This post is older than 2 years and might not be relevant anymore
More Info: Consider searching for newer posts

JLink Info: Could not power-up debug power domain.

image descriptionHi all,

until now, I used nRF51 Chips on my board. I programed them trough the nRF51DK(PCA10028). But now I would like to work with nRF52832 Chip. I have a nRF52DK(PCA10040) and an application which is working fine on the 52DK.The application is based on peripheral examples from SDK 11. SD is not used. But when I try to download the same application on my own board with the nRF52832 Chip, will get this error:

  • JLink Info: Found SWD-DP with ID 0x2BA01477
  • JLink Info: Could not power-up debug power domain.
  • JLink Info: Found SWD-DP with ID 0x2BA01477
  • JLink Error: Communication timed out: Requested 20 bytes, received 0 bytes ! Could not transfer JTAG data. DAP error while reading DP-Ctrl-Stat register.

Error: Flash Download failed - Target DLL has been cancelled.

What means: Could not power-up debug power domain? How can I fix this error?

If I want to program the nRF52832 Chip is it possible to use the Segger from the 51DK, or do I have to use the segger from DK52?

Do I have to use an external crystal for programing the chip? Because I would like to work only with the internal RC oscilator.

Voltbox_Classic 1.0 BRD.pdfVoltbox_Classic 1.0 SCH.pdf

image description

This is what I get when I want hwinfo over jlink commander

image description

image description

image description image description image description

Parents
  • have you tried shorter wires? That's a pretty non-standard debug hookup. There's a reason that debug cables are usually short ribbon cables with ground every other conductor; when running at 4MHz extra capacitance or crosstalk between them is going to degrade the signal quite badly. If you can at least get the wires short, parallel and not bundled up together in one place you might eliminate some issues.

    Basically the JLINK is able to read the registers, but it seems to be having trouble writing them, if the debug signal itself is right on the edge of good enough, that's quite possibly what you'd see.

Reply
  • have you tried shorter wires? That's a pretty non-standard debug hookup. There's a reason that debug cables are usually short ribbon cables with ground every other conductor; when running at 4MHz extra capacitance or crosstalk between them is going to degrade the signal quite badly. If you can at least get the wires short, parallel and not bundled up together in one place you might eliminate some issues.

    Basically the JLINK is able to read the registers, but it seems to be having trouble writing them, if the debug signal itself is right on the edge of good enough, that's quite possibly what you'd see.

Children
No Data
Related