Modular Panel WIP

Hook indicator lamp is finished in CAD and all box mods to house the components are completed. Now onto printing it.

This is a pretty awesome project. I have been considering a switch box of my own and this thread has given me lots of ideas!

How are you doing the electronics. I see that all of the boxes are connected by a 'bus'. Is that encoded somehow? In my box I was (am) considering shift registers to handle all of the buttons but yours has waaay more than what I have at the moment (30 buttons). Curious about how you are collecting all that data electronically.

Plus, would love to eventually see the STL files in the future, however you want to make them available.

Awesome!

STL's + BoM would be doubly awesome.

Thanks dwpenny.

The electronics side is very simple thanks to DCS-BIOS. It handles all master/slave comms across a RS-485 bus. With the time it takes to design and print the controls and boxes I'm glad the software and electronics side are straight forward.

I feel just putting the STL's out there wont do my project justice, it's going to need some effort to describe how I put it all together. It will happen though.

cheers

John

This is a pretty awesome project. I have been considering a switch box of my own and this thread has given me lots of ideas!

 

How are you doing the electronics. I see that all of the boxes are connected by a 'bus'. Is that encoded somehow? In my box I was (am) considering shift registers to handle all of the buttons but yours has waaay more than what I have at the moment (30 buttons). Curious about how you are collecting all that data electronically.

 

Plus, would love to eventually see the STL files in the future, however you want to make them available.

Cheers!!

Awesome!

 

STL's + BoM would be doubly awesome.

Well Done!

What printer are you using? I've got the Formlabs Form 1+, which I love, and Solidworks 2018 sp2.0, which often makes me want to switch to Fusion 360...

No electronics yet, just manually powered up the hook-down indicator LED. I have some 8mm orange LEDS on the way, for now I'll use a red LED. The indicator lamp has gone through many revisions to reach something that is easy to print and requires no paint for masking. This is design four, the fifth is on the printer right now. This one has a clear/frosted core with a solid black shell.

The empty boxes around it are rejected older designs and are used as place holders. It's all sitting in the new right hand side vertical stand.

cheers

John

Cheers mate :)

I have a Zortrax M200. I've rambled about this printer in the past :), it's a real work horse, just keeps printing and printing without fault.

No way I can afford Solidworks, Fusion 360 has everything I need and it's free.

Well Done!

 

 

What printer are you using? I've got the Formlabs Form 1+, which I love, and Solidworks 2018 sp2.0, which often makes me want to switch to Fusion 360...

I am almost finished assembling my Prusa MK3 and cannot wait to start printing.

Sent from my iPhone using Tapatalk

Nice, great printer!

I am almost finished assembling my Prusa MK3 and cannot wait to start printing.

 

 

Sent from my iPhone using Tapatalk

Cheers mate :)

 

I have a Zortrax M200. I've rambled about this printer in the past :), it's a real work horse, just keeps printing and printing without fault.

No way I can afford Solidworks, Fusion 360 has everything I need and it's free.

Struggling right now with Solidworks, working on designing the tooling for my client to restart manufacturing the Stewart 51D...

http://www.stewart51mustang.com/

and my Solidworks is kicking me in the junk with broken sketches to files that I had finished months ago...

Arrgh.

Doesn't Fusion 360 also do multi-axis CAM toolpaths?

Paths are real finicky if you move files around Ive found for assemblies.I have same problem with ProDesktop 2000ic especially with real old files.With sketches I export them off as a Dxf file as a backup instead of trusting the SW sketch.Im doing something crazy as well.

You're operating in a completely different league than my 3D print boxes.

I think I heard you need the professional version for that feature, but now it's in all versions.

Struggling right now with Solidworks, working on designing the tooling for my client to restart manufacturing the Stewart 51D...

 

 

http://www.stewart51mustang.com/

 

 

and my Solidworks is kicking me in the junk with broken sketches to files that I had finished months ago...

 

 

Arrgh.

