Monday, September 22, 2008

Three Fine Assemblies

With all three assemblies completed, we of course had to line them up on the mockup stand to see how they looked all in a row like soldiers.

Be sure to click on the images to see the larger 800x600 images. These show a lot more detail.

Don't these things look terrific?

The next steps will involve dismantling the pedal assemblies so we can drill half inch holes in the aluminum channel sections for the 1/2" galvanized steel rods that will support the pedal platforms.

We waited to drill these holes until we had completed the assemblies because we wanted to make sure that wherever we ran the steel rods through, they wouldn't interfere with bolt heads or rivets or anything else attached to the sides or undersides of the channel section pedal platforms.

Now that we know where everything has ended up, we can position the holes for the rods with confidence.

After that, all that will remain will be construction of the actual pedal base and installation of the load cell controller and wiring.

By the way, the day after we did this work on the pedal project, Amos totally destroyed the absolute lap record at Rocky Ridge in his stock-engined shifter kart. He won the heat race going away and was charging in the feature when a hub broke and put him out, leaving the way clear for his dad to take his own first shifter kart feature win. With one race day to go, Amos has a healthy lead in the Vermont Shifter Kart Stock 80 championship.

Way to go, Amos!

Pot Mounts and Actuators

No, this has nothing to do with either cooking or weed.

The throttle and clutch pedals work by operating a potentiometer, which is a variable resistor. Each of these will be connected by a couple of wires to the Leo Bodnar load cell controller, which has circuitry to deal with switches and pots as well as the load cell.

The little bracket that mounts the pot to the top of the pedal platform is made from a 3/4" long piece of 1" angle aluminum. The pot is the black plastic box with the little solder tabs sticking out to the left.

The pot actuator arm is made from a piece of gray square plastic rod we ordered from McMaster. The minimum was three feet and we only need about six inches to do both pedals on both my and Amos' (yet to be built) pedal sets, so we have a lot left over! Fortunately this stuff cost about a dollar.

It's also very easy to work. Which is lucky because I drilled the hole for the pot shaft a little bit oversize in the first one I made, so I had to make two more. This took about ten minutes each.

Once we'd cut some 6-32 threaded rod for the pot pushrods we were able to assemble them together with the little R/C aircraft rod ends and 4-40 screws, nuts, and washers, and figure out exactly where we wanted to mount the pot mounting brackets to make sure we got free motion throughout the range of travel of the pedals.

We were also hoping for linearity of operation of the pot, but given the geometry that's pretty much impossible. However, we think that the nonlinearity will be advantageous, with slower ratio at the beginning of the pedal travel and higher ratio at the end of travel.

I think this is wicked cool!

Throttle and Clutch Final Assemblies

While Amos was doing all his work on the brake pedal and actuator assembly, I fabricated the bracket that the front end of the throttle spring would be pressing against.

I also drilled out the rivets in the corresponding bracket for the clutch spring and cut a quarter of an inch off the bottom of the bracket to make it match the length of the new throttle spring bracket, and then riveted the little angle bracket back on.

I did this to lower the height of the front end of the pedal pushrods so the pushrods would be more nearly level. I thought this would look better, and also it fixed it so the up-stop (the large steel and rubber washers at the front of the assembly) were resting nearly flat against the front side of the spring end bracket.

At this point I was ready to assemble the spring end brackets and their diagonal braces to the clutch and throttle platforms. After a bunch of drilling and riveting, the job was done and the throttle and clutch pedal assemblies were complete.

Brake Pedal Final Assembly

More progress! In fact, at this point it feels like we're nearly done.

On Saturday we spent a long day completing and assembling each of the three pedal assemblies.

This day felt different from the previous days, however. At this point we are so far along that there was very little time spent dithering and making decisions, then reconsidering and deciding agaian, as we'd been doing most of the previous days of this project.

Instead, we found that things were falling into place. Both of us simply kept working: measuring, drilling, cutting, filing, riveting, and bolting together. It was a rewarding day!

Amos spent most of the day working on the most complex assembly, the brake. He drilled the pivot holes in the u-bracket that mounts the load cell actuator and then spent some time tinkering with the alignment to try to get the actuator perfectly square with the tongue he'd fabricated earlier and mounted on the load cell.

He also determined where he wanted the u-bracket to sit on the pedal platform, and drilled the holes for that.

Along the way he decided to put a shim under the rear end of the load cell, and then during the alignment process he also made a shim to go between the u-bracket and the pedal platform.

