Anyone tried a force feedback shifter for non synchro boxes

[RiftFlyer]

I recently started driving ATS in the rift. Along with a physics mod I disabled synchronized gearbox - uset g_hshifter_synchronized "1" to "0" in config. This makes the Eaton Fuller boxes a LOT more realistic as you have to double clutch or float and match RPMs to find the gear. The issue is there is no feel for when the gears will mesh and slip in so finding the gear can be very difficult.

I was thinking of ways this could be improved and made more realistic. I was thinking of servos operating a gate at each opening on the H pattern. The servo arm could be closed when the engine rpm doesn't match the roadspeed and open for a given gear when the rpms match.

I have picked up a new shifter with range and splitter (not CSIO) and was already toying with extending my G27 gear stick. The mechanical arrangement for H box and servos won't be too difficult but before I go any further I need to know if the necessary data can be output from the game. Is this something anyone else has tried? I'd welcome feedback from those more experienced.

[RiftFlyer]

There doesn’t seem to have been much interest in this but I revisited it again lately now that clutch and gear behavior has been updated.

I’ve been doing some coding and have the electronics functioning as I need them. I’ve used two Arduinos. The first is a sparkfun pro micro which has the gear switches for the H pattern along with the splitter, range and engine brake.

The second board is an Arduino Uno which controls two solenoids (to restrict movement of the shifter) and a vibration motor (to simulate grinding).

I have the micro successfully recognized in windows as a game controller and it is also outputting when a gear button is pressed over i2c to the second Arduino.

The second Arduino is getting the gear data from ATS over a serial connection. This info along with the button info from the micro is all I need. I have a few functions written and tested them in the sim and it seems to be working as I intended. I now need to build the physical shifter. I have many of the parts needed already gathered up.

The gear buttons are to be mounted on spring arms so they can be pressed without the shifter actually slotting forward or back into position. If a button is pressed and the game hasn’t output a gear change then the Arduino fires the vibration motor on the shaft via a relay. This is to simulate grinding. If the button is pressed and the sim does recognize a gear change then the solenoids fire which allow the shifter to slot into position in the H pattern. I’m using a 12v powersupply to power the solenoids and motor.

The result should be that it won’t be physically possible to move the gear stick into position unless the box has been synchronized. The added vibration should provide some nice feedback for when the rpms don’t match.

I’m still waiting on some spring hinges to arrive from China (these are needed for the gear switch mechanism). Once they arrive I’ll start building and share my results.

[theclumsygeek]

Good luck man! We've been talking about this, too for quite some time. Really hope you succeed with building a real ffb shifter.

[BK Vissers]

This is interesting, I’ll have to experiment with non-synchro transmissions.

[supersobes]

This is an awesome concept! This would be the next step in making the gear shifting experience more realistic.

[RiftFlyer]

Thanks for the feedback guys. My big worry at this stage is timing. There are two possible issues:
If I simulate the larger throw of a truck shifter then I fear, for upshifting, the rpm in the game may drop too quickly to physically pull the stick into neutral, fire the relay then slot it into its new position. It may be possible to overcome this with a mod if necessary but I’m really hoping not to have to use any. Do those with real trucking experience think the current time for rpms to drop in game is realistic?

The second concern is again to do with timing. I’m hoping the length of time between the game recognizing a change from neutral to ‘in gear’ and the length of time it takes the solenoid to fire and me to physically move the stick isn’t too long as to cause issues. I won’t find that out until I’ve built the shifter!

This could be a deal breaker. During testing I’ve noticed that the dash shows a selected gear while the clutch is pressed but the actual telemetry output doesn’t show it until the accelerator is pressed. I’ve only been testing this with the sequential gear box and the keyboard on a laptop. I need to view the telemetry on my main pc where I have the g27 currently hooked up. I’m hoping it’s different when I have the full Eaton Fuller selected.

[RiftFlyer]

I just tested the telemetry output with my G27 with an Eaton Fuller 18spd in my truck. I am delighted to report that the last hurdle mentioned above regarding gear output from the sim won’t be an issue. The sim is outputting the gear when the clutch is pressed and even when the ignition is off.

Now I just need the final parts to arrive. I’m really hopeful that this is going to work!

[aujagor]

Saw this mentioned in the SKRS thread. I’ll be followinng this, and should you decide to build more than one I may have to pick one up. Been wanting a more realistic shifter myself for ATS. Keep us posted buddy!

[RiftFlyer]

Some progress!

I have the main mechanical components built and tested. It is all working as intended. I decided to build it as a floor standing unit. Having used all metal construction the result is a heavy lump which has a great feel when slotting into gear. I have a few bits left to do. I need to install the vibration motor, sort the wiring and add two additional limit switches for gear selection. I need to wait before mounting them due to issues with throw (see below). The two Arduino boards have their respective sketches installed and are working. The Uno is controlling the solenoids and vibration motor via some relay shields. The power is provided by a 12v mains powered supply.

[ external image ]

Here is the internal 'synchro' mechanism which prevents shifiting unless the rpms are correctly matched. The limit switches are mounted on spring loaded arms. The arms are locked by the barrel of the solenoid. When the solenoid is retracted (via arduino) the arms are free to move.

[ external image ]


This is the forward and aft mechanics. The bearing wheels are tensioned with springs. Once the stick is free to move (solenoids retracted) the wheels ride over the MDF track to slot the stick into position with the familiar 'clunk'. This mechanism also serves to hold the stick in place and stops it falling out of gear.

[ external image ]

The Road King has arrived from Tech Affliction.

[ external image ]


I have 'dry' fitted all components to take some measurements. I quickly discovered that with the shifter knob at a comfortable height the throw is way too much to be comfortable. I think I need to re-cut the H pattern in the top plate. As can be seen in the images below the throw is excessive. I'm going to recut it from a plastic chopping board and use narrower spacing between the gates. The current metal on metal arrangement is loud and will eventually show signs of wear. If anybody has access to a rig with an 18spd box could you share some measurements. If I could get the height of the grip from the floor and also the range of movement left and right, forward and back that would be a great help.

[ external image ]

Please give me some feedback and any ideas are welcome.

[RiftFlyer]

I have redone the top plate and H pattern to try and get more realistic throw. I purchased a cheep chopping board and cut the new pattern. Awful stuff to work with. I really struggled to cut it without melting the plastic. It came out reasonable ok and can be hidden with a shifter boot.

[ external image ]


You can see the throw has been greatly reduced at the shift knob and I think it will work out a lot better.

[ external image ]

[ external image ]


Now I need to fix the vibration motor in place and make two more switch arms for the gear output.