Status July 2020

MCP and EFIS

The MCP and both EFIS panels are fully functional and after I realized that I had forgotten to connect the GND of my electronics to the GND of the external backlight dimmer and had fixed this, the whole system is stable, also with the backlight LEDs embedded at the PCBs.

Radio panels and backplane

Since I have switched from 74HC595 (common GND) to DM13A (common power) for the signal LED drivers there was a problem with the backlighting of the radio panels because at the radio panels the signal LEDs and backlight LEDs have a common GND.
As other builders are using these PCBs I wanted a solution that works with the current radio panel PCB design.
In the latest radio backplane design I have solved this by re-introducing the powerFET, but now its input is driven by the external backlight dimmer so that all backlighting is controlled with the same dimmer.

Also I experienced that connecting the First Officer radio panels with long cables is fragile and as I need to order the PCBs with a minimum amount of five anyway I have modified the latest radio backpanel design such that a second one can be used as a passive backplane for the First Officer radio panels; the active and passive one are connected with a flatcable.
In this design I have also modified the power and ground connections for the radio panels for more reliability because I noticed that these are very important for stable MAX7219 operation. There is a separate P/G connector for each radio panel at the backplane now.
To keep the backplane compatible with the old radio panel design the P/G signals are still connected at the large IDC connector, but I designed a slightly modified radio panel where these pins are not connected and this has a separate P/G connector instead.
In the latest SimVim configurator I saw that the ADF panels can also have a TEST switch, that NAV and ADF panels also can have an OFF switch and that my ADF design was missing the ADF and ANT LEDs. As there were enough spare inputs and outputs left at my backplane multiplexers and LED drivers I’ve now included all these options to the latest backplane PCB design.
Because of the finished radio PCB’s I’ll map the ADF and ANT LED’s on the existing VHF1 and HF1 LED positions for the ADF variant of the panel.

Update: later I realized that there’s a more elegant solution for the ADF and ANT LED’s. As only 4 to 5 digits are used for the ADF displays, the top and bottom segment of the sixth leftmost digit can be used as ADF and ANT LED.
This matches more with the real location and makes it possible to have them both for the standby and active display.
I did a quick check in SimVim and it looks like this must be possible: the ADF_Mode_ADF and ADF_Mode_ANT signals can be mapped to 7-segment text and ‘_’ and ‘~’ custom text is displayed as the lower and upper segment.
Add a small mask with ADF and ANT text between the the 7-segment display and a smoke acrylic window on top of it and it may resemble the real thing.
To be tested ….

MIP and Arduino Master PCB

For easier connections I’ve moved some of the multiplexers and output drivers from the MIP PCB to the main Arduino PCB so that the glareshield switches and LEDs that are very close to this PCB can be easily connected with short flat cables: two flatcables the for CP and FO sixpacks, fire and master warning inputs and outputs and two flatcables for the CP and FO AT/AP/FMC warnings and light test.
This also made it possible to make the MIP PCB design a lot smaller.

Overhead

Initially I was afraid that there are not enough I/O connections in the system to also connect the overhead, but it is possible to connect most of its functions after all.
While still working on other parts of the cockpit I have started to design three PCBs for the three overhead panels that contain 7-segment displays: the IRS, pressurization and electrical panels. The MAX7219 drivers for these displays are controlled by the spare output pins of the MCP panel clock multiplexer, which saves an extra Arduino pin. Besides the MAX7219 drivers these PCBs also contain some input multiplexers and LED drivers for their own panel and other parts of the FWD and AFT overhead panels.
As my MIP PCB is basically an input/output expander board I am going to reuse a second one of it for the remaining overhead inputs and outputs.
The electrical panel feeds part of its signal bus coming from the main Arduino board through to this second MIP PCB.
In this way all panels of the overhead except the AFT overhead audio panel can be connected with these three custom overhead PCBs and the extra MIP board.

On my YouTube channel I posted a video some time ago about 3D printed V-slot connectors that I designed to build an overhead frame using 4040 V-slot rails.
Meanwhile I have extended the design with the same T-profiles that I used in the pedestal to attach the overhead panels to, connected with custom 3D printed connectors to 20×20 V-slot rails.
The actual shape of the frame is probably going to change, because of space limitations that I have for the two front legs, and I want to add hinges to the FWD and AFT panel boxes or the frame with rails inside them so that they can be lowered for easier maintenance.