 

 

Doesn't Fusion 360 also do multi-axis CAM toolpaths?

Here is the final design for the hook lamp. I'll repeat this method for the advisory lights beneath it. The top HOOK plate is bonded to the clear section by changing the material from clear to black during the final layers. It's then slotted into a black case to prevent light leak.

cheers

John

Do you use dual extruder or just pause and change?

It's a pause in the slicer then a material change during printing.

Everytime i think this thread cant get better....it gets better. The innovation is crazy.

Can I just say that it is truly inspirational watching the development of your panels in this manner. Early on you were wondering if this was the place to keep such a build log, I would say that it really shows the effort and time you put into developing this :D. It is also nice to see you reply to other people's questions and remarks, as it also helps them further their own build stuff.

I cannot get enough of going through these posts, as they outline kind of the same process I wish to go through for building my own 'simpit': I wish to go a bit more frankenstein however and build a number of panels that really would not make sense for anyone else, like having a number of small displays to the side of my monitor with airspeed and g loading.

To the end of my own insatiable curiosity (and hopefully without infringement on you hypothetically selling these panels), do you mind if I shoot you a few questions?

1. Did you use two HE sensors for each lever type switch, with a magnet in the handle? Or did you put a magnet on each end (with different magnetic properties) with a HE sensor in the handle?

2. {Probably related to 1.} How did you do the magnets for the parking break? You put magnets (with different properties) on the outside, with a HE sensor in the sliding shaft?

3. How did you work out the gearing to get from the hornet's ejector wheel to correspond with your absolute rotation switch?

Again, thanks so much for your effort in the development and documentation of this!

Thanks nicka :), I do get frustrated with the pace, I want to get so much done but can't find any extra time.

Everytime i think this thread cant get better....it gets better. The innovation is crazy.

Thankyou Nanne for your comments, really keeps me motivated! I wasn't sure if people wanted a build style thread, as I didn't receive a response to the question, so I kind of went middle of the road. I think showing a CAD diagram and the completed pieces with some construction detail is enough detail without bombarding everyone with redundant pictures.

1. Yes two HE sensors with individual rare earth magnets, this solution lends itself very well to these boxes as they're only 3-4 cm deep. Basically the moving part gets the magnet side. It's also far cheaper than buying regular switches and offers great flexibility and reliability.

2. Yes the sensor at the base of the shaft to sense the PULL action and one on the side to sense the ROTARY action. I was going to use a state-machine to detect the valid positions then send the necessary commands to DCS, but the ROTARY command doesn't work yet. I'll have to circle back to this later.

3. Gearing was determined from simple division, 180deg swing on the dial needs to swing the rotary switch 120deg (30 deg per position). I then started with the smallest diameter and tooth size for the smallest cog then determined the number of teeth for the largest cog. Basically, Num teeth small cog * 180/120 = num teeth large cog. The teeth need to be the same size so the diameter of the large cog is determined by the required number of teeth.

Thanks again.

cheers

John

Can I just say that it is truly inspirational watching the development of your panels in this manner. Early on you were wondering if this was the place to keep such a build log, I would say that it really shows the effort and time you put into developing this :D. It is also nice to see you reply to other people's questions and remarks, as it also helps them further their own build stuff.

 

I cannot get enough of going through these posts, as they outline kind of the same process I wish to go through for building my own 'simpit': I wish to go a bit more frankenstein however and build a number of panels that really would not make sense for anyone else, like having a number of small displays to the side of my monitor with airspeed and g loading.

 

To the end of my own insatiable curiosity (and hopefully without infringement on you hypothetically selling these panels), do you mind if I shoot you a few questions?

 

1. Did you use two HE sensors for each lever type switch, with a magnet in the handle? Or did you put a magnet on each end (with different magnetic properties) with a HE sensor in the handle?

 

2. {Probably related to 1.} How did you do the magnets for the parking break? You put magnets (with different properties) on the outside, with a HE sensor in the sliding shaft?

