It's easy to get misled during a "debugging" session, being hardware or software. And often we tend to focus on things that we think must be the most reasonable culprit, and spend a bit too much time on these.
The key is the following:
- Be systematic. Test every little thing that is a possible cause of failure, however silly it seems, one by one. As long as it's factually a possible cause.
- Don't run in circles.