Power Measurement Tool 'Unable to connect to and initialize probe'

[mikkoroth]

Hello,

I am trying to conduct a series of power measurements using the Power Measurement Tool plugin.

The first measurement using the plugin after connecting the USB and downloading the CMSIS-DAP (default) firmware works fine. When trying to start a second measurement by pressing the "start data collection" button in the plugin, the plugin reports "Unable to connect and initialize probe". Connecting to the probe via the Redlink Server seems to work though. Could this be a bug in the plugin? The only way I am able to conduct a new measurement is to power cycle the board (disconnect USB cable, reconnect and re-download the firmware).

I have tested this on two development boards with the same result: LPCXpresso54114 (OM13089) and LPCXpresso54102 (OM13077). I'm using the LPCXPresso IDE v8.2.2 on a Xubuntu 16.04.

Is there a way to avoid power cycling the development board after the first measurement to conduct multiple measurements in a row?

Best Regards,

Mikko

[brendonslade]

Hi Mikko,

The power measurement works via a separate virtual USB device, so is independent of the CMSIS-DAP probe (this is why Redlink could still work if power measurement doesn't). You shouldn't have to do any rebooting to re-run the power measurement, so something isn't right.

How are you booting the probe? Are you starting a debug session with the DFU jumper installed or are you using LPCScrypt to flash the on-board debug probe then using the board with the DFU jumper removed?

Regards,

Brendon

[mikkoroth]

Hello Brendon,

Thanks for your reply. I start the debug session with the DFU jumper installed: I use the "Flash Program" for flashing the CMSIS-DAP probe and program code onto the board, and then run the measurement tool. But your hint about the "separate virtual USB device" seems to have solved my problem: I forgot to mention that I run the 32-bit Xubuntu in Virtualbox. Now I installed LPCXpresso on my 64-bit host and could re-run the power measurement multiple times on one of the boards. I did not think about that as the the other connections (Redlink, debugging etc) work without problems from within the virtual machine.

Best Regards,

Mikko