 

3. How did you work out the gearing to get from the hornet's ejector wheel to correspond with your absolute rotation switch?

 

Again, thanks so much for your effort in the development and documentation of this!

Hi,

Wiring complete for the hook module.

The suspension design has been improved for the right hand air module. Much more stable and more clearance.

cheers

John

Cheers for your previous answers, they have been helpful and definitely insightful as to what goes into designing / developing this stuff! :D

I hope you don't mind me shooting another couple questions your way:

1. How did you design and print the brake pressure indicator? Multiple colour extrusion? (It looks really good!)

2. Any particular reason why you used VGA connectors for the implementation of the RS-458 (MAX-485/-487)? Would that not leave you with a number of spare connections on the VGA connector, seeing as you only need 4 wires (VCC, GND, A, B)?

3. I presume you use the optocoupler in conjuction with the (mos)fet to avoid reverse current or overvoltage, seeing as the fan can source these issues? Would a capacitor over the poles of the fan not also suffice to suppress any unwanted voltage/current spikes? I mean, using an optocoupler is a much nicer solution, I am just wondering if there are any other advantages to using it.

4. How much of a pain was it to get those louvers from the ACU correct in Fusion? I cannot imagine that has been a painless and easy process (or I am just really crap at Fusion, which given my track record seems more likely)

Hi Nanne, happy to answer any questions.

1. It's much simpler than that, colour laser printer. I did the design the artwork in Fusion360, then printed it out and stuck it in. :)

2. I use 6 pins on the DB9: +12v, Aux, +5v, GND, A and B. I wanted a push/friction connecter, the DB9 is perfect for that.

3. I was aiming for near total isolation without that nasty reverse EMF stuff getting into my MCU and logic bus. I went the optocoupler driving a MOSFET route. However for my right hand air module I've bought fans which accept a TTL level PWM signal directly so only need the optocoupler. I'll be building/testing this shortly.

4. These worked first time, the beauty with CAM is once you know the tolerances of your machine most designs are successful on the first iteration. You can swing parts in Fusion360 to make sure everything mechanically works without collision. Sometimes some minor sanding is required as the 3D printer leaves behind dags. Also you need to design around the 3d printer's weaknesses. For example, I could get near perfect louvers if I oriented them vertically but then they wouldn't be as strong. So I printed them horizontally, but then I'd lose resolution in the hinges. This meant the tiny hinges needed to be sanded smooth to restore the curvature. I also splash on some acetone to these working joints to improve the inter-layer strength.

One other thing, printing in ABS has given me much better flexibility than PLA, it's not as rigid so allows for snug fitting parts, allows for glue-welding with acetone, can be smoothed with acetone, can be strengthened with acetone, can be more easily sanded, is better for mechanical joints and can handle much higher temperatures. There's many things I couldn't do if I was using PLA.

cheers

John

QUOTE=Nanne118;3714482]Cheers for your previous answers, they have been helpful and definitely insightful as to what goes into designing / developing this stuff! :D

I hope you don't mind me shooting another couple questions your way:

1. How did you design and print the brake pressure indicator? Multiple colour extrusion? (It looks really good!)

2. Any particular reason why you used VGA connectors for the implementation of the RS-458 (MAX-485/-487)? Would that not leave you with a number of spare connections on the VGA connector, seeing as you only need 4 wires (VCC, GND, A, B)?

3. I presume you use the optocoupler in conjuction with the (mos)fet to avoid reverse current or overvoltage, seeing as the fan can source these issues? Would a capacitor over the poles of the fan not also suffice to suppress any unwanted voltage/current spikes? I mean, using an optocoupler is a much nicer solution, I am just wondering if there are any other advantages to using it.

4. How much of a pain was it to get those louvers from the ACU correct in Fusion? I cannot imagine that has been a painless and easy process (or I am just really crap at Fusion, which given my track record seems more likely)

Thanks ever so much! I really appreciate the detailed explanation of what is going on inside, it really helps to get things into perspective (and fuels me wanting to build my own panels for DCS).