It also took some tweaking to get the spring situated so it didn't flex sideways and hang up on the nut on the threaded pushrod inside it as the pedal is actuated.

By late afternoon the complicated brake pedal assembly - the part of the project that had required the most design work and the largest amount of precision fabrication - was complete.

Didn't Amos do a fantastic job on this?

Wednesday, September 10, 2008

Hot Shoes

And now, at last, we get to see the pedals in action (sort of).

Here are some shots of the clutch and brake pedal assemblies sitting on top of the mockup floor stand. Don't they look cool?

We haven't got a throttle assembly yet. I know I said we did in an earlier post, but actually that was the clutch assembly masquerading as the throttle assembly so we could gauge the relationship of the throttle pedal to the floor.

In reality, the throttle and clutch assemblies will be identical except that the throttle will have a longer pedal pad to facilitate heel and toeing.

But in our case, the first pedal shaft and platform assembly Amos made has a little bit of play in the pivot between the two, because Amos was learning how to deal with a slightly bent chuck in his dad's drill press. This makes the drill bit wobble and makes it difficult to produce holes that are precisely the right size and in precisely the right place.

By the time he made the second and third pedal shafts, Amos had mastered a technique that allowed him to drill extremely precise holes despite the defective chuck. Since a small amount of side to side wobble is less critical in the clutch than it will be in the throttle and brake (again, for heeling and toeing we want them to be stable laterally) we've decided to use the first assembly for the clutch and use the more precise later assemblies - which have no play at all - for throttle and brake.

Anyway, in these shots you can see how the assemblies will look when finally mounted on the finished pedal stand (which we haven't started building yet).

Amos kindly contributed his shoes so that we could demonstrate the operation of the pedals. The shoes are having a rather hard time heeling and toeing, though, because of the absence of the missing throttle pedal.

The Actual Load Cell

Finally we were able to assemble the actual load cell to the aluminum channel section part that I'm calling the pedal platform, and set all the load cell actuation pieces in place.

The right-hand end of the load cell in these photos is bolted to the pedal platform by two metric bolts that go through holes from the bottom side and into the load cell, which is tapped with metric threads from the factory.

To the other end of the load cell we've bolted a part we made. This is just a flat piece of steel bar stock that is slotted to accept the load cell actuator fulcrum.

The fulcrum projects upwards and the plastic knob, whose stud screws into the fulcrum, clamps the fulcrum to the horizontal section of the load cell actuator. So when you push on the pedal, the pushrod will move to the left (in these photos), rotating the actuator counter-clockwise, and putting upward pressure on the fulcrum.

This will lift the left end of the load cell, which will cry uncle and send a resistence change through its wires to the load cell controller (which has yet to make an appearance in this blog). The controller will notice this change and produce numbers which will, in turn, be sent via USB to the racing sim to let it know you are pressing on the brake - and precisely how hard.

We still need to drill the holes in the steel U-bracket that will accept the actuator's pivot bolt, and also we need to drill holes in the U-bracket and the aluminum channel that will allow us to bolt the two together.

We put these off till last because the size of the fulcrum and load cell and the actuator all dictate where the pivot holes have to be.

At the moment we're debating whether to cut down the top of the fulcrum a little, which will allow us to lower the actuator so the pushrod will end up being more nearly parallel to the top surface of the pedal platform. (See the top photo in this post for the actual position of the actuator and fulcrum once everything's assembled.)

This is just a question of aesthetics, but, hey, aesthetics are important, right?

Actuator Fulcrum: the Bogey Part

Throughout the entire project up to this point, there has been one component that was our "bogey part". This is the assembly that we are calling the load cell actuator fulcrum.

Part of the reason for this being such a head-scratcher is that there are no photos of Todd's production pedals on his web site that show this part. And it doesn't exist in his DIY design.

So we had to infer not only the function of the part, but its configuration.

We spent a considerable part of the first day of the project (the design day) brainstorming different ideas for what this part should look like and how to make it. We made several sketches, and finally settled on a design and ordered the materials.

But as the project went on, each time we looked at this part, we found ourselves reconsidering its design, trying to come up with something that was going to be easier to build and yet properly effective.

In the end, this past weekend, after yet another round of reconsiderations - which included perusing a variety of specialty bolts on - we ended up settling on a design which was essentially the same as one of our early concepts.

Amos started cutting metal. He said it was annoying to do all the hacksawing this part required (the steel bar stock we had ordered was bigger than necessary) and he seemed to dread drilling and tapping the necessary holes.

