable internal pull-up for the selected pins */
// Pull-ups can be enabled or disabled disabled by the user.
// This will enable the pullups for the specified pins.

    GpioCtrlRegs.GPAPUD.bit.GPIO28 = 0;    // Enable pull-up for GPIO28 (SCIRXDA)
//    GpioCtrlRegs.GPAPUD.bit.GPIO7 = 0;     // Enable pull-up for GPIO7  (SCIRXDA)

    GpioCtrlRegs.GPAPUD.bit.GPIO29 = 0;       // Enable pull-up for GPIO29 (SCITXDA)
//    GpioCtrlRegs.GPAPUD.bit.GPIO12 = 0;       // Enable pull-up for GPIO12 (SCITXDA)

/* Set qualification for selected pins to asynch only */
// Inputs are synchronized to SYSCLKOUT by default.
// This will select asynch (no qualification) for the selected pins.

//    GpioCtrlRegs.GPAQSEL2.bit.GPIO28 = 3;  // Asynch input GPIO28 (SCIRXDA)
//    GpioCtrlRegs.GPAQSEL1.bit.GPIO7 = 3;   // Asynch input GPIO7 (SCIRXDA)

/* Configure SCI-A pins using GPIO regs*/
// This specifies which of the possible GPIO pins will be SCI functional pins.

    GpioCtrlRegs.GPAMUX2.bit.GPIO28 = 1;   // Configure GPIO28 for SCIRXDA operation
//    GpioCtrlRegs.GPAMUX1.bit.GPIO7 = 2;    // Configure GPIO7  for SCIRXDA operation

    GpioCtrlRegs.GPAMUX2.bit.GPIO29 = 1;   // Configure GPIO29 for SCITXDA operation
//    GpioCtrlRegs.GPAMUX1.bit.GPIO12 = 2;   // Configure GPIO12 for SCITXDA operation

    EDIS;
}

Which document do I need for the the I/O selection/definition?

Hope you can help.

ernst van spronsen said:

Which document do I need for the the I/O selection/definition?

The SCI TX data seems fine. Did you monitor this data on your PC? You should be able to receive 'a' on hyperterminal.

The DSP TX line works fine. It is the DSP RX signal that isn't working.

if I send a 'a' from my pc I see it the scope picture above.  If i disconnect the wire from the DSP the signal is correct.(0  - 3.3V signal)

ernst van spronsen said:

if I send a 'a' from my pc I see it the scope picture above.  If i disconnect the wire from the DSP the signal is correct.(0  - 3.3V signal)

So do you get any data on the Watch window or not? Esp. add SCIRx register in the Global registers to the watch window to observe the incoming data.

Regards,

Gautam

Look at the scope picture. This signal isn't correct.

The signal doesn't make it to ground. Why? the FTDI usb-> serial converter isn't capable to sink the signal. Because it is broken(Don't think so)  or the control  source more than the FTDI can sink. But why is the controller sourcing a input signal. When I disconnect the pin. The (!!RX!!) pin is still high (without pull-up).

 I think that is strange and can't find why. Also tried it on different demo boards and they have the same problem.

To proof: This are both serial signal. Green is controller TX(at start-up). Showing the welcome text.

Yellow is the RX signal: floating(Not connected to PC) with 3.3k Pull-down resistor!!. So I think there is something wrong with the demo.

The odd thing is. When I select the GPIO28 as GPIO input. I see the same behavior.

code:

    EALLOW;
    GpioCtrlRegs.GPAMUX2.bit.GPIO28=0;
    GpioCtrlRegs.GPADIR.bit.GPIO28=0;
    EDIS;

(I'm not that far in the documentation. To know for sure if this is correct GPIO input initialization).

Very wierd! The TTL signals should not cross the 0 towards negative side. Are you sure these are the TTL signals towards & from the controller?

There are internal PUs on the IOs, but these are very weak, ~200uA.  You could go into GPAPUD register and disable the one on GPIO28 to be sure, but this wouldn't hold the signal high against it's own driver or a 3.3kOhm PD as you mentioned.  I just want to take that out of the equation.  Also, just to confirm J12 is removed, correct.

I have a F28069 control stick that I use the FTDI chip for the same echo back and it works; so the only difference(other than the MCU) is the ISO chip:  Can we probe the in/out of this guy?

I did the measurements on the J12 pins. so yes the jumper is removed. If that is what you mean. I also did measurements on Docking-Stn USB-EMU [R3] on GPio28 and GPio29.

 I have the F28069M.

I removed my serial-ttl (FTDI ttl-232r-3v3). for the Pull-down measurement test. So I also removed that out of the equation. So the only think I can think of: is that the F28069M or my test-software is the problem.

-  Is it for you possible to test with the M version.

 - can you please sent me you're version of the demo.

-  can you tell me how to configure GPio29 as input (with no pull-up) and how do I read the the pin in the debugger.

 Thanks

If i'm correct this piece of code. Sets GPIO28 as input without pull-up resistor.

    EALLOW;
    GpioCtrlRegs.GPAPUD.bit.GPIO28 = 1;
    GpioCtrlRegs.GPAPUD.bit.GPIO29 = 1;

    GpioCtrlRegs.GPAMUX2.bit.GPIO29=0;
    GpioCtrlRegs.GPADIR.bit.GPIO29=0;
    GpioCtrlRegs.GPAMUX2.bit.GPIO28=0;
    GpioCtrlRegs.GPADIR.bit.GPIO28=0;
    EDIS;

I breakpointed direct after this code and the signal is still 3.3V. And that is odd because of my 3.3k resistor at the input. (checked with scope) 

I changed the module for an other one,  with the same result.

After a week searching what I did wrong. I changed from SCI-A to SCi-B. And redid the I/O pin test and modified the example code to work with SCIB GPIO18 and GPIO19. And this worked. directly as expected

I have the same problem with 2 of the same boards. So I think you have to check you're F28069M silicium.

The problem:

     EALLOW;
    GpioCtrlRegs.GPAPUD.bit.GPIO28 = 1;
    GpioCtrlRegs.GPAPUD.bit.GPIO29 = 1;

    GpioCtrlRegs.GPAMUX2.bit.GPIO29=0;
    GpioCtrlRegs.GPADIR.bit.GPIO29=0;
    GpioCtrlRegs.GPAMUX2.bit.GPIO28=0;
    GpioCtrlRegs.GPADIR.bit.GPIO28=0;
    EDIS;

Is you execute this code. You expect the signal to go low. But is doesn't . It is stays a low impended high signal.

 If you configure it as output is works fine(low and high output).

Is there a work around available for this problem?