I am having some issue wrapping my head around the RS-485 (and respective chips MAX-485/-487) though. I have ordered the chips and they should be here next week, so hopefully having the hardware will un-complicate things a little.

The issue I am struggling with is that of communication between the Arduino's, with respect to the DCS Bios programming language. Because of the 'slow' data rate, you obviously want to have your transmissions be as concise as possible (to save on transmission time), and thus avoid having any lengthy descriptors attached to your values from the Arduino's. Also only transmit when there is a change, presumably.

Does that than mean that you have labelled each slave device (Arduino Nano) with a simple descriptor, like AA or A1, and then sort everything out at the Master Mega? That would also make multi-aircraft compatibility easier to do, right?

Hi Nanne,

I haven't peered under hood but I assume the slaves are only identified by their assigned address. I imagine the packet format is something like,

<SLAVE ADDR><PAYLOAD LENGTH><PAYLOAD:Name,Value>

I've got 30+ switches on one 250kbps bus and 12 on another, I'm not sure how much messaging the Mega can handle but so far so good. There is a third bus I haven't used yet. If I reach the limit of the Mega I'll create a new one.

DCS-BIOS makes things very simple,

Code for Master,

#define DCSBIOS_RS485_MASTER

/*
 Define where the TX_ENABLE signals are connected.
 You can connect up to three half-duplex RS-485 transceivers.
 
 Arduino Pin             RS-485 Transceiver Pin
 TXn ------------------- DI (driver input)
 RXn ------------------- RO (Receiver Output)
 UARTn_TXENABLE_PIN ---- /RE, DE (active low receiver enable, driver enable)
 
 If you have less than three transceivers connected, comment out the corresponding
 #define UARTn_TEXENABLE_PIN lines for receivers that are not present.
*/
#define UART1_TXENABLE_PIN 2
#define UART2_TXENABLE_PIN 3
//#define UART3_TXENABLE_PIN 4

#include "DcsBios.h"

void setup() {
 DcsBios::setup();
}

void loop() {
 DcsBios::loop();
}

Code for Gear Slave,

#define DCSBIOS_RS485_SLAVE 11

/*
 The Arduino pin that is connected to the
 /RE and DE pins on the RS-485 transceiver.
*/
#define TXENABLE_PIN 2

#include <DcsBios.h>

const byte gearLeverPins[2] = {7, 6};
DcsBios::SwitchMultiPos gearLever("GEAR_LEVER", gearLeverPins, 2, true);
DcsBios::LED gearLeverLight(0x7478, 0x0800, 9);

DcsBios::Switch2Pos gearSilenceBtn("GEAR_SILENCE_BTN", 5);
DcsBios::Switch2Pos gearDownlockOverrideBtn("GEAR_DOWNLOCK_OVERRIDE_BTN", 4);


void setup() {
 DcsBios::setup();
}

void loop() {
 DcsBios::loop();

}

cheers

John

Thanks ever so much! I really appreciate the detailed explanation of what is going on inside, it really helps to get things into perspective (and fuels me wanting to build my own panels for DCS).

 

I am having some issue wrapping my head around the RS-485 (and respective chips MAX-485/-487) though. I have ordered the chips and they should be here next week, so hopefully having the hardware will un-complicate things a little.

 

The issue I am struggling with is that of communication between the Arduino's, with respect to the DCS Bios programming language. Because of the 'slow' data rate, you obviously want to have your transmissions be as concise as possible (to save on transmission time), and thus avoid having any lengthy descriptors attached to your values from the Arduino's. Also only transmit when there is a change, presumably.

 

Does that than mean that you have labelled each slave device (Arduino Nano) with a simple descriptor, like AA or A1, and then sort everything out at the Master Mega? That would also make multi-aircraft compatibility easier to do, right?