3D printed connectors

Front panel engraving

Some time ago another cockpit builder told me that he used Troply engraving panels for his cockpit. These are ABS or acrylic panels in various colors and thicknesses with a 0.08 mm engraving layer on top.
So far I have only been able to buy them in panels of 244 x 122 mm (or two halves or four quarters of them), which makes them relatively expensive, but the engraved front panels that you can make with them look great.

A big advantage is that engraving and cutting panels with this material can be done in one go without the need to remove them from the CNC machine for painting and carefully realigning on the CNC again for engraving.
I get the best engraving result with v-shape engraving bits with 90 degree 0.2 mm tip, for cutting the Troply panels a downcut (left helical) bit gives the best results.

For the panels that don’t have a PCB I’m 3D printing a construction of a 3 mm grey back with a 1.5 mm thick PLA inlay and on top the 2.4 mm thick engraved Troply panels. In the back part there are slots that hold LED strips for the backlighting.
Because the LED strip is very close to the engraved panel, at high backlight brightness the individual LEDs shine a bit through the engraved panel; possible solutions that I want to try for this are printed paper, painting the backside of the engraved panel or less transparent material for the inlay.

Where a LED strip segment with three LED’s and resistor doesn’t fit I cut and soldered the individual LED’s and resistor. An alternative might be to use a LED strip with more LED’s per length unit.

Initially the inlay was in transparent PLA. I’m going to experiment with more opaque PLA material

The sides still must be painted

Update: while ordering a second panel of smoke grey on white Troply 2.4 mm for the overhead panels I learned that unfortunately exactly this smoke grey on white variant is not produced anymore.
The only alternative in the same color combination is the Trolase variant (intended for laser engraving), but this has 1.6 mm thickness instead of 2.4 mm, so I’ll need to make the 3D printed transparent backside for the panels a bit higher to get the same total thickness.

Gauges

I’m using 2.4 mm smoke grey on white Troply for the front panels and as the results looked so good that I bought a 1.6mm thick black on white one for the overhead gauges dials and needles.
The needles are moved with X.27 stepper motors controlled by VID6606 drivers as described at this SimVim page and I designed a small slave PCB for the Arduino Nano and four VID6606 drivers for three dual needle and ten single needle gauges.
Because both the FMC slave and stepper slave boards use the same serial Arduino pins, a jumper is available at the stepper slave board that can be used to define if both boards are daisy chained or that only the stepper board is connected.

Sketchup Make was used to design a 3D print model of the various gauges and for the dual-needle gauges and yaw damper I designed some small 3D printed gears. For engraving and cutting the dials, needles and  a 1 mm clear acrylic top glass I used the CNC.
Once everything has been succesfully tested all designs will be made available for free download at this site.

 

Single dial gauge

Top side of dual dial gauge

Bottom side of dual dial gauge

Yaw damper gauge

Yaw damper gauge internals

Work in progress

I’ve also used the same Troply 1.6 mm black on white panel for the annunciator text plates. The texts are probably a bit too white when the annunciators are off, but look great when on, better than they look in the picture, I need to experiment a bit with my camera settings to make it look more like in reality.

Blue / dual brightness annunciators

For the blue dual brightness annunciators I’m using blue transparent 3mm thick acrylic, the engraved texts are filled with white engraving filler.
A three millimeter thick opal acrylic plate is used to diffuse the light of the two LEDs and a simple two resistor network is used to support dual brightness using two output pins.

MIP accessibility

For easier accessibility of the MIP connections I have connected the MDF bottom plate of the MCP and EFIS with hinges to the horizontal beam so that the top of the MIP can be lifted and the MIP front panel rotated or even removed.

 

Above the large monitors in the MIP there is a small gap for which I designed a new narrow 20 mm I/O breakout board to which all the MIP connections can be made so that only a few flatcables need to be disconnected to completely remove the MIP front panel for maintenance.

Reliable wire labels

At internet I learned about the existence of printable heat shrink tube that can be printed with a labelwriter.
It is available in various diameters and for cables with a large diameter or with fixed connector (like UTP cables) the normal cartridges with sticker labels can be used.
The business model for the labelwriters seems the same as for inkjet printers so the printers are relatively cheap.

