Converting an ADI for Simulation

[Warhog]

I've been working on this project off and on for the past year. Whenever I get frustrated with it, I put it aside until I have settled down and decided that destroying it would be a bad idea. It wasn't as easy as I had originally anticipated but I have most of the bugs worked out so I thought I might share some pics before its all back together.

 

The first area I started to rework were the indicators. When I began work on them I didn't have the same knowledge base as I do now. So my immediate thoughts were to replace the coils with servos. In hindsight, I should have left the coils/indicator in place and undisturbed as there are other ways to drive them. Gadroc had mentioned this in an earlier thread and said he may post how he reworked them electronically.

 

Here is a pic of the coils for the glide slope indicator.

 

 

The next pic is with the coils having been replaced with a servo. You can also see a coil spring that I added to preload the indicator. There is always a bit of play when gears are involved. Preloading the indicator ensures that the play will be completely eliminated.

 

Adding servos wasn't that difficult. The easy part was that all four indicators have identical mounting bases. I re-used those parts and then built 4 identical frames to hold the servos.

 

Here is the original bases that I re-used:

 

 

I added acrylic frames...

 

 

... and then glued and screwed (drilled and tapped holes) them to the original parts.

 

 

 

 

The servos were now well seated and ready to use.

 

But then I discovered a problem with swing. Servos have a swing of 180 degrees. The indicators have a swing of may 10 to 20 degrees at the very most. So re-mapping the 2 different swings in the code would reduce the precision of each indicator considerably. I wanted to use as much of the swing available in a servo to maximize the precision I would get from each servo. Hence...GEARS. I dislike using gears but this was a situation that warranted their use. And it wasn't that difficult to implement has I had a reasonable inventory of spare gears to choose from.

 

 

 

 

 

So I used gears on each indicator and although it looks a bit like an afterthought (because it was):music_whistling: it worked.

 

 

Next on the list was the ADI ball. I started with the roll as it was definitely the easiest. After removing the synchros and their gears I was able to reuse the gear from the synchro as well as the slot it occupied in the ADI frame and replaced it with a small NEMA stepper motor.

 

 

 

I did have to build an acrylic frame for the motor and that frame incorporated slots for the screws that would ultimately hold it in place. By doing that, I had some latitude in moving the motor around on the ADI frame before it was actually tightened down. It allowed me to mesh the gears together with a minimal amount of play. These gears already have a built in backlash reduction device which preloads them in either direction. Very sophisticated stuff.

 

 

 

As I was saying the roll was really easy to do. Just find a big flat spot from a synchro you removed and close to one of the main gears. Reuse the synchro gear and fabricate something to let you adjust the motor in relation to the main gear so it meshes smoothly. It requires a lot of creative thinking in how you approach this.

But what about zeroing the motor. Next pic please....

 

On the right side you see a gear with a splotch of black paint on it. I removed that gear and drilled a small hole in it. I then fabricated a small frame to hold an IR LED Emitter that lined up with this hole and installed the same on the opposite side with an IR detector. That what I use for establishing a zero reference point for the code.

 

 

Now comes the hard part, the pitch. I started by dismantling the entire ball/bracket assembly. This ADI is from an F4 and it rotates in roll pitch and direction. Inside the ball there are 2 synchros and their associated gear trains that allow movement in 2 directions.

 

 

Here you can see the 2 shafts that the ball rotates about.

 

 

At first glance I immediately realized trying to insert 2 motors into this frame was going to be extremely problematic. I decided to retrofit just the pitch axis and forget about anything beyond that. From looking at the photo you can see the synchros and the existing gear train. I removed all of the synchros and kept their gears. In some cases however, that meant breaking the synchro down so I could retrieve just the shaft as the gear was cut as part of the shaft. With some basic machining I was able to make an adapter to attach the gear to the shaft of the NEMA motor and then install this motor in the exact same location where the synchro and it's little pinion gear were previously located.

 

 

 

 

The next problem was how to zero this axis. After close inspection I discovered that there was already a zero detection device installed on the pitch axis. Its the big bulky device on the side of the frame. Inside, it has a central arm that swings about 360 degrees and as it turns it makes contact with several points that protrude into the drum. You can see the attachment points around one half of the drum. So I actually had my choice as to which point I wanted to use for a zero point.

 

 

 

One more issue I had to deal with was the wiring coming out of the ball at the shaft. It was badly damaged when I got the device so I had to install a new slip ring (thats what its called) which would provide the wiring into the ball for control of the motor and the zero detection device. You can see where all of the wires were cut at the end of the shaft in the above picture.

 

And this next pic show the new wiring. The slip ring is actually installed behind the zero detection drum so you can't see how it is installed.

 

 

 

Slip rings can be purchased on eBay. The price of these little components is quite reasonable considering what they allow you to achieve when working with devices like and ADI or HSI.

 

So I now have to end this post. I have apparently maxed out the number of pics you're allowed to use in one post. As to completing the ADI retrofit, I have made the PCB's. They are installed on the ADI and the Arduino boards and MAX487 IC's are in place and ready to be programmed. I'm just now starting to write code to synchronize the simulators movement with the physical movement of the ADI. This is my next challenge. I will post more when I complete this next phase.

 

I hope you have gained a bit of insight into what it takes to retrofit an ADI. It definitely requires some creative problem solving to get it functional.

 

If you are interested in seeing more pictures of the ADI in various states of assembly or disassembly please follow this link: http://s1168.photobucket.com/user/PitBldr/library/My%20A10C%20Cockpit%20Parts/ADI%20Retrofit%20Project?sort=3&page=1

 

Thanks for looking. I hope you enjoyed it.

[ClayM]

Very impressive John! ( I know I have said before....but WOW!)

 

Your work always pushes me to keep going. Looking forward to see some code for that beast.

 

Good luck and hope to see it in action soon.

 

Clay

 

P.S. Lets get some accurate MFDs going next!!

[WhoMadeWho]

Very nice

 

Very nice work Warhog! :thumbup: I will be digging into my ADI sometime in the future and I'm sure I'll have some of the same challenges as you. It will be fun trying to figure things out. It is amazing the engineering that goes into these things! Here's a few pictures of my ADI.

 

[bnepethomas]

Nice work John, its the mother of all challenges!

[Hansolo]

Very nice of you to share John :thumbup: I have a simulator ADI which I hope will be just slightly less complicated that what you have achieved. Really looking forward to see you get it all working :thumbup:

 

And thanks for all the photos and explanation. This Christmas present is very much appreciated. And a Merry Christmas back to you :smilewink:

 

Cheers

Hans

[Braeden108]

This is beyond cool