Crumar D9U: What went wrong? (Day 3) - Sand, software and sound

People who build stuff are optimists. Even just a little.

After slaving over a hot soldering iron (or bit of code), there comes the moment of truth. Of course, we all hope and believe that everything will run just fine when power is turned on.

Thus, I was a little bit bummed when I ran my drawbar test sketch (Arduino program) and saw that the second drawbar was unresponsive.

Hey, PJ, how did you debug and fix this thing?

First and foremost, I want to emphasize the importance of diagnostic tests. My test sketch told me conclusively that the second drawbar was busted. The test not only said, “Houston, we have a problem,” but the test also told me where to look for the problem — the second drawbar. Explicit testing is much better than trying to test and debug hardware with the application sketch itself, i.e., the sample sketch provided by Crumar.

Knowing that the second drawbar was bunged, I whipped out my digital multimeter and did a few simple electrical tests. I usually assume good components, but it was easy to check the second slide pot. Yep, the meter read out the expected resistance. Then I did quick continuity and resistance checks back to the appropriate Arduino analog pin. Thank heavens for the sample sketch because I quickly worked out the pin map (Arduino Pro Micro):

          TX o    o RAW
          RX o    o GND
         GND o    o RST
         GND o    o VCC
         SDA o    o A3    Draw4
         SCL o    o A2    Draw3
  Draw5   A6 o    o A1    Draw2  <---
             o    o A0    Draw1
  Draw6   A7 o    o SCLK
             o    o MISO
  Draw7   A8 o    o MOSI
  Draw8   A9 o    o A10   Draw9

I numbered the drawbars (left to right) starting with one. The Crumar schematics numbers the drawbars from zero. [Oh, well.]

This was a good time to check the batteries in the multimeter. Whoops, 2017. Good, fair or poor, I replaced the batteries with fresh ones.

I didn't find anything out of the ordinary, so I began the usual prayer, "Dear Lord, I hope I didn't blowed up the chip." I did not relish the thought of replacing the Arduino Pro Micro -- all 24 pins of it.

My soldering skills are good, but not perfect. Plus, the large holes and the excessive amount of solder consumed by the slide pot leads immediately made me suspect a bad soldering joint.

In order to test this theory, I tried jumpering the voltage from the first drawbar slider to the Arduino A1 pin while watching the sketch's output in the IDE Serial Monitor. Perhaps an internal pull-up (or something) didn't change the A1 signal or the output from the sketch.

Then it was time to play "which of these isn't like the others" with the multimeter and with the power on. Ah-ha. The voltage at the SIGN #2 pad did not change when sliding the pot to a new position. The rest of the SIGN pads responded correctly. Not conclusive evidence for a bad solder joint, but more hope than a blown up Arduino.

The next step was to fire up the iron and retouch the pot leads for the second drawbar. It takes longer for the iron to heat up than to do the touch up. Damn, these leads suck down solder.

After letting things cool, I plugged the Arduino back in and voila, the second drawbar now responds correctly. Thank you, test sketch.

When testing pot resistance, I noted some variability in resistance across the 9 pots. I hope the variability doesn't affect behavior when I get down to transmitting MIDI values. Maybe everything will act like a well-worn B3? Sometimes higher resistance is due to a marginal solder joint, but I'm not anxious to touch up every pot connection lest I introduce a worse problem. More testing ahead.

If you landed here, check out Day One in this series of blog posts about the Crumar D9U Drawbar Controller kit.