Showroom (In Progress) DIY FFB Wheel and pedal set

Some work in progress for full size, gross, sim racing cockpit... clutch and brake pedals from BMW E39, throttle pedal from BMW E46. Quick made board with 3 arduinos converting sensor signals to 0 - 5V analog for feeding to Arduino Leonardo FFB wheel. Hall angle sensor module AS5600 on the clutch, cheap load cell on the brake, gas pedal uses original hall sensor

Really admire you these will be the real car parts of the craftsmen, for me this novice only worship the share.

The steel frame is under construction...

Can you use one arduino for the wheel and the pedals ?

Unfortunately... i don't have that much knowledge in software programming. I use AI-Wave firmware for the leonardo which has only three 10-bit analog inputs.

BMW gas pedal angle sensor outputs between 0.4 - 4.3V and it's better to use external arduino in combination with 12-bit ADC (ADS1015) and 12-bit DAC (MCP4725) with some mapping to achieve full swing 0 - 5V. That way the loss of resolution is smaller...

AS5600 hall angle sensor module on the clutch has I2C interface... it cannot be connected directly to Leonardo.
Again some code and MCP4725 for analog output 0 - 5v.

For the brake i use HX711 and load cell. I tried the option for direct connection (pins 5 - 7) but i don't like how it behaves. I wrote some messy code with automatic zeroing every time brake pedal returns (ISR triggered by stoplight sensor).

Bottom line is... i just use what i have and what i can . Maybe is possible to combine these three codes into one arduino but i don't know how to do it without increasing overall latency of the pedals...

Now added latency is 2.2ms for the Gas, 1ms for the Clutch, 12ms fog the Brake.

Throttle pedal mounted in place and some very crude spray painting....

I admire the work but I think you're over complicating things there that are just going to add to the delay in the signal and affect reliability.

it's great to try to get 'full swing' but in reality you are compromising more important factors (speed) to achieve precision.

All the A-D/D-A conversions take time and for no realistic benefit. You're ensuring a 12 bit signal but then sampling it at 10bit anyway. You'd be better off with OSH which samples at 12bit to start with.

That load cell interface is slow - I think it can only do 100hz from memory. Stick with a normal Load cell amp (i.e INA122PA) and stay fully analog and feed the signal straight into your controller.

Not hating on your work at all - If you're happy with it then all the best to you but coming from a background in electronics I can speak from experience that the KISS strategy is often the best.

Hey, i know it's not the best way to do it but for amateur gamer like me... it's fine

I'm getting kinda old and my reaction times are probably far worse that those 12ms brakes latency...

Initial tests results are not bad at all... the worst problem so far are my driving "skills"



The steering wheel produce 6Nm holding torque at 6.2v. Linearity is good above 10% minimal force which i thing is normal for such heavy wheel with timing belt drive.