How to use the keyboard in the "GRBL_MPG_DRO_BoosterPack" project

[CHTUZKI]

Hi.

I want to use the keyboard, but I have encountered some problems. I use MSP430G2554 to drive the keyboard. The IIC interface of MSP430G2554 (keyboard board) is connected to tiva-c (DRO board), but the keyboard does not seem to work, and MPG cannot be used.

I don't know where to check. I looked at the code. The keyboard seems to be related to the UI. Can you give me some help?

Thank you.

[terjeio]

Assuming you have buttons for jogging the first thing to check is the keypad strobe signal (PC4), is it going low while pressing a jog button? Second thing is to check if the I2C bus is not stuck - both SDA (PA7) and SCL (PA6) should be 3.3V. Do you have access to an oscilloscope or a logic analyzer?

[CHTUZKI]

Thank you for your reply.

I have many questions about the keyboard processing of the "GRBL_MPG_DRO_BoosterPack" project. I am not very clear about this process.

I found that MSP430G2554 is connected to the keyboard and communicates with the tiva-c board through IIC, but MSP430G2554 is an IIC slave device. If a key is pressed, how does tiva-c know that a key is pressed?

I found the following lines of code in the "GRBL_MPG_DRO_BoosterPack-master\Firmware\TM4C123\driver. h" file:

#define KEYINTR_PIN               GPIO_PIN_4          
#define KEYINTR_PORT            GPIO_PORTC_BASE     

What is this for？

I found a function in the "GRBL_MPG_DRO_BoosterPack-master\Firmware\TM4C123\i2c. c" file:

void i2c_nb_send (uint32_t i2cAddr, const uint8_t value)
{
    while(i2cIsBusy);

    i2c_m.buffer[0] = value;
    i2c_m.count     = 1;
    i2c_m.data      = i2c_m.buffer;
    i2c_m.state     = I2CState_AwaitCompletion;

    I2CMasterSlaveAddrSet(I2C1_BASE, i2cAddr, false);
    I2CMasterDataPut(I2C1_BASE, *i2c_m.data);
    I2CMasterControl(I2C1_BASE, I2C_MASTER_CMD_SINGLE_SEND);
}

I search for this function throughout the project. I found only two calls，

leds_setState and keypad_setLeds,

They just seem to get the status of LED, where to use i2c_nb_send() function gets the keyboard data？ Can you give me some help? Thank you very much。

[terjeio]

where to use i2c_nb_send() function gets the keyboard data？

Data is read from the MSP430 in an interrupt handler, the interrupt is triggered by the MSP430 pulling PC4 low. For most keys this is vith a short pulse, but for jogging the pin is held low until the key is released (key up event). When released the jog is terminated.

[CHTUZKI]

I'm sorry. I accidentally turned off this issue.

I'm trying. But there are still problems.

I use MSP430G2554 to drive the keyboard, and I use "Code Composer Studio 12.0.0" to debug. It seems that everything is normal, but the I2C interrupt of MSP430G2554 will never enter. I have connected the pull-up resistance on P1.6 and P1.7, which is the data displayed by my logic analyzer.

The MSP430G2554 I use does not have P3 port. Will it affect me? I see that P3 port is only used to drive LED. I think it will not affect me?

I only use STM32. I feel MSP430G2554 is very complex😁😁😁

[terjeio]

I2C interrupt of MSP430G2554 will never enter.

It should be entered after P2.5 (or P.2.4 depending on the configuration) is driven low as this should interrupt the Tiva and prompt it to read the keycode. For this to happen the MSP430 has to be waken up from sleep to perform a keyscan and detect a valid keypress. Does this happen?

The MSP430G2554 I use does not have P3 port. Will it affect me?

No, it should not. AFAIK the chip inside the package has the port but it is not wired to pins.

I only use STM32. I feel MSP430G2554 is very complex

IMO all processors seems complex until you get to know them... BTW there is nothing to stop you from porting the MSP430 code to another processor. I chose the MSP430 since it is cheap and I had a bit of experience with them beforehand.

