nrfjprog, pynrfjprog - Intro, Mac OS X & Linux now supported, New Feature

RE: nrfjprog, pynrfjprog - Intro, Mac OS X & Linux now supported, New Feature

helen — Fri, 02 Aug 2019 03:19:02 GMT

When I use MultiAPI with multi thread

thread.start_new_thread(pro_hex)

pro_hex():

with MultiAPI.MultiAPI("NRF51") as api:

.........

Sometimes an error occurs like below:

RE: nrfjprog, pynrfjprog - Intro, Mac OS X & Linux now supported, New Feature

Rick Gould — Tue, 04 Sep 2018 09:23:55 GMT

sdafasdfasdfsdaf

RE: nrfjprog, pynrfjprog - Intro, Mac OS X & Linux now supported, New Feature

JDW — Tue, 17 Apr 2018 18:56:52 GMT

is this cask still current? I'm getting an error message: "ERROR: JLinkARM DLL not found. Please reinstall latest JLinkARM."

RE: nrfjprog, pynrfjprog - Intro, Mac OS X & Linux now supported, New Feature

jdog — Fri, 10 Nov 2017 11:44:58 GMT

I had to use this to install via Homebrew:

brew tap caskroom/drivers
brew cask install nordic-nrf5x-command-line-tools

RE: nrfjprog, pynrfjprog - Intro, Mac OS X & Linux now supported, New Feature

Lee Daniel Crocker — Mon, 15 May 2017 16:25:55 GMT

It was #1, many thanks.

RE: nrfjprog, pynrfjprog - Intro, Mac OS X & Linux now supported, New Feature

David García Polo — Mon, 15 May 2017 10:57:15 GMT

@Lee Daniel Crocker

nrfjprog does not enable pinreset by itself, and I am certain of that since I am the programmer. But I can think of two ways it happens:

You build your project with the macro CONFIG_GPIO_AS_PINRESET defined. If that is the case, on the first run of the SW the SystemInit() function in system_nrf52.c (or system_nrf52840.c) file will write those values.
Your code in some way defines a const variable placed at 0x10001200 with the value. You can open yourself the hex file and take alook at the end of the code.

If those are not the case, please tell and I can take a look at it.

RE: nrfjprog, pynrfjprog - Intro, Mac OS X & Linux now supported, New Feature

Lee Daniel Crocker — Fri, 12 May 2017 16:21:53 GMT

BUG REPORT:

nrfjprog version: 9.3.1 JLinkARM.dll version: 6.12g

I haven't figured out the exact way to trigger this yet, but about half the time when programming either the softdevice or bootloader from hex file, nrfjprog pre-emptively enables pinreset without being asked to in any way (no command line, and nothing in either hex file). It's quite clear--I read address 0x10001200 before the --program statement with 0xFFFFFFFF, and afterward it contains 0x00000015.

I have had to eliminate the use of nrfjprog from my Makefiles and use Segger's JLink scripts to program my device to work around this bug. But nrfjprog is handy, and I would use it if it were fixed.

RE: nrfjprog, pynrfjprog - Intro, Mac OS X & Linux now supported, New Feature

David García Polo — Fri, 10 Feb 2017 11:55:47 GMT

The problem is that the debug port is not abkle in nRF51 to get out of debug interface mode, so some regulators and oscillators are kept on. A pin reset will stop those current drains. Perform a pin reset and measure again.

Note that this is only valid in nRF51 devices.

RE: nrfjprog, pynrfjprog - Intro, Mac OS X & Linux now supported, New Feature

Sven — Fri, 10 Feb 2017 11:18:59 GMT

At the production tests the device is powered via programmer's VCC output line. Between that line is a current meter. We have noticed that if one burns the firmware and measures the current, it is biased since somehow the programmer(nRF 51 DK) also introduces some current due to some functionality. If one cancels the session by powering on and off the programmer and no new connections are made by the host PC, one can see the real current. Is there any API call where I could disconnect or reset the debugger so that the current introduced by the after-burning process is not accumulated into current meter's reading? The disconnect_from_emu() call does not decrease it.

RE: nrfjprog, pynrfjprog - Intro, Mac OS X & Linux now supported, New Feature

wlgrd — Sat, 20 Feb 2016 06:57:55 GMT

Great package, love the MultiAPI! I would like to have a function returning what speed RTT is at, and what I can set as maximum. For example, when using the DK as debugger, what speed can I set the RTT at? Is it also depending on the processor I am debugging?

RE: nrfjprog, pynrfjprog - Intro, Mac OS X & Linux now supported, New Feature

FormerMember — Thu, 04 Feb 2016 08:49:50 GMT

@Michael Great!

@Nikolaus Thanks. What instrument did you use for this? Sorry for going off topic. Maybe you could write a blog post about the test procedure and instrument used.

RE: nrfjprog, pynrfjprog - Intro, Mac OS X & Linux now supported, New Feature

Nikolaus Wittenstein — Wed, 03 Feb 2016 17:54:56 GMT

@Michael: I'm looking forward to March! :)

@PrithviRaj: I was able to get down to about 3µA when sleeping with the radio off, which is what I was aiming for. Is there anything in particular you're eager to know?

RE: nrfjprog, pynrfjprog - Intro, Mac OS X & Linux now supported, New Feature

FormerMember — Tue, 02 Feb 2016 09:07:52 GMT

Yes, please make as much as the tools open source friendly as possible. It'll help us tweak it as per our needs and also to help others in the community. Also please bring in features such as programming the bootloader and generating the dfu init zip files in nrfjprog.

@Nikolaus, please post the findings from your sleep current tests. :)

RE: nrfjprog, pynrfjprog - Intro, Mac OS X & Linux now supported, New Feature

David García Polo — Tue, 02 Feb 2016 08:52:36 GMT

We have plans to setting certain parts of the tools code in github to improve the feedback on issues and suggestions for improvement, but due to licensing issues we will not be able to open source the nrfjprog.dll since it makes use of Segger API. nrfjprog.exe (and equivalent) could be open sourced, but it makes heavy use of the dll, so the functionality is quite simple.

Regarding the SDK, the request has been forwarded to the responsible.

RE: nrfjprog, pynrfjprog - Intro, Mac OS X & Linux now supported, New Feature

Michael Dietz — Tue, 02 Feb 2016 07:37:17 GMT

Thank you nikolaus, thats great! I will add this to the blog post in an edit. Well keep an eye on github in late march ;)