He even fired up his laptop and looked up on the Internet to find the drill sizes for the 10-32 and 8-32 threads he had to tap. Of course, these are number drills, and naturally we found that his dad didn't have any number bits. So he had to make do with inch sizes, which was even more annoying.

But by the end of the day Amos said that the process had been incredibly fun and satisfying. And he'd produced a part that was simple, elegant, and (we hope) effective.

Isn't it pretty?

Tuesday, September 9, 2008

Load Cell Actuator, Part 2

The load cell finally arrived from China, and that allowed us to confirm the dimensions of the actuator and the other parts related to it.

As it turned out, the dimensions of Amos' mockup load cell were perfect, so we needn't have waited.

Anyway, once we knew that the dimensions were correct, we were able to construct the load cell actuator fulcrum (see next post) and once that was made, we knew the dimensions of the slot that had to be cut in the load cell actuator.

Here are some photos of the finished actuator. Didn't Amos do a brilliant job?

Mockup Pedal Stand

Last Friday while Amos was at school or running cross country practice or something frivolous like that, I knocked together a mockup floor stand for the pedals out of scrap wood. This is to help us determine the height and angle of the pedal platforms from the floor before we build the actual stand.

The height of the pedals from the floor is critical because I don't want to have to keep lifting my foot up off the floor to actuate the pedals. I am hoping I'll be able to operate them with the ball of my foot, except that I'll need to lift my heel and rotate it to the right to blip the throttle when heel and toeing.

Here are a few photos of the throttle pedal assembly sitting on the stand. Note the pieces temporarily clamped in place on the top of the pedal platform; once everything is finalized, we'll rivet them in place.

Yes, that handsome fellow checking out the pedal operation is Amos!

Load Cell Mockup

The load cell came from China via registered mail, and it took three weeks or so to get here.

This load cell and its associated parts dictate the length of the aluminum channel part to which all the other pieces of the pedal assembly will be attached. So it's a critical part.

While we were waiting for it to arrive from China, Amos made a wooden mockup of the load cell using dimensions from its web page so we could verify the dimensions of the load cell actuator and associated pieces.

Here are some photos of the mockup load cell, actuator, and actuator pivot bracket, all pieced together with the pedal shaft and platform. This made us confident that the dimension we'd chosen for the length of the platform (9") would be long enough to accommodate all the parts.

Pedal pad

Here's a shot of the pedal platform and pedal shaft with a pedal pad attached.

Todd's DIY plans specify pedal pads like this, made out of 2" wide aluminum strap, but with an option of using pedal pads made by aftermarket suppliers for real cars.

Since Amos and I both drive with stocking feet, we're going to try this design first. We think the smooth surface will allow our socks to slide easily up and down on the surface as the pedals move.

If we don't like this, we can always substitute real-car pedal pads later.

Load Cell Actuator, Part 1

The load cell which produces the numbers for braking force is the heart of this design, and is more than anything else its reason for being.

The load cell is fairly large, and its dimensions to some degree determine the length of the platform on which it sits, and to which the pedal shaft is attached. This in turn determines the length of the platforms holding the other pedals, and the spacing between the rods which will support all three pedal platforms.

Since Todd's production design differs in the way it mounts the load cell from his DIY design, and we had decided to copy the production design, we had to work out the layout and dimensions of all the components on the brake pedal platform before we could finalize the dimensions of the other pedal platforms and the floor mount.

This meant that designing and building the brake pedal assembly, with load cell and actuator assembly, came next in the process.

A key component of the load cell actuation assembly is a right angled part which will pivot around a bolt mounted to a bracket which is in turn bolted to the pedal platform. The brake spring pushrod will project through the vertical section of this part (we're calling it the load cell actuator) and one end of the spring will rest on the rear face of this section.

When the brake pedal is pressed, pressure on the spring will tend to push the vertical section of the load cell actuator toward the front of the assembly, which will cause the actuator to want to rotate around its pivot bolt.

This in turn will produce a lifting force on the horizontal section of the load cell actuator, and this lifting force will be transferred via a couple of other parts to the end of the load cell, subjecting the load cell to a bending force.

In Todd's production design, the load cell actuator is a weldment, but although we had access to both a MIG welder and a gas torch, I didn't want to weld anything on this project because that would have meant painting it, and because of my chemical injury, I don't want to have freshly painted things in my apartment.

