Texas Instruments has released a new version of the Hercules LaunchPad.

And I have one since two weeks. These are my first impressions.



Whats new in MK II


More of everything. I've worked with the original LAUNCHXL-TMS570004 Hercules LaunchPad with TMS570LS0432 safety microcontroller at 80 MHz. And that TMS570LS04x is at the low end of the family.

A very nice controller, with loads of safety features and power - but light on the peripherals side.

The new version received a controller upgrade. The LAUNCHXL2-RM46 that I have has an RM46L852PGE at 220 MHz

It's significantly faster, has more memory, has DMA - and my two favorite changes: it has I2C and the launchpad has better prototyping features (finally we have female headers, and breakout pins at the side of the pcb).



Here are two links for a detailed comparison from TI:

Quick overview: Hercules ARM Cortex-R4 MCU LaunchPads

More detailed: Hercules LaunchPads - Texas Instruments Wiki


As a side effect, the new version is bigger. It's all those extra headers and pads that have cause that. The core schematic hasn't changed a lot.



My first exercise: test i2c


The quick start card does not mention that there is i2c (even the LaunchPad overview on the TI side doesn't list it as a capability).


You can see that in the picture above, where the marked pins only list timer and SPI functionality, no i2c.

But if you look at the controller's reference manual, it shows that these pins are i2c too.


To be certain that these pins could be used, I asked TI, and all was ok.


When you look at the LaunchPad schematic, you'll see that the SDA and SCL are available on ttwo locations:

Pin 8 and 9 of boosterpack connector J5, and Pin 8 and 9 of the new breakout pads J11 (to annoy me, they've switched them in the two connectors).


I have used the breakout pins for my test. They work well with those tiny colourful breadboards.


As test, I used one of the examples that come with HALCoGen. You will not find the i2c examples listed in the help for this RM46 controller, but the ones in the RM48 help file works without issues.

The TI example uses an internal loop - and they have not enabled multiplexing. That means that your test program will work, but the signals are not available on the controller's pins.

To make that happen, you have to do a few additional steps :

Put external pull-ups on both pins.

Enable the right multiplex mode for them:




That'll do the trick. You can then probe the signals .



The next trial is to talk to a Fuel Gauge IC. I have a Fuel Tank Booster Pack (the one that was designed by element14) at home. I am in the process of making the RM46 talk to the on-board fuel gauge.

Here's my test setup - spot the GadGet Factory Papilio Pro that serves as logic analyzer:



The first results look promising:



Here you can see the result of the RM46 starting the communication with the fuel gauge. The picture shows the scope output (insert) and the result of the Papilio Pro data capture.


This is where I got until now. I won't give up before that difficult fuel gauge ic talks back.