Upcoming Events

Rookie friendly WTCR sereis Weekly BMW races on Simracing.GP Weekly BMW races on Simracing.GP Other regular AC events on Simracing.GP Weekly GT3 Sprint Races on Simracing.GP Weekly GT4 Sprint Races on Simracing.GP

DIY Pedal Build - Load Cell, Linear Pot & Clutch Cover

Messages
67
Points
21
This thread will cover the design and build of my sim racing pedals.

First a few photos of the work so far before I get into the detail.

Beardmore30.JPG

The Brake Pedal (work in progress)

Beardmore03.JPG

The clutch pedal (Work in progress)

Background:

I used to play Grand Prix Legends back in 2000 with a red Logitec wheel and pedals, dial up internet and a large speaker behind my seat (early tactile!) Then I stopped to concentrate on the build of my race car, which I developed and competed in. It was sold 8 years ago.

Fast forward to 2019 and I happen to see modern day sim racing on you Tube and was very impressed and decided I needed to get back into it.

After lots of research I decided I could build my own pedals quite cheaply. I had previously built a car from scratch, have a work shop with a lathe and TIG plus lots of material left over.

I intend to build the pedals using as much material from stock. So the designs may not be perfect but they make use of the material I have.

Electronics

Using Leo Bodnar components, it appears to be very easy to make transducer and load cells into game controllers. Time will tell if it is really that easy!

Brake

My race car experience taught me that it is easier to modulate pressure than movement, this then dictated that I need a load cell based system rather than a rotary or linear potentiometer. I looked at beam, button and S load cells and looked at prices.

My original plan was to use a beam vertically and mount the pedal directly to it, so in effect a solid pedal. Talking to the electronics guy at work I was advised that the cheap Chinese load cells do not tolerate any side loads, so this idea would not work. He also advised not to overload the load cell.

To get straight motion on the load cell I needed a linear bearing. My race car also had a small about of pedal travel then went solid. So I wanted to replicate this feel.

To understand brake loads so I could specify a load cell I put my bathroom scales in the passenger footwell of my car and pressed hard and noted the loads. I could press 50kg, which equates to around 500N. With leverage and availability I decided on a 100kg (1000N) beam load cell.

Most commercially available pedals feature large amounts of adjustability. I decided I didn’t need this as I was making them for me and not trying to replicate a particular car.

I also considered a hydraulic set up but considered it too complex.


Clutch

I decided I wanted a clutch. I looked at many designs to understand how to get the correct pedal feel of a clutch (increasing pressure then less pressure). All these designs needed springs. I quickly realised that springs are expensive. Plus, I probably would have to play with rates to get the correct feel. That would take time and cost money.

So, I decided to use a real clutch cover. A quick look on eBay indicated new ones can be had cheaply. That would definitely give the correct feel. Originally, I was going to use a master cylinder and concentric release bearing, but then I realised I was again making it over complicated. Why not have a mechanical direct acting set up? It is only a matter of getting leverages correct. Hydraulics give mechanical advantage just as levers do.

I kept an eye on eBay looking for a new small diameter cover, plate and release bearing. Eventually I bought a new small Renault kit for £14 delivered!

After discovering linear bearings I decided linear bearings would be ideal to actuate the clutch.

A rotary pot will be used.

Accelerator

Originally, I was going to use a rotary pot but was concerned about only being able to use a maximum of 120 degree of rotation. I have read articles stating it doesn’t matter that you are not using the whole rotation but I thought if I could it will not harm. I looked at many ways of converting linear motion to rotary such as rack and pinion, cable around a pulley and increasing rotary such as gearboxes but decided that a linear potentiometer would be best. The only downside of linear pots is cost, but I found a couple of options at reasonable cost.

I also looked at hall sensors and discussed them with the electronics guy at work but decided they were not the best solution.

Having discovered linear bearings I decided to use them for the accelerator too.


Brake Detail
Beardmore30.JPG


The brake is a linear bearing block mounted on a block of aluminium. A 16mm dia rod slides through the middle. A 100kg load cell is bolted vertically. The advantage of the block is it is easy to keep everything square and in line, plus is easy to drill and tap into.
There is a rubber buffer acting on the load cell. This screws into the 16mm dia rod. It is set up with a small gap to the load cell. This will give the initial slight pedal movement.
At the pedal end a Lancia Integrale inner valve spring (had it in stock!) ensures the pedal returns and there is some initial pedal pressure.
The pedal pivots on cartridge bearings which are old ones from my mountain bike pivots. To be honest bearings are not really required due to the very small amount of pedal pivot but as I had them it was cheaper than buying plastic bearings or bushes.

