nrf51 wont run without debugger

[Jim Peschke]

Have an nrf51 board design that doesn't want to run without a debugger. When programmed, the debugger can run the board. When the debugger is removed, the board does not seem to run at all.

The SWDCLK line has a 12k pulldown. The SWDIO line has a 12k pullup. I suspect these pins, especially SWDIO, might be the culprit. Sometimes I can get the board running by touching SWDIO with an oscilloscope probe.

[Roger Clark]

When you connect the debugger, which wires do you connect ?

GND, SWDIO, SWCLK , 3.3V etc

Have you tried just connecting the ground and other wires, one by one and see if any particular wire makes it work

[Jim Peschke]

I connect four wires: SWDIO, GND, SWCLK, and Power (which is 2.8V for odd reasons, not 3.3V). I noticed that one of the reference designs does NOT have any external pullup on SWDIO. The manual says this pin is internally pulled up. I have a previous design where these four signals have identical wiring, and I have not encountered this problem before.

[Daniel Wang]

You have 12k external pulldown/pullup on the SWD lines?? Do you have this issue if you try to program an application that doesn’t use the softdevice? Or does the issue occur only for the BLE examples in the SDK?

[Jim Peschke]

Yes, I have an external 12k pullup on SWDIO and a 12k pulldown on SWDCLK. I am not the software developer. I know that the application we are testing on now uses the softdevice. The chip very much acts like its stuck in an undefined state until the SWDIO/RESET line is moved. I am asking the software developer to test this my manually cycling the SWDIO/RESET line.

[Roger Clark]

Have you tried removing those resistors. They are not mandatory, they are just a precaution, and are not supposed to be so necessary on newer versions of the nRF51 (though older versions were supposed to be prone to being locked in debug mode)