You’ve been in this situation before. The output from your test set up or analytical model is garbage. You’ve looked at all the inputs and they appear to be ok. So…what do you do now?
I’ve often found myself staring at my test rig or a new prototype or (fill in the blank). It should be working, but it’s not. I delve into the datasheets and triple-check jumper settings. I switch out hardware, hoping it’s that simple. Maybe I’m not reading this correctly. Everything looks good, yet the output is still coming through as static. I am not a happy camper.
It’s times like these I’ve found you need to really think through your assumptions. You assume your off-the-shelf components are working correctly. You assume your material datasheets are correct. But be careful; starting with incorrect assumptions can lead you astray.
Identifying every assumption is never easy. Here are a couple methods I use. I usually take a tiered approach: start by confirming the most obvious and likely assumptions (a switch is wrong, a power supply isn’t providing good power) and move down the list to the less likely scenarios (electro-magnetic interference from a cell phone, physics is not applicable here). Hopefully, this approach will lead me to my problem. If I ever get to the point where I’m questioning the laws of physics, I’ve usually missed something. Then again, perhaps not.
Other times, I might bring in a “cold-body”, someone with the right skills to understand the problem but that hasn’t been involved with the work, for a fresh perspective. In this case, the less I explain, the better; I won’t taint this person with my own assumptions. I’ll tell them that if they can quickly solve the problem and embarrass me with something I’d been overlooking, I’ll happily spend my newly found free-time taking them out to lunch. Otherwise, we might brainstorm ideas or just discuss the issues to develop a plan of attack.
It’s never easy when something is not working. Take a break…walk away…get a soda or water. Start making up your list of assumptions…confirm the easy stuff first and go from there. Find a cold body and offer to trade lunch for a solution.