Open-Source JTAG/SWD debugger with built in GDB and STM32 dev board

[esden]

Hi everyone,

You might already know or have heard of 1Bitsy & Black Magic Probe. But maybe not.

A few days ago we launched a Kickstarter for the next version of the Black Magic Probe JTAG/SWD debugger with built in GDB server. This means you don't need to use texane/stlink or OpenOCD as a middleman. To make exploring and learning about JTAG debugging you can bundle it with the 1Bitsy STM32F415 development board.

By backing the Kickstarter you are not only getting the hardware, but you are also supporting the open-source projects that make it possible. All the earnings from the campaign will allow us to spend more time developing new features and software making ARM Cortex-M and Cortex-A programming and debugging easier.

I hope this is interesting enough that you consider taking a look and maybe even supporting our project.

https://www.kickstarter.com/projects/esden/1bitsy-and-black-magic-probe-demystifying-arm-prog

Cheers,
Piotr Esden-Tempski

[thm_w]

Nice work, its already funded.

What was the green bodge wire for shown in the gallery?
If you are working with STM32 devices and already have st-link, does black magic probe provide any advantages?

High speed data interface to the Device Under Test 4.5MBit

Do you know how this compares to st-link v2?

[esden]

@thm_w Thank you very much. It is really great to see so many people come together to support our project.

The bodge wire is related to the RST line circuitry on the V2.1a of the board, it has already been fixed on the V2.1b revision that is on the way. For more background: RST has to be driven as an open-drain output. The version V2.0 of the board used a bidirectional level shifter. It was able to detect the direction automatically and the way it was implemented allowed us to translate the open-drain signal too. Because how "magical" the part was it was also not a very strong drive. When developing the updated version V2.1 I decided to use dual supply transceivers to improve it's performance when used with high impedance JTAG/SWD lines. For version V2.1a I forgot that the new drivers can not be used as open-drain drives. The green bodge wire was the first attempt to fix the issue. By now I have tested a circuit that uses two N-FET to drive and sense the RST signal. This way we have nice and clean RST signal, the STM32 GPIO is not being stressed if someone decides to short the RST to 5V, and we can sense that the target has pulled down its RST line even if we are connected to a 1.7V device.

I have posted some detailed pictures of the new V2.1a and V2.1a with the new circuit on Twitter:
https://twitter.com/esden/status/807036692368474112
https://twitter.com/esden/status/807384774163603457

There are a few advantages of the Black Magic Probe. BMP is open-source, meaning that you can look inside it if you need or want to. We are getting support for new ARM Cortex-M based chips on a regular basis, so you are not limited to just the STM32. We have preliminary support for Cortex-A this will result in the ability to use the probe with Raspberry PI and Beagle Bone Black and many others. The Black Magic Probe also supports JTAG not only SWD, because not all microcontrollers use SWD. The GDB server is implemented on the probe itself, this means we do not use some proprietary protocol to talk to your debugger software. All you need to do is fire up GDB and connect to the virtual serial port of the Black Magic Probe, no special setup necessary. It is physically small, you can plug it into your hardware even if it is buried deeply inside your killer robot. We have equipped the Black Magic Probe V2.1 (this is the version the Kickstarter backers will receive) with dual supply level shifters that make it possible to connect to targets that run on voltages as low as 1.7V and as high as 5V. The transceivers are more robust than the STM32 GPIO meaning that if the BMP is connected wrongly to the target it is less likely you will hurt either the Black Magic Probe or your target.

Regarding your last question. I have pulled out a nucleo board and measured that the ST-Link is also running the SWCLK line at 4.5MHz.

I hope I did not miss anything. Let me know if you have any other questions.

Regards,
Piotr