(NOTE: this is a copy past of a post made by me, in the iRacing forum. Because some of you don't have access to iRacing forum, i will duplicate it here
Hi,
My intention with this project is to make a DIY H-Pattern Shifter. But, one of the main objectives, is to make it low cost. But not too crappy!!!
For reference, probably the best H Shifter in the market is the Fanatec CS Shifter with the tag price of 260€/200$
So, the objective is to stay under this budget. How much? It depends of the components I need to mill... and which material I will use.
1.
Since you can find in eBay, some used parts from real cars, for very low price, I decided to use the most important component that gives the input for the transmission.
You can find this components quite cheap. I bought mine for 20€! And I choose to use a VW shifter, since we can find allot of used parts from this brand.
Other good alternative is the Porsche 996 shifter (normally, expensive) or Honda Civic.
Note: for this concept, I'm using a shifter that has 2 Cables to transfer the movement of the Gear Knob to the transmission. (Normally FF MR and RR cars use this, FR aren't good...) And an old Shifter form a Porsche isn't good as well...
2.
The challenge is to make a mechanism that replicates the operation of the gearbox.
My first solution is this:
On the Left Plate (grey) is connected the cable for the Lateral movement of the lever
On the Right Plate (purple) is connected the cable for the Longitudinal movement of the lever
The Shifter pattern will be made with the combination of the Lateral and Longitudinal movement of the lever.
The Shifter support, normally don't have spring, included, so it is required to add the spring and step effect, that we are use to it IRL. For that, I have in each Plate a path, that forces a Ball Bearing connected to a spring/rubber bands, to generate tension on the plates, to return to the Neutral position.
Here, I have made a basic GIF showing the shift pattern from 1 to 6 (reverse no included)
3. Software
The connection to the computer, I'm thinking (already have it) to use a Leo Bodnar board, and 2 Hall Sensors, to read the position of both Cable Endings. (I believe it is possible)
I don't have a program, at the moment, to interpret this information. And my programming skills are really limited... But, it isn't the biggest priority.
Update:
I give up on the Hall Sensors... It required a very strong magnet, because the travel of the cables is too big. So, will use Spur Gears with Potentiometers.
4. Material
I'm in doubt what the next step I should do. I have a close friend with a 3D printer, But I believe if I print it, it will break really easily.
The best option at the moment is to mill from a sheet of Plastic, like PC or POM (probably 10mm thickness).
The other option, and probably I will do this for the final version, in Aluminium.
If anybody have a Desk CNC machine, feel free to offer your service!!! Send me a PM.
Final thoughts:
Unfortunately the final package is really big, and not pretty. But removing the cables, and try to connect directly to the Shifter, in my opinion, would make this more complex and expensive...
I have other solution, with similar concept, that uses less space. But, since this is my first concept, I will not care much about the package. Only after, I will share my conclusions and thoughts, and share a 3D model.
I will soon share the 3D models. But, I need to improve some details. I'm using CATIA V5, if you have it, I will share the parametric version. For the rest... the STEP file.
UPDATE:
Today I Bring more pictures.
Overal View:
Section View
Lateral Plate
Longituninal Plate
Cost (so far):
The Shifter (VW model) : 20€
10 ball bearings (12x6x4 with flange): 13€ (will use 8 )
Leo Bodner USB Board: 37€
3D Printing: probably for free
Spur Gears set: 2x 6€
Potentiometer: 2,40€
Other stuff to help the bluild (nuts, screws, etc): probalby around 25€
Update from 28/01/2016
So, Finally i have a working prototype. It took more time then expected... But i have a good starting point for improvements.
The printed parts are printed by a friend Henrik Samulesson (if you are interested in printed parts, fell free to contact him at http://www.henriksamuelsson.se/)
The quality is much better than i had expected. And i believe the printed parts can survive a daily use. Making in a stronger material, will imply CNC Router machine, and the final cost will probably not justify. Print spare parts can be mor cost efficient.
The final assembly (without electric parts. Still missing)
Neutral
1st Gear
Next step:
my next step is to attach the spur gear that i bought in a RC shop. I will use in the plates 40 teeth, and in the Potentiometer 15 teeth. This will be ok, but not the ideal situation. I will not archive a full rotation of the potentiometer, but a shifter don't need a big precision, like a steering wheel.
The working product:
[youtube]
(PS: sorry for the vertical video, my wife suffers form this syndrom... )
Final thoughts:
As you can ear, the sifter makes a bit of noise. The main reason, is the metal hitting some aluminium plates that i was forced to use, because when i was trying to make the oblong holes in the acrylic, i completely destroyed the holes... so, i had to find an alternative.
the other problem is the feeling. I was expecting a strong effect when changing the gear, but i don't have it. In the video, there is the Illusion that is a strong feeling... But it is not... The reason is the springs, they aren't strong enough. And were i found this ones, i cannot find stronger. The next concept I will take this in consideration!
The support i used is in acrylic, and used regular nuts, but this is a bad solution for the future, the nuts are getting loose quite easily. So, a locking nut or something to lock is required.
Because of that, the shifter plates jams and makes the shifting difficult.
So, I'm at the moment making a better version of this concept, with a smaller package, and better spring system.