Beardmore32.JPG


Beardmore31.JPG

Detail of the pedal end of the linear bearing block. The aluminium disc between spring and linear bearing is held onto the linear bearing with Loctite 638 bearing fit.
The spherical rod ends are M6.

Beardmore18.JPG

Detail of the rubber buffer and aluminium spacer.




Beardmore27.JPG

This photo shows how I ensured the linear bearing block was mounted inline. Using the 16mm rod I was able to check everything was in line. The fact that the rod was long amplified any error and made it easier to ensure it was straight.


The 16mm rod is surface hardened. The only way to cut it to length was to grind it using a 1mm thick cutting disc. It was also hard to tap. I had to turn the tap a few degrees and then back it off. Time consuming but I didnt want a tap breaking off in the rod.

Work to do on the brake
Brackets to hold the block onto the box section
Brackets to hold the spherical bearing to the pedal
Spacers either side of the spherical bearing
Weld everything up - My gas has run out and now lock down is easing I should be able to get some more.
Mounts to my rig (which I will build)
Tactile

Tactile
I intend to add some tactile to the set up. First thoughts is an exciter to the pedal stem. I can easily isolate the above set up and isolate my heel pad. In fact I can easily isolate every pedal and every heel plate separately from each other and the rig. My plan is to build this in so I can add tactile as I learn about it / choose what to do.
I am currently researching isolators so I know the best to use.

Next time the clutch.
 

Attachments

  • Beardmore33.JPG
    Beardmore33.JPG
    102.9 KB · Views: 73
  • Beardmore29.JPG
    Beardmore29.JPG
    268.2 KB · Views: 63
Last edited:
Messages
207
Points
121
This provoked me to realize what is missing from throttle feel, namely inertia of linkage and throttle bodies. I added mass to free end of throttle pedal, for a subtle improvement.

Its actually NOT inertia, its friction. I started with a real MX-5 pedal assembly, and a standard linear spring. That didn't feel right, and the spring was too soft. But, with enough spring force, the pedal was too "snappy". I tried a bunch of different spring types and geometries, without much success. I then bought an old MX-5 throttle body, and connected it directly to the throttle with a short piece of cable. It still didn't feel right---not enough pressure, and too snappy on return. So, I bought a real MX-5 throttle cable, which just adds a bunch of drag. Voila. The friction requires more foot pressure to overcome, and adds resistance to slow the return spring.
 
Last edited:

blekenbleu

Premium
Messages
829
Points
1,027
Its actually NOT inertia, its friction.
OK, adding inertia was enough so that G29 throttle feel at least no longer distracts me,
just as adding degressive linkage to its clutch made that no longer distracting,
although inauthentic feeling at launches.
Still fiddling with load cell brake, though...
 
Messages
5
Points
0
Thanks for sharing your pedal build! I was just researching a clutch pedal build using a "real" clutch disk and found your post!

I'm going to go the hydraulic route and build my own clutch disk using a Harley diaphragm spring to keep the overall size down. But like the manual setup you built so I'm rolling that around in my head now instead :)

I'm wanting to replace the ramp/spring clutch I built and am currently using.

You might have already posted this but curious what POTs you are using? Currently, I'm using just cheap wiper pots and wanting to upgrade to hall effect.

Thanks again for sharing!


1230201042.png
 
Last edited:
Messages
67
Points
21
The accelerator is detailed on the first page.

Many of the cheaper ones only have a life of under 100,000 cycles which I calculated wouldn't last that long. There are plenty around with long life but many are over £100. However I found a good range from Variohm Here for reasonable money. I bought mine from RS Components and went for a 50mm long version.
1591901704461.png

I used linear bearings to make a sliding carriage that would operate the transducer. As the linear transducer is quite delicate I wanted to ensure it didn't get any side loads.


The clutch is a rotary unit from Bodnar.

The loadcell is a cheap Chinese unit.
 
Top