I am so excited! Need help with 2DOF restoration

Hey guys I think I found the right place! I am a military pilot and build my own simulators, but until now they have been fixed. I am very picky that the aircraft operate and feel like they should and I was researching putting the F-18 simulator I built, to be placed onto a full motion platform, when I discovered these old commercial simulators already on a full motion platform. These simulators have been moving from warehouse to warehouse for years after the previous owner’s passing. They are a really cool “barn find” I have two of them and various spare parts, but as you can see in the pictures they need complete restoration. My plan is to completely restore the f-16 with all new avionics and accurately duplicate an actual f-16 panel with modern software. On the other I will place my already built F-18 as soon as I can get the full motion simulator running from the simulation software. Below are pictures of what I am starting with. I am completely new to full motion simulation so I will need to be guided by the experience here to get this thing up and running. I am really excited about the possibility of a Department of Defense grade simulator when I am finished. So the first thing I need to learn is how to interface my Lockheed Martin P3D software to this simulator to be able to control the pitch and roll axis via the software. Below are photographs of what I am starting with. Any help with this project will be greatly appreciated. Thanks for the help, Lee

UPDATE: I figured out how to use an Arduino mega 2560 to send signals to the motor controller and I am able to run the original motor on a test bench with the Arduino and respond to code. Next step is to figure out how to get the motor to respond to the simulator software. Any advice on how to get P3D to drive the motors?

What type of motors are they ?

Are they dc motors or stepper motors ?

Hey! Thanks for the interest. They are stepper motors with matching controllers and I have them running with an Arduino circuit so I am thinking that next step is to get Simtools. Am I on the right track?

See if you can find out what type of signal they use for control

There may be no position feedback because they are stepper motors and they move in steps

I am not sure how to answer the question, they are stepper motors and i have them responding to the Arduino Mega so I assume that the motors are DC based on the cover plate on the motor above and they respond to a digital signal, thus the response to the Arduino circuit. Is that the question you are asking? Thanks again, Lee

@Lawwolf7567 - I could be wrong so please double check, but it could be that you might want to just scrap the entire motor drive and controllers and rebuild with more modern and more powerful parts.

Reason I say is those are steppers and don’t appear to have feedback like Gadget says. Those rely on reliable driving of the platform to not get out of register with where the platform actually is. The system would need to go through a calibration each time it fires up. That’s no big deal as a number of other approaches also use startup calibration.

The issue I see is just low power and slow response. Those controllers appear to only be rated at between 100 and 300W depending on input voltage (not bad depending on how many motor outputs each box has), but the real oddball is the motor data plate that looks like those are rated at only about 15W (7A x 2.24v). That’s maybe per phase but it’s still pretty low for good motion use I think. They are also geared down with the belt drive. That increases torque but reduces speed and response speed can be important in sims.

I’d bet those frames in stock configuration just do gentle sweeping motions to simulate flight but that the system lags and only approximates the movements. I’d rather be wrong but I checked those motor ratings and they are indeed 7A x 2.24V (about $150 on ebay too). Steppers aren’t generally meant to pull high currents and don’t have super high torque.

If what I am seeing is right, I’d gut the motors and drives and look at beefier motors and controllers and then you may need to reinforce the frames where the drives attach to take the higher stresses. But if that system is designed for only gentle movements, it will probably be disappointing for use with modern sims and SimTools. New controllers and motors would also let you skip reworking the Arduino programming to work with that system. You may need to also replace the power supply(ies?) depending on specs. Other approaches are largely canned and setup is much easier.

But I could be missing something. That’s just what I can see from the ratings on the parts. There may be more to the story. The cool part is you already have frames, shells, and a lot to work with!

is your arduino circuit driving the motors them selves or the motor drivers ?

Wow!! @Zed That is great help! I really appreciate the attention and detail of your response, it is super encouraging to get the help here and I am excited you see the potential here. This used to be a commercial dogfight simulator that people would pay for at a mall, and I have both of them. I really am lost when it comes to electrical ratings and clearly you have great knowledge, thank you again. I would love to upgrade this as you suggest I just don’t know where to start. I want good response but not jerky because neither slow response nor jerkiness is like a real aircraft and I want to simulate flight and actual flight physics coming from modern P3D Lockheed Martin software. With that in mind could you recommend a paired set of motors and drivers that can be easily adapted to the existing framework? I am excited to hear your response! Lee

Thanks, the Arduino is driving the motor controllers.

