Could not find AHB-AP to communicate with Cortex-M core

[nemunoz]

Update: Hi again to everyone!

After a while, I've received a new custom board and a new cable to connect through P19. But everything are not working again.

Now, in nRFgo, when I click in Segger (under nRF52 development boards) I receive a message in bold: "No device detected or readback protection may be set"

If I press the recover button, an unknown error occurss.

Maybe I have to change something, in the development board, to be able to program the custom board?

Original post: Hi to everyone!

I have been testing and programming a nRF52832DK board (Bluetooth, SPI, NFC, ...) with, at the beginning problems and learning how to fix them, and in the end everything works successfully.

I am been programming in Keil uVision 5 and using this program to upload the code compiled to the development board and nRFgo Studio to program the SoftDevice binaries.

Ok, everything works fine.

Now I have a custom board (with nRF52832) and I want to upload a simple code to test I can work with it.

But I do not know if I can upload the code in the same way that I uploaded in the development board (with Keil).

Also I do not know if I can debub in the same way (I am using J-Link RTT Viewer, selecting SWD as target).

After wiring all connections in P20 on the development board (to custom board), when I power the custom board, power on the development board and trying to upload the code from Keil (or connecting through J-Link RTT Viewer or J-Link Commander) I always receive the same error:

"JLink Error: Could not find AHB-AP to communicate with Cortex-M core. Communication timed out: Requested 20 bytes, received 0 bytes ! Could not read hardware status!"

Also, in the Keil, I receive: "No Cortex-M SW Device Found".

(If I do not power the custom board, everything works fine, because I am working with development board and this board does not detect the custom board).

Any Idea where the problem might be?

Could it be a bad wiring or something like that?

Thanks in advance!

[endnode]

Normally 99% of SWD debugging/flashing problems are wiring so first thing I would check all the wiring on nRF5x DK's P20 header. If you are sure that GND/VDD and other are having right levels but still no luck with J-Link then I would use logical analyzer and spy on SWD wires to see if at least CLK and some reasonable commands are going through. If this is first prototype of custom HW design/layout then you need to be careful if there isn't some principle flaw (once you make it work for the first time then "99%" rule applies again;)

[nemunoz]

Yesterday, I also used the P19 header with the same problem so, as you said, I think the problem would be in the wiring.

Anyway, I will answer in this post when I find a solution to the problem or If I finally check that the wiring and the custom board are not the problem.

Thank you very much!

[nemunoz]

Finally, it looks like is a wiring problem. After braiding the wires very well I have a connection! But only 20% (more or less) of the attemps work.

[endnode]

(again it just supports my experience that 99% of flashing/SWD problems are wiring problems, great to have more and more evidence;)

[nemunoz]

In the end, it was a wiring problem, more or less. In fact the problem was the nRF52832 chip wasn't welded very well.

A new custom board, with the chip welded again and re-checked, is fully working.

Thanks for your help, endnode!