I had similar problem that I traced to rf feedback getting into the Nano controls. That does not mean that it is your issue, just one other thing to test. That fact that you changed the IRF510 is a potential pointer to the RF feedbackissue, as it is very lead and postion orientation.
Have you tried into a dummy load so that htere is minimal RF in the shack? If you do not have the issue on the dummy load, then I would guess feedback is the issue. If the problem still persists, then I would recheck the wiring and the solder joints on the devices that you replaced.
I would also use the DMM to measure the current to the PA (brown wire) and do the bias adjustments.
If all that fails to point you to the problem, then may need to build an RF probe to go with your DMM and trace the RF from TP2 through each TP until you get to the antenna output. (TP2 to TP7). You should see a signal growing in strength from roughly 200mv PP to over 50v PP at the antenna. Here is a link to scope traces of a version 5 (has the same power circuits) https://groups.io/g/BITX20/message/72786