As you can see the motors do not actually take much of a load because the weight is on the frame and the motors turn the ball screw/lead screw to tilt it on the fulcrum taking the weight.

And after the reduction belt drive, they run screw drives so another reduction. I bet stock that runs slow. ;-)

Are the screw drives part of what holds the rig upright? I can’t see the full bearing setup but is it two sets of pillow blocks at right angles to each other to allow the seat to pivot front to back and side to side? That could still be ok, though.

It looks like basically a full-frame 2 DOF rig and should be easy to convert over. I can’t really tell you what you need to do but most 2 DOF rigs put a U-joint under the seat and the seat can flop around pivoting about the U-joint. Two servos of some kind then act on the upper frame and are generally mounted in the rear corners of the sim to push and pull on the chair to tilt it.

That framing doesn’t look too heavy but your total upper frame weight will be a factor in the size motors/electronics you need. So will the geometry/attachment points and the location of the pivot point. Basically, the pivot point is ideally located at the center of gravity. Any offset can impose a static load on the servos and increase the forces required to move the platform. It’s a compromise, though, and any compromises mean bigger motors and more powerful electronics.

If you read through the various sim builds people have posted here you will see lots of approaches and ideas that can help you on your journey. Just a guess but I bet you will be looking at something bigger and more powerful than the crab pot motors some of us have used and those bigger motors might need Sabertooth drivers. But basically you get to sort what forces and throws you need for your attachment points, then sort what that translates to for motor requirements, and then size power supplies and controllers for those motors.

But I do think you’re looking at a new drive and control system to get those frames airworthy. Cheers!

What inputs does your motor controller have ?

I was thinking about this and you could also just replace those stepper motors depending on how much slop there is in the mechanism. I tend to think in terms of worm or planetary drives and output arms but those long screws are very similar to how linear actuators work. You could use higher performing DC motors and rig feedback to turn them into servos. That’s the more common control method here - servos. You send a desired position and the servo/controller electronics look at the error and move the servo output to minimize that error. Steppers can do the same thing but are a lot more complicated in their internal logic and not as powerful.

If the slop is little to none with the stock setup, you might just swap motors and add feedback and controllers to turn them into servos. If it does have slop, you may want to look at worm gear output motors with lever arm outputs you can link to the top frame for motion. This would look a lot like other 2 DOF systems here but I do suspect you’ll want slightly bigger motors than many of the rigs here.

The pivot could be another source of slop. It looks to be made out of two sets of pillow block bearings so the pitch and roll axes aren’t the same. I don’t know if that matters but if there is slop there, you might want to swap for a single sturdy U-joint.

Also, there is a program available called SimCalc (IIRC) and that can take geometry and motor ratings, etc, to help you get answers to what’s what. (https://www.xsimulator.net/community/marketplace/simcalc.184/)

Last, keep the CG stuff in mind. If the CG of the top frame (with pilot) isn’t coincident with the pivot point, it can put static loads on your motors and kind of pervert the motions. With standard 2 DOF rigs, the pivot points are under the CG but out of necessity are below the seat. This could be an argument for replacing the pillow block bearings if their axes are lower than they could be. What happens with a low pivot is pitch, for example, becomes pitch plus a translation forward or back. Similarly for roll. It isn’t just a roll but is actually moving you side to side along the arc defined by the pivot point and the distance to the CG, your head, etc. And as your top frame moves away from center where the CG is over the pivot point to where the CG is offset horizontally from the pivot point, then your motors start having to support the mass which increases power demand, heat, etc, and slows moving the top frame back to center.

Depending on how your seat is mounted, some put their seats on adjustment rails instead of fixed mounting. I set up little ammeters on my motors so can adjust my seat to get that proper balance - minimum current. It also gives flexibility for smaller folks to get closer to pedals or whatever since their mass won’t cause too much trouble if they need to move away from the pivot point.

Cheers!

From your first post it sounds like your expectations are too high for a simple 2dof setup. Everybody has varying degrees of good enough, so I could be wrong. But to me it sounds like you're going to invest a lot of time and resources into a motion rig that will leave you wanting more in a very short time.

I think you should get what you have running before you invest in more powerful motors etc. If you've got the motors turning on the bench then what is stopping you? Connecting to simtools? If that's all then then that's easy enough.

