Lessons from An Electronics Noob

One of my great heroes, Jeri Ellsworth, has spoken publicly (and made a great video) about the importance of trying and failing in learning electronics.

I personally am tempted to generalize her statement to everything else. In fact I would say that is the great genius of Agile methodologies and Test-Driven Development in particular.

On Saturday, I got my "test preemie" working---a success.  Monday I started on the incubator, thinking it would take about 3 hours.

And failed, and failed, and failed some more.

I had it all working on the breadboard---easy as pie.  But breadboards are fragile. I don't want to go to Mini Maker Faire with a breadboard.  I wanted to solder up a prototyping board to make a sturdier solution.

Using tips suggested by this gentleman for prototyping, I laid out my protoboard, and bent the components into place.

I soldered it up, and it didn't work. I had missed a connection.  I fixed it, and tried again, and it didn't work. I realized I had soldered the MOSFET in backwards.  This is easy to do---the little ICs are little on 3/10ths of an inch wide and they aren't "marked" in anyway---you have to look at the data sheet to which pin goes where.

So I painstakingly desoldered it. Desoldering is harder than soldering; you have to use a "solder sucker" and then "soldering braid" to wick up the molten solder.

And, if you bent the pins of the components over as hinted above, it is even harder to get them out.

Then I carefully resoldered the MOSFET back in --- the WRONG WAY AGAIN. Imagine a face palm that evacuates the brain pan.

So I carefully desoldered the MOSFET again.  With each operation, my board was getting messier and messier.

Then I put a clean MOSFET in---and it worked!  For about 4 minutes. Then I broke one of the pins leading to the Arduino off.  No worries---I could just solder a new lead in place. To make it sturdier this time, I stripped a solid 24 gage wire and soldered it in.

Then, nothing.

It seemed to draw current but wouldn't heat the little 8 Ohm resistance cloth to keep the baby warm.  So, to try to isolate it, I took it over to my power supply, away from the Arduino, and tested the MOSFET and heater load in isolation---or so I thought.  It soaked up every amp my power supply would give and converted into heat---INSIDE THE MOSFET, which began to smoke, right before it melted the plastic off my leads.

Trying again and again, it behaves as if there is a short INSIDE the MOSFET.  I don't know if this possible, or if somehow I have a stray gob of solder which neither my eye, nor the multimeter can see.

I spent at least 10 hours debugging this, and am completely stuck.  So, I am retreating but not surrendering---I ordered some new MOSFETs and will build a completely new board when I get them.

Until then, I will try to do the same thing a "relay", a physical switch which does what a MOSFET does more reliably.

I have learned:

• This stuff is hard and takes practice to develop basic skills.
• It is easy to make mistakes which make things very confusing.
• You could probably learn a lot from watching someone experienced do this. YouTube videos are the best I have in that department.
• Breadboards are fragile.
• Soldering sucks.
• This is an excellent way for a grown man to feel stupid.