Soon i will post a 3d model.
See you
Tiago Viana
Hi,
My intention with this project is to make a DIY H-Pattern Shifter. But, one of the main objectives, is to make it low cost. But not too crappy!!!
For reference, probably the best H Shifter in the market is the Fanatec CS Shifter with the tag price of 260€/200$
So, the objective is to stay under this budget. How much? It depends of the components I need to mill... and which material I will use.
1.
Since you can find in eBay, some used parts from real cars, for very low price, I decided to use the most important component that gives the input for the transmission.
You can find this components quite cheap. I bought mine for 20€! And I choose to use a VW shifter, since we can find allot of used parts from this brand.
Other good alternative is the Porsche 996 shifter (normally, expensive) or Honda Civic.
Note: for this concept, I'm using a shifter that has 2 Cables to transfer the movement of the Gear Knob to the transmission. (Normally FF MR and RR cars use this, FR aren't good...) And an old Shifter form a Porsche isn't good as well...
2.
The challenge is to make a mechanism that replicates the operation of the gearbox.
My first solution is this:
On the Left Plate (grey) is connected the cable for the Lateral movement of the lever
On the Right Plate (purple) is connected the cable for the Longitudinal movement of the lever
The Shifter pattern will be made with the combination of the Lateral and Longitudinal movement of the lever.
The Shifter support, normally don't have spring, included, so it is required to add the spring and step effect, that we are use to it IRL. For that, I have in each Plate a path, that forces a Ball Bearing connected to a spring/rubber bands, to generate tension on the plates, to return to the Neutral position.
Here, I have made a basic GIF showing the shift pattern from 1 to 6 (reverse no included)
3. Software
The connection to the computer, I'm thinking (already have it) to use a Leo Bodnar board, and 2 Hall Sensors, to read the position of both Cable Endings. (I believe it is possible)
I don't have a program, at the moment, to interpret this information. And my programming skills are really limited... But, it isn't the biggest priority.
Update:
I give up on the Hall Sensors... It required a very strong magnet, because the travel of the cables is too big. So, will use Spur Gears with Potentiometers.
4. Material
I'm in doubt what the next step I should do. I have a close friend with a 3D printer, But I believe if I print it, it will break really easily.
The best option at the moment is to mill from a sheet of Plastic, like PC or POM (probably 10mm thickness).
The other option, and probably I will do this for the final version, in Aluminium.
If anybody have a Desk CNC machine, feel free to offer your service!!! Send me a PM.
Final thoughts:
Unfortunately the final package is really big, and not pretty. But removing the cables, and try to connect directly to the Shifter, in my opinion, would make this more complex and expensive...
I have other solution, with similar concept, that uses less space. But, since this is my first concept, I will not care much about the package. Only after, I will share my conclusions and thoughts, and share a 3D model.
I will soon share the 3D models. But, I need to improve some details. I'm using CATIA V5, if you have it, I will share the parametric version. For the rest... the STEP file.
UPDATE:
Today I Bring more pictures.
Overal View:
Section View
Lateral Plate
Longituninal Plate
Cost (so far):
The Shifter (VW model) : 20€
10 ball bearings (12x6x4 with flange): 13€ (will use 8 )
Leo Bodner USB Board: 37€
3D Printing: probably for free
Spur Gears set: 2x 6€
Potentiometer: 2,40€
Other stuff to help the bluild (nuts, screws, etc): probalby around 25€
Update from 28/01/2016
So, Finally i have a working prototype. It took more time then expected... But i have a good starting point for improvements.
The printed parts are printed by a friend Henrik Samulesson (if you are interested in printed parts, fell free to contact him at http://www.henriksamuelsson.se/)
The quality is much better than i had expected. And i believe the printed parts can survive a daily use. Making in a stronger material, will imply CNC Router machine, and the final cost will probably not justify. Print spare parts can be mor cost efficient.
The final assembly (without electric parts. Still missing)
Neutral
1st Gear
Next step:
my next step is to attach the spur gear that i bought in a RC shop. I will use in the plates 40 teeth, and in the Potentiometer 15 teeth. This will be ok, but not the ideal situation. I will not archive a full rotation of the potentiometer, but a shifter don't need a big precision, like a steering wheel.
The working product:
[youtube]
(PS: sorry for the vertical video, my wife suffers form this syndrom... )
Final thoughts:
As you can ear, the sifter makes a bit of noise. The main reason, is the metal hitting some aluminium plates that i was forced to use, because when i was trying to make the oblong holes in the acrylic, i completely destroyed the holes... so, i had to find an alternative.
the other problem is the feeling. I was expecting a strong effect when changing the gear, but i don't have it. In the video, there is the Illusion that is a strong feeling... But it is not... The reason is the springs, they aren't strong enough. And were i found this ones, i cannot find stronger. The next concept I will take this in consideration!
The support i used is in acrylic, and used regular nuts, but this is a bad solution for the future, the nuts are getting loose quite easily. So, a locking nut or something to lock is required.
Because of that, the shifter plates jams and makes the shifting difficult.
So, I'm at the moment making a better version of this concept, with a smaller package, and better spring system.
Soon i will post a 3d model.
See you
Tiago Viana