Load Cell Brake Pedal - Feel? Travel? Pre Load?

I'm in the middle of building a load cell brake pedal and the only thing I really have to go on is the feel of my real car (which is actually a pickup truck! so it's not much help) and my Logitech G29 which is only based on springs. Can anyone answer these questions?

I've broken my brake feeling down into 3 stages. preload, under load, full compression.

1) How much if any braking pressure should be applied under the pre-load? To me when I press the brake of my real car I don't think I feel any brake until I start to feel resistance, ie after the preload. Is this the correct setup? So the brake cut off should be set to the end (or near the end) of the preload travel?

2) Brake Travel - I've come up with a repeatable system to adjust the brake travel under load by compressing different pieces of rubber. On the hardest setting, there is almost no compression at all, so after the preload spring, the pedal might move 1/4 an inch maximum, basically pushing on the load cell. On the softest setting, the pedal will compress 2" or more after the preload. I know that a lot of this is personal preference but what would the "out of the box" setup feel like on a decent set of pedals?

3) Position of the load cell. I've looked at a lot of the higher-end sets and most of them have the load cell in a horizontal position with a force being applied to the end at a 90-degree angle. I've added a picture of a similar load cell and a drawing for reference.

I imagine this is a basic physics question.... when you push the pedal down in these setups the spring pushes some type of L bracket against the load cell. The bracket doesn't bend so the only "give" is downwards against the load cell. How much force is "lost" when compared to applying the force directly to the load cell?

This vaguely feels like high school physics and a lever issue. I imagine I'll be embarrassed by the answer.

you need to get rid of the movement in the system

get rid of the spring and fit the load cell

I haven't ever tried commercial load cell pedals but I have made my own load cell mod for my g27 pedals. I think you have the right idea with how it doesn't apply the brakes till after the pre load and then just progressively increases. In a real brake system, there's usually a bit of free play in the pedal before the brakes engage, then the travel after that would depend on how much air is in the system, how much the rubber brake hoses expand, etc.
From your drawing it is hard for me to say how the force from the spring would affect the load cell. It looks like it would most of the force would be stretching the load cell lengthways which doesn't seem optimal since they're designed to work as a cantilever beam and detect the difference in compression and extension between the bottom and top of the load cell.
Here's how I did it in case you're interested: https://grabcad.com/library/g27-load-cell-brake-modification-1

Hello Erik,

Thanks for the reply. Nice mod, I am using the same style load cell.

You are right in your assumption that the force is "stretching" the load cell in my picture. Here is another picture I found that shows a load cell in a similar arrangement (not exact but same orientation).

My original thought was to use the load cell exactly like you did. I'm just wondering if there is some type of advantage to this style setup as it looks to be used quite frequently.

Thanks for the answer on the other questions!

Here is a picture of how the load cell is used in another application.

That second pedal is a bit easier since there's a larger angle (about 45 degrees) between the push rod and load cell, and the end of the push rod is mounted nearer the load cell centerline. In that case the load cell would get about 0.707 times (from sin(45)) the force in the push rod spring, which I guess would let them use a lighter weight load cell.

I've tried going into a bit more detail drawing on your original sketch:


A = angle between load cell and push rod
Fs = force in spring/push rod
Fy = force perpendicular to load cell (bending it)
Fx = force parallel to load cell (stretching it)
c = load cell length
d = distance parallel to load cell from load cell mounting point to push rod mounting point

So the vertical force Fy on the load cell would be equal to sin(A) * Fs. The force Fx stretching the load cell would be cos(A)*Fs. I think Fx wouldn't change the output of the load cell as I think it would stretch the strain gauges on top and below the load cell equally, so can probably be ignored.
Because the mounting point is out past the end of the load cell, the vertical force on the cell would be increased and would be equal to Fy*d/c.

To figure out the force Fs in the spring/push rod:

F = force you push on pedal
a = distance from the pedal hinge that force F is applied
Fs = force in spring/pushrod
b = distance perpendicular to the pushrod from pedal hinge
c = distance of push rod mounting point above load cell (oops I used c again but am running out of time)

Fs=F*a/b.

Because in your drawing the push rod/spring is mounted significantly above the load cell, it would also produce a torque Fx*c on the load cell (Fx from previous drawing, c from this drawing) which would generate a bending force in the load cell which would affect it's output.
But in the photo of the other pedals, the push rod mounting point on the load cell isn't so high above the load cell, so could probably be ignored.

Sorry, I've kind of run out of time, have to go to work. I hope that was helpful.

Thank you Erik for the very detailed explanation. I very much appreciate it!

I am concerned about the torque (as you described ) causing the load cell output to be less linear. Clearly, my understanding of basic physics is lacking. Thanks again Erik for doing the math.

I'm just waiting for my load cell to arrive and I can post a picture of my test setup.