New MIP display bezels

For better esthetics  and somewhat larger display sizes I have replaced my original MDF bezels by 3D printed ones, based on the design found in the DIY Cockpit Sim Builders Facebook group.
Because I’m using large, unmodified monitors (the housing has not been removed) there is some room between the MIP front panel and the screen of the monitors, so I have added a 3D printed frame behind the bezel with a 2 mm slot in which a 2 mm thick clear acrylic window fits.

Slot for a 2 mm thick acrylic window

New and old bezel

45 degree rotary switches

Some of the cockpit panels have 45 degree rotary switches. Unfortunately the bigger variants of these are quite expensive compared to the 30 degree ones. There are cheaper, smaller alternatives, but the disadvantages of these are that they have thin axes and are sealed (making it very hard to drill through the axis for a combined rotary switch/encoder) and the number of positions cannot be limited like in the bigger ones. I’ve solved the last problem by inserting a small pin in the knob and milling a slot in the front panel with the CNC.

Also the pins of these small 45 degree rotary switches are quite fragile. To improve this I have designed a small breakout PCB so that the wires need not be connected directly to the pins but can be soldered on the board.

 

Breakout board for small 45 degree rotary switches. The PCB has just arrived, so it is not soldered yet in the picture

 

The MIP N1 Set and SPD REF switches also have 45 degree positions, but do require combination with a rotary encoder, so for these I have bought the more expensive large rotary switches. These switches are in front of one of my large monitors, so there is very little depth for the combined switch.
I’m trying to solve this with a DIY 3D printed rotary encoder that uses 20 tiny 1mm3 magnets and two digital hall effect sensors.

Throttle quadrant

Meanwhile I’ve learned a lot on interfacing the cockpit functions with the simulator and as a result of that I’m going to make some changes to the throttle quadrant. I had an idea to replace the incremental flaps position sensor wheel by one that uses three optical sensors and two switches for a binary encoded absolute position sensor.

But this will cost a digital pin per flaps position and actually it can be done much easier: in X-Plane joystick axes can be used for the flaps and speedbrake handles, so a simple gear with analog potentiometer can be used instead. I was wondering how in this case the exact positions can be calibrated until I realized that there is an option to edit the joystick response curve in X-Plane which can be used for this.
Also I’m going to replace the optical wheel sensor for the trim wheels by a simple rotary encoder with gear and the trim wheels continuous servo by a DC motor.

The ideas for the throttle quadrant PCB have become more clear: I’m planning an Arduino Nano for easy interface with the simulator: the 6 analog inputs are seen as joystick axes by Windows and the 14 digital I/O pins can be mapped as joystick buttons to the throttle quadrant switches.
The six analog pins will be used for both throttle handles, reverser handles, speed brake and flaps handle, 10 of the 14 digital I/O pins for trim wheels rotary encoder, parking brake, TOGA, A/T disengage, fuel cutoff handles and optional horn cutout and stab trim switches.

To control the motorization I’m planning an additional Arduino Mega that controls all the clutches, servo’s, DC motor etc. This will be an interesting job because it requires its own interface with the simulator for which I want to have a look how to exchange data with the extPlane plugin.

At the first version of the throttle quadrant PCB I want to keep several options open to experiment with: interfacing with SimVim, as a joystick or standalone with the Arduino Mega. Using servo’s for the motorization, but also the option to embed four Polulu A4988 stepper motor driver boards or two L298HN DC motor driver boards for max four DC motors.

The PCB has been ordered, so soon I can start connecting the throttle quadrant to the simulator.

The first version of the throttle quadrant PCB design

Finally the trim wheels have been 3D printed and I have designed a new foldable 3D printed handle for them. Because of the handle, the wheels are not nicely balanced, so I’m going to stick some coins with double sided tape opposite to the handle to see where to put how much counter weight to balance them.

Although not finished yet the throttle quadrant has now become part of the cockpit.

Tiller

Because of space limitations my cockpit doesn’t have sidewalls, so I needed a custom solution for the tiller.
I used a slightly modified version of the hall sensor and center mechanism that was used for the yoke roll function and designed a combination of 3D printed clamps and a piece of MDF to mount the tiller to the desktop that is next to my cockpit.