Odd behavior upon new print after cancel

So, as the subject line states, the Duet seems to act strangely if I pause/cancel a print and then try to start a new one.

If I notice something is going wrong with the current print and I pause and then cancel it, I will re-home the axis and then start a new print, The printer will maybe wait for the bed to heat, but then it just goes ahead and starts printing, but not even printing correctly. The head will be bouncing up and down across the build plate. It's clearly not starting the new print correctly, and it's clearly not waiting for the print head to heat up.

The only way for me to fix this is to manually power down the printer and start it back up. This happens both when stopping a print via the web interface and via the PanelDue.

Is this normal? Is there some setting I'm missing or have set incorrectly?

Thanks.

I power down my RADDS too, because the reset circuit is flaky. Don't know if you should reset the duet, but I'm sure there is some old data in the pipeline that confuses the restart.

Sound like you don't have a T0 at the start of the print gcode. You need to make sure the correct tool is selected at start of the print, put a T0 in the slicer startup gcode section. Also, if you add a cancel.g file in sys, you can add things you want done when canceling.

I second that.
I had the same behavior previously. And if you would had a look the the g-code logging, there should be a message, that you were trying to print with no tool selected.
Although I would agree, that this behavior is not 'Normal', but there're is an easy fix for it.

The addition of the T0 seems to have fixed it. Thank you all.