Group Actions
Engagement
  • Author Author: Jan Cumps
  • Date Created: Date Created
  • Views 56 views
  • Likes 6 likes
  • Comments 5 comments
Related
Recommended

switch up to 5V and 8 mA with MSPM0 Open Drain IO

Jan Cumps

The Texas Instruments MSPM0 microcontroller has a few open drain GPIOs. TI calls them ODIO. These are 5V tolerant, and can switch up to ( see   comment below) +-8 mA. The main purpose is to support multi-level IO. 
They can do general low side switching too. That's what I'm checking out in this post. I use a ODIO to blinky an LED.

image
image: EasyL1105 controling an LED with open drain IO

The standard GPIOs can source 3.3V, 6 mA (with caveats). Can be used to modestly blink efficient LEDs. But if you want to control a small LCD's backlight, this 5V, 8 mA capability can save a few components.

From TI's documentation:

image

"... the open drain IO only implement a low-side NMOS driver and no high-side PMOS driver. The 5V-tolerant
open drain IO are fail-safe and may have a voltage present even if VDD is not supplied"

"ODIO can be used to implement communication with a 5V device."

image
source: MSPM0 L-Series MCUs Hardware Development Guide and MSPM0L110x Mixed-Signal Microcontrollers datasheet

What does this project do?

It's the blinky. TI calls it gpio_toggle_output. It blinks an LED every second. The only change I made, is to define open drain pin PA0 as the LED pin. The code is just a simple loop that toggles the pin, then delays half a second.

But what it actually allows you to do, is test that the MSPM0 ODIOs are capable to switch a 5V powered load.

Hardware resources

LED1 PA0 Led with 330R resistor in series
Powered by 5V

The circuit is simple. A red LED with a 330V resistor in series. When PA0's gate is driven high, the current in this circuit, and through PA0's channel, will be around 10  7 mA.

I used the 5V of my EasyL1105 kit. If you use a LaunchPad: they have a 5V supply too.

image

Testing

Program the MSPM0LL05 with the attached CCS project code. If you have a debugger, you can do this from CCS. I've also provided a .hex firmware that can be used with  ' bootloader utility.

Code highlights

In SysConfig, PA0 is set up as an open drain output.:

image

The code is extremely simple:

/* This results in approximately 0.5s of delay assuming 32MHz CPU_CLK */
#define DELAY (16000000)

int main(void)
{
    SYSCFG_DL_init();

    while (1) {
        delay_cycles(DELAY);
        DL_GPIO_togglePins(GPIO_GRP_ODIO_PORT,
            GPIO_GRP_ODIO_PIN_ODIO_0_PIN);
    }
}

What's next? PA0 can be driven from a PWM peripheral. That allows you to dim a small LCD's backlight without external transistors.

Have fun!

ccs project adapted to EasyL1105 (with bootloader compatible .hex firmware): EasyL1105_odio_sink_20251118.zip

Related posts

Parents

  • You're quoting the 20mA from the absolute max ratings table. The IOL is 8mA (see the table further down the datasheet).

    That 20mA basically comes down to a dissipation limit for the output transistor (20mW is quite a lot in a small area on-chip) - the metallisation and the bond wires will manage much more than that before fusing.

    Your example is no problem at all because you're only running about 7.5mA through the red LED (not 10mA - you haven't taken into account the rDS-on value for the output MOSFET), and you're pulsing it at intervals which reduces the average dissipation even further. But sinking 20mA DC for a backlight with it wouldn't be good design practice, though it could work for a personal project (the ratings are somewhat conservative and you're not going to be operating it in an ambient temperature anywhere near the top of the spec range, anyhow).

Comment

  • You're quoting the 20mA from the absolute max ratings table. The IOL is 8mA (see the table further down the datasheet).

    That 20mA basically comes down to a dissipation limit for the output transistor (20mW is quite a lot in a small area on-chip) - the metallisation and the bond wires will manage much more than that before fusing.

    Your example is no problem at all because you're only running about 7.5mA through the red LED (not 10mA - you haven't taken into account the rDS-on value for the output MOSFET), and you're pulsing it at intervals which reduces the average dissipation even further. But sinking 20mA DC for a backlight with it wouldn't be good design practice, though it could work for a personal project (the ratings are somewhat conservative and you're not going to be operating it in an ambient temperature anywhere near the top of the spec range, anyhow).

Children