First of all steppers are a pain in the but because they really need to be ramped up and down in speed. If you are using the regular stepper code in the arduino libraries then you can get things to move easy enough but it wont work for our purposes. The reason is because the arduino library will finish moving to the first commanded location before it stops and then heads to the new commanded location. The problem with that is we are issuing a new command every couple of milliseconds. Our motion data often changes quicker than the rig gets to its commanded position. So the rig will act very strange as it finishes moving to the old location before even listening to where to go next, maybe this wont matter so much with flight.

Someone here did make code to drive a stepper directly without the acceleration ramps and if you data is smooth enough it may not feel bad (driving has more high frequency data than flying).
https://www.xsimulator.net/community/threads/arduino-code-for-steppers.5434/

The other thing that sucks about stepper is they are noisy. You can microstep them with modern drives but you loose torque and speed. Overall it's usually not worth it.


As for Zed's comments about your motors being pathetically low wattage, that is a common misunderstanding of stepper motors. The voltage that should be used in your stepper motor power supply is different than the voltage listed on the motors. I read why somewhere on the net a long time ago but I can't remember the details. Something about coil saturation... google if interested.
Gecko has a good description of what power supply voltage to use with what motor.
https://www.geckodrive.com/support/step-motor-basics/power-supply-basics.html

But that shouldn't matter to you as you already have all the electronics. Get what you have working. Learn simtools or flypt mover, learn what your 2dof can offer in terms of data into the body. Read up on the great threads here for 6dof flight sims, and in my opinion most importantly, learn about gseats as they are the best way to get data into the body - especially for sustained g force like you see in flight.

@cfischer - That’s putting words in my mouth I never said. If you read what I said I was surprised the motor rating was that low and called it “oddball”. Quote: “But I could be missing something. That’s just what I can see from the ratings on the parts. There may be more to the story.” Not “pathetically low wattage”. Never said that. Please be accurate when you attribute anything to me.

But also, the reason I suggested gutting the existing motors was because OP repeatedly said he wanted a more realistic movement capability. The existing motor and drive setup is going to be slow with sweeping motions - hardly realistic. It’s using steppers and then has two stages of reduction with the belts and then the screws. Those steppers probably play back canned motions and it would be very difficult or impossible to modify that system for use with something like SimTools.

Why rebuild that back to stock? The stock drive system is unsuitable for a motion sim like this site is devoted to. Not that that is a reason rebuild it into a more traditional sim — again, I was just advising based on OP’s stated desire to rebuild it into a more realistic sim. Ultimately he will end up gutting that freshly rebuilt system and selling, recycling, or just throwing the old bits away, to then build it the way he says he wants it. And it will be difficult to “learn SimTools” without a motion sim that can act on commands from SimTools. It can be simulated but at least for how I learn, that’s not quite the same.

Nothing wrong with presenting options. Just please don’t build me as a strawman to attack something I never said.

Great input guys! I really appreciate all the information and I think you are both correct. My plan now is to simultaneously restore the existing system as a learning opportunity and really understand full motion sims and how to set up Arduino code for full motion simulation as a restoration project as I rebuild the A model and simultaneously learn about the upgrade options. In phase two I will redesign an X Model with new electronics and modern interfaces to new motors. With that said, I consider myself very familiar with actual defense department simulation systems and the F-16 fuselage that mounts on that frame was created with actual F-16 dimensions, so I will completely want to keep that as a big part of the project into phase 2. Inside the fuselage i will install a new interior for modern displays and operating weapons systems. I already have a built F-18 flight deck made from actual F-18 blueprints for correct dimensions that I hope to be able to mount on one of the platforms. I have two, the other will be the modern F-16. The point is that I know I will have to upgrade to get the authentic flight sensations. In the mean time, I will have a shorter path to some sort of operation and good hands on experience while I restore what I have to the original designer’s intent. At least I know at one point this same system operated commercially for years and it was formerly driven by software, so I have a great head start! I just need to figure out how to get the software driving the motors now that I have done my bench testing of the motors and power supply and Arduino control a single stepper motor. I have read about the Arduino code not processing fast enough and just ignoring things it can’t process. I now want to see how it operates as designed and then select the upgrade path. If I treat the lead/ball screw as an actuator with Arduino code for roll and pitch, am I thinking correctly for the first phase of my two phase plan? Again, all input is great learning and greatly appreciated, Thanks again!