desensitize the electronic steering assist

Here is the original article:
Modifying Electric Power Steering
However, it is more complex than it needs to be.

The problem is two, variable pots are used. Pots are more expensive than fixed resistors and if one is turned a little too high, there would be a bias towards on or the other sides.

If I choose to do this, I will create a fixed, voltage divider network between +5V and ground. Then I would have the fixed inputs of a single pot tied to the middle of the resistor divider network and the sensor network. The pot sensor would sweep between the two. This allows a single control to adjust the gain. Two fixed resistors will be more stable than two pots and any variation in a single pot simply changes the gain without asymetrical signaling. I've not discussed the resistor network analysis that is needed.

With the car off, I would measure the resistance to ground of the sensors. I would then apply a left and right torque on the steering wheel to measure the change. This information is used to model the circuit and choose values for the fixed resistors and pot. The risk is making sure there is no increase in current flow through the torque sensor wiper.

There is an even better circuit that uses a linear amplifier with the pot tied to gain control. As a complete buffer, there is no possibility of increased current to the torque sensors. But this is something I would pass on to hobbit, a world class linear amplifier circuit designer.

Bob Wilson