[CHTUZKI]

yes. When P2.5 is low level, tivaC will actively request data from MSP430G2554. At this time, MSP430G2554 should enter the iic interrupt, but it did not enter. I plan to migrate MSP430G2554 code to STM32.

[CHTUZKI]

Whether the "GRBL_MPG_DRO_BoosterPack" project can independently control grblHAL? I want to make an offline controller to control the grblHAL board without using the "IOsender" software. Can this project be achieved

[terjeio]

Can you zip your MSP430 binary and add it to a comment so I can test it here?

Whether the "GRBL_MPG_DRO_BoosterPack" project can independently control grblHAL?

Yes, but it will need the gcode to be stored on a SD card in the controller as the MPG do not have a local SD card. Without a SD card manual control is possible and was the reason for why I made it in the first place.

I'll look at updating the Pico version to support a local SD card - and also add support for Pico W to allow wifi access to the card via ftp or WebDav.

[CHTUZKI]

Certainly. Thank you very much for your help.

There are some Chinese code comments in it, because I want to understand the running process of the code so that I can migrate to STM32. But it's still difficult for me, because I don't know what kind of data the program outputs. I hope code comments won't cause you trouble. KeypadProcessor.zip

This is my board

[terjeio]

I don't know what kind of data the program outputs

It outputs a single character based on the key pressed - or rather makes it available for reading via I2C. There are several mapping tables - from keypad scan codes to the character they represents. This is the default map.

[CHTUZKI]

I tried to use arduino to simulate MSP430G2553. I manually triggered the KEYINTR_PIN pin of Tiva-C to let the arduino slave send a character, such as "U" and "F", but Tiva-C did not respond. I debug TivaC and found that TivaC could not trigger the EVENT_KEYDOWN event.

Did arduino send the wrong data,The program will use cnc_ map_ Default() function。

[terjeio]

I have tested your version with the Launchpad connected directly to a STM32F446 grbrHAL controller with the keypad plugin enabled and it works, at least for the one button I tested - flood coolant toggle. This with a 2553 in the LaunchPad and a switch between P1.2 and P2.2. Are you testing with the LaunchPad connected to the Tiva?

Did arduino send the wrong data,The program will use cnc map Default() function。

I cannot answer that - I do not have your Arduino code. And triggering the interrupt pin manually with a switch might be problematic due to switch bounce.

The logic analyzer trace above - was that triggered by a keypress? If so the Tiva attempted a read from they keypad processor but failed - could be due to incorrect wiring?

[CHTUZKI]

Yes, I also suspect that there is a problem with my I2C connection. The picture of logic analyzer is triggered by cable short circuit. I use a wire to short P1.2 and P2.2,

I use 10K resistor as I2C pull-up resistor

[CHTUZKI]

I will continue to test, thank you for your help

[terjeio]

I use 10K resistor as I2C pull-up resistor

Perhaps reduce that to 4K7 or 2K2. Anyway - the MSP430 should respond to a read request regardless of any key beeing pressed. 0 is returned if not. Or your MSP430 has blown pins?

[CHTUZKI]

I re-ordered an MSP430G2553 from the Internet. I feel that there may be a problem with my chip. When the chip is mailed to me, I will continue to test it.

[terjeio]

One thing to be aware of with the MPS430 LaunchPad is that P1.6 is connected to user LED2 via a jumper. IIRC this may interfere with I2C operation so perhaps it is good practice to remove the jumper? BTW it worked for me with the jumper in place but then I have pull-up resistors with a lower value than you have.

[CHTUZKI]

There is no problem with the keyboard. happy. It's just a small problem. My I2C resistance may be too large. After I replace the resistance of 10K with the resistance of 4K7, the problem is solved.

It seems that my knowledge of hardware is still lacking😂

Thank you for your long help。