So after some thought and debate Amos and I came up with a design for this part which is made from a piece of galvanized steel angle and a small section of the rectangular aluminum tubing that we were planning to use for part of the floor stand.

Making this piece required some serious precision work with the drill press in order to put small rivets into places in the two parts where they wouldn't interfere with either the pivot bolt or the spring, pushrod, and stop washers.

Amos did a brilliant job with this. Note the closely spaced rivets, in just the right place to attach the two pieces together without interfering with the moving parts.

The actuator isn't quite finished; we will need to cut a slot in its horizontal surface for the attachment of another part, which we're calling the load cell actuator fulcrum. We can't cut the slot until we've determined the dimensions of the fulcrum, so this is it for now.

Friday, September 5, 2008

Building a Set of Load Cell Pedals

I've always wanted a better sim racing brake pedal setup than anything I've ever tried on standard and even high end pedals. Using a spring and a potentiometer just doesn't cut it, expecially if you want to heel and toe. When resisted by a spring, but with enough range to operate a pot, a sim racing brake pedal simply moves too far. (Unless it's got a really stiff spring and some fiendishly complex linkage that translates tiny pedal movements into major pot traversal. Translation: big bucks.)

With conventional sim racing brake pedals, by the time you've got some braking, the pedal is so far down compared to the throttle that you haven't a hope of getting your heel on the throttle for that nice blip that makes a smooth downshift.

Enter my nephew Amos (above, hard at work in his dad's garage shop). Amos is a champion shifter kart racer who is hoping to move up to racing full size cars soon. He wants to use iRacing to practice his heeling and toeing. With the standard Logitech G25 pedals we are both using at the moment, this is just about impossible due to its long-travel, squishy potentiometer-driven brake pedal.

I'd read about Todd Cannon's CST pedals in the RSC forums. It seemed that a lot of people were extremely happy with them, and I learned that some real-world hotshoes like Dale Jr. were using them with iRacing.

These pedals are not only gorgeous, with a very elegant design, but they incorporate a load cell for the brake instead of a pot. And Todd sells a nice Do-It-Yourself manual so you can build a set of them yourself. Sounded like the way to go!

In early July of this year I talked this over with Amos and we decided to build two sets of Todd's pedals, one for each of us. Mine would be the prototype.

I ordered Todd's DIY manual and we spent quite a while poring over it, as well as the photos of pedals on Todd's web site. There were beautiful photos of his production pedals, and also a number of shots by people who built the DIY version.

We decided to build a design that was a hybrid of Todd's original DIY version and his new, more sophisticated (and more difficult to build) production version. This meant doing a lot more design work than if we'd just built the DIY version straight from the manual. But it would give us the advantage of more adjustability, particularly in terms of lateral pedal spacing and also in the area of brake pedal effort.

The first day we got together, Amos and I spent almost all day just going over the manual and the photos, making sketches, and creating parts lists in a Google spreadsheet.

Some of the design changes we were making were fairly straightforward, but others required a significant amount of thought, particularly because we were constrained by my chemical sensitivities; I didn't want anything with fresh paint on it in my apartment when we were done, so that ruled out welding.

Todd's production pedals have a number of weldments, so we had to come up with ways to make parts with the same functions but without any welds. Fascinating. And brain-teasing.

Finally we reached a point where I felt ready to start ordering materials and parts. A big order went off to McMaster-Carr for aluminum channel, angle, and bar stock, some steel bar, and assorted hardware. Another order went to China for a load cell, another to California for tiny R/C rod ends, and others to spring and electronics companies. And I spent a good part of an afternoon haunting the nuts and bolts section of a couple of local hardware stores.

In late July we met again at Amos' shop, which is actually the home garage/shop of my brother Nate. This is the same garage where Nate and I built my Cobra in back in the summer of 2000. Nate and his friend John Spain operated a Spec Miata out of this shop for several years. For the last year this is where Nate has been building and developing his hairy TDI Special, a former SCCA Spec Renault that now has a highly over-boosted diesel engine from a VW New Beetle and goes like stink. So to speak.

Anyway, Amos and I got together and finally started cutting metal. Or rather, Amos started cutting metal while I crashed on an air mattress he set up for me because I have this damnable illness that limits my capacity for normal activity to about five minutes an hour.

By the end of the first day, Amos had fabricated and assembled the first pedal assembly, which will be the clutch.

While Amos made the drill press sing, I'd happily assembled the springs, rod ends, nuts, washers, and other bits onto the threaded rods that will be used for all three pedals. The products of our combined efforts are evident in this photo.