A-10C Standby Compass: Basic Functions Working


After getting the stepper motor to move yesterday I worked on the LEDs and getting the IR sensor into the mix today.  My LEDs are green ones that have the resistor built-in, and therefore are expected to be wired to my 12v power supply in parallel.  That generally went ok even with my sub-par soldering skills but there was very little room for things.  If I had made my own 3D print for the LED holder (or had the original 3D design files) I might have improved that a bit, but I made it work.

Here's the code I used for this today.

Things I learned while copying this awesome work:

  • Be careful with how wires need to route before putting heat-shrink on them, that stuff doesn't bend well.  Also consider the type of wire in use: I switched to easy bending wire for the things I had to route around tight spots.
  • If the motor spins the wrong way, swap the pins on ONE of the coils, A or B - or use the AccelStepper's setPinsInverted() function in the Arduino code to invert the DIR pin instead.
  • My stepper seems to skip sometimes when doing its startup calibration to find the nail with the IR sensor.  Seems to work ok with DCS though, so hopefully nothing is wrong.  I played with the current limiting potentiometer on the EasyDriver but that didn't seem to help much.  Left it in the middle setting (the silk screening showing which way adds current may be backwards so be aware).  This improved when I went with the 1/8 step mode that the EasyDriver defaults to.
  • My EasyDriver board requires MS1 and MS2 be grounded to get full-step action.  Before I did that the motor turned way too slow and would not align with the heading data from DCS. I ended up NOT grounding those and just setting my max-step count 8 times higher.  Was better that way.
  • Sizing that nail for the IR sensor was a bit tricky because I had to pretty much assemble the core parts to take a measurement - still got it wrong the first try.
  • I ended up using thick CA glue to glue the nail in and to attach the IR sensor to it's holder.  Hopefully that will hold up.
  • I also carefully used some of that same CA glue to provide a better friction fit for the main wheel that sits on the motor.  My 3D print made the hole way too small, and unfortunately I did not have a perfectly sized drill bit for my pin vice - so I oversized it a bit and then used CA glue to fill in back in.  I used a spare motor's drive shaft to keep the hole clear while the fresh glue setup - moved the motor in and out of the wheel center hole to keep it from getting glued in.  Very fussy.
  • I could only get one screw in to secure the 3D printed IR holder to the unit.  I may use some glue if need be, but likely not required.
  • The accuracy seems to be with 5 degrees of the real data.  Maybe that would be better with a thinner nail (or other obstruction) for the IR sensor to calibrate against.  That did improve when I changed to embrace the default 1/8 step mode from the EasyDriver.

Comments

Post a Comment

Popular posts from this blog

Updating data on the Garmin Perspective system

Image
After changing from the Avidyne system to the Cirrus Perspective (G1000) system, I had to relearn an important part of owning an airplane: how to update all the data the systems use. There's even more data in the Perspective, including AOPA directory data and more charts. In this walkthrough, I make use of the Jeppesen app for Apple OS X: JDM .  It has worked well for me, though I did need to call Jeppesen when I first set this up to get my subscriptions fixed up.  They gave great customer support when I did. The process starts (optionally) by hooking up a ground power unit (GPU).  In the case of the Cirrus, you can certainly do this without it, but I tend to hook it up just to be sure.  The systems draw about 8-10amps while updating, but you won't generally have the power on for long.  One tip I've read: power up the GPU before plugging it into the plane, at least for the one I bought.  That saves the plane from seeing any power spikes that might have occurred when t
Read more

Hiking to Rocky Peak in Steamboat Springs

Image
Did a great hike today that is not in the commonly used hiking apps.  The trail is generally easy to see, though there were times that we lost it.  There are plenty of bushes and undergrowth draping over the trail, so bring long pants. Leg gaiters might be nice too, just to protect your legs a bit. Many parts of the trail are actually drainage ditches.   Here's a KML file from our trip down - we followed the trail much better after learning what to avoid on the way up. We went right after a thunderstorm had come through the previous night, so got pretty soaked due to the plants being wet.   Found the path to the top and saw 2 moose along the way - gorgeous views for the entire trek.  2 miles to get to the top from the parking near the hot springs with about 1800' of elevation gain.  Certainly not a beginner's hiking trail, but not a problem for those with some experience (and lungs that are used to the altitude).
Read more

G1000 transistion training - emergencies - final flight

Image
For the final flight in my G1000 transition course we focused on failure scenarios. Not the "engine goes POOF" kind but the avionics kind. Normal situation This picture shows what the 2 Garmin screens show in a normal (non failure) situation. You can see we have 2 screens - the left PFD focusing on flight performance (airspeed, heading, attitude, altitude) and the right screen (MFD) focusing on engine data, mapping, checklists, and everything else. Failed PFD screen In this scenario the screen that is right in front of you (the PFD) has failed. This basically means the screen has gone dark. We simulate that by dimming the screen down to 0% brightness. After darkening the PFD we shot a GPS approach into Rock Hill. When this failure has occured you shift the G1000 into Reversionary mode. That means the info that is normally split across 2 screens is now merged into 1. It actually duplicates the same data on both screens, but since the PFD screen is no longer functioning
Read more