J-Link Programming external nRF52 : Failed to connect to target, no Idcode detected

[StephaneJ]

Hi,

I'm currently programming a custom board using a nRF52832 (using this wiring). Everything was working great but I tried to play with SAADC and maybe did a wrong configuration and/or added too much debug log. Since a previous code upload, I'm unable to connect to my board again.

I had another board with the same behavior (same code uploaded to confirm this was the code), but somehow I found a way to unlock it, I was able to upload a basic blinky example while removing the GND DETECT and switch from VDD to VDD nRF. I don't know if it related though.

But I still have one board completely locked, how to unlock it ?

> JLinkExe -device NRF52 -if SWD -speed 1518 -autoconnect 1

SEGGER J-Link Commander V6.22a (Compiled Nov 28 2017 17:57:01)
DLL version V6.22a, compiled Nov 28 2017 17:56:49

Connecting to J-Link via USB...O.K.
Firmware: J-Link OB-SAM3U128-V2-NordicSemi compiled Jul 24 2017 17:30:12
Hardware version: V1.00
S/N: 682621274
VTref = 3.300V
Device "NRF52" selected.


Connecting to target via SWD
Cannot connect to target.
J-Link>

Thanks

[Håkon Alseth]

Hi,

 

Since it detects 3.3V, it looks like you have the wiring setup correctly.

Have you tried using nrfjprog to recover the device, like this: "nrfjprog --recover -f nrf52"?

The current consumed by your custom board can also be a good indicator if there's something wrong. Have you checked this?

Cheers,

Håkon

[StephaneJ]

Hi Håkon,

Did you had the time to check my previous reply ? I'm still not able to find a solution by myself :/ 

Thanks for your support

[Håkon Alseth]

Hi,

I am sorry for the late reply.

Just as a reference, you can find how to measure current in the DK user-guide:

Measuring current

Note: if this board has been used for current measurements before, it is likely that the SB9 is already been cut. In this case, you need to short header P22.

Since you measure 2.8V on VDD_NRF, this should be fine on your kit.

Did you connect the pins "VDD" and "VTG"? This is required for external programming, if not; the SEGGER device will try to program the nRF on the development kit.

Pins "VTG" and "GND_DETECT" are actually ADC-input pins, and should not be confused with GND and VDD, as these pins will not provide power to your external board.

Also, could you make sure that you have the latest nrfjprog? latest one is v9.7.2: nRF5x-Command-Line-Tools-OSX

 

Best regards,

Håkon

[StephaneJ]

Hi Håkon,

I measured the current for both working and broken custom board, same result : 2.06V

VDD and VTG are well connected since I'm already able to program other boards.

About nrfjprog :

> nrfjprog -v
nrfjprog version: 9.7.2
JLinkARM.dll version: 6.22a

Hope this help you help me :)

Thanks

[Håkon Alseth]

2.06V? I assume you mean 2.86V?

Since one of your boards are working, the issue might be on the "bad board", it could be a loose connection or a soldering. I would recommend connecting a multimeter in series with VDD_NRF going from the DK to the custom board to see if there is any current going into the custom board.

Best regards,

Håkon

[StephaneJ]

No I really mean 2.06V with a coin cell battery.

Where would you put the multimeter on the custom board ? If I put the multimeter between VDD_NRF on the DK and the GND of the custom board I got ~0.7V for both boards (broken and working ones)

[Håkon Alseth]

Hi,

2V on a coincell means that it's dead (0 % capacity). If you look at this image, there's not much juice left once you reach 2V on a CR2032.

In order to program externally via the DK, always power your external board from the DK. This is because there are level triggered logic that is used to detect which device the SEGGER should program, and it is _very_ important that your board has the same voltage as the DK uses.

Cheers,

Håkon

[StephaneJ]

To program the external board I always plug the USB, but to measure current I thought your link mentioned that we should release the USB.

But whatever, even if I plug the USB and remove the CR2032 from the DK, I got 2.1V even on a working board. But shunting SB11 or SB12 doesn't change this value.

I don't really understand how to be sure my board as the same voltage than the one DK uses.

[Håkon Alseth]

If it works on the other board, powered at 2.1V, then you are barely hitting the required voltage for a registered logic '1' (0.7*VDD).

It's no problem to measure current when the USB is connected, it is just that you will get more noise on your measurement if you're measuring in the micro-amp area.

You must have the same VDD level on the custom board as the DK, as there is no level shifters or similar in between your board and the DK. Connect "VDD_NRF" from the DK to the '+' of your battery terminal (remove battery first) and the same applies for the GND connection.

If you follow the schematic in this thread, you effectively power your custom board from the DK to ensure that VDD_CustomBoard == VDD_NRF.

Best regards,

Håkon

[StephaneJ]

Hi Håkon,

First I removed the battery soldered on my Custom Board (aka CB). Here are some measurements I've made with my amp meter on the broken CB :

• DK VDD_NRF  -->  CB '+' = 0.280V (unsteady, keeps dropping few mV every second)
• DK GND  -->  CB '+' = 1.1V (unsteady, keeps dropping few mV every second)
• DK VDD_NRF  -->  CB GND = 0.797V (steady)
• CB '+'  -->  CB GND = 1.1V (unsteady, keeps dropping few mV every second)

And here are the same measurement on my working CB which I can program normally using JLINK :

• DK VDD_NRF  -->  CB '+' = 0.280V (unsteady, keeps dropping few mV every second)
• DK GND  -->  CB '+' = 0.5V (unsteady, keeps dropping few mV every second)
• DK VDD_NRF  -->  CB GND = 0.806V (steady)
• CB '+'  -->  CB GND = 0.5V (unsteady, keeps dropping few mV every second)

I don' really understand.

Is it because on the working board 0.806V - 0.280V ~= 0.5V   while on the broken board 0.797V - 0.280V /= 1.1V ?