[Gouky]

So It's a bit early in this design process to really bring this up, i don't usually talk about parts until i'm much closer to a sellable product but i need some feedback from you guys to make the right product.

I'm starting from a blank slate and I want to make a better accelerator pedal design that gives everyone a consistent pedal feel.

So i started by ordering a whole pile of Toyota's pedals

And first things first, i did confirm that there are different curves in the different pedals, there seems to be two categories of pedals, 13 degree pedals and 16 degree pedals but even in each sub category there seems to be minor differences. I also haven't done an exact measurement but they also seem to have different slopes within that 13 or 16 degree window.

So that gives me a good idea of what pedal to start with but there's a few preferences that may affect that.

First i wanted to match the proper pedal travel so i measured the stock MR2 and then everything else in my driveway and that's where things got weird. Stock accelerator pedal travel varies WILDLY from car to car and until i measured it i can't say i ever really noticed when driving before. The shortest travel was in the Cayenne at a little under 2" and the longest was in the WRX at just a bit over 4". No other car went over 3" though. The stock MR2 is about 3.2" but i don't have a motor in place so i don't know if the actual travel is shorter than that.

So my intention was to aim for 2.75 - 3" of functional travel but at least an extra 3/8" of unused travel at the end to account for carpet thickness differences (and potential added sound deadener material. But that full 3/8" of dead travel will only be there for people that have a stripped car without carpet or sound deadening.

Also, based on the complaints over the years i was going to make a metal stemmed pedal. I've had lots of complaints of people not wanting to see a plastic pedal in the footwell. This is compounded by the unintended accel service bulletin that resulted in most of the pedals being ground down to be shorter.

I also want to make it possible to use the "sport" face that was available in some models.

I'm also debating between the stock pedal space or moving it closer to level with the brake pedal. The stock MR2's accelerator pedal is almost 3"closer to the firewall than the brake pedal. Most cars in my driveway seem to be much closer to 2" Some feedback here would be appreciated. In theory it would make heel-toe work a bit easier.

and finally, i'm going to move the pedal over to the right by about 5mm or so. everyone with big feet that i've talked to has always found the MR2's pedals to be a bit crowded.

Now would be a great time to provide feedback on the design criteria.

 

[DefSport]

My 2 cents:

1. MR2 stock accelerator pedal is too low in stock form compared to brake pedal. Bringing it up 0.6-0.8" sounds about right (if you bring it up too high, it's hard to heel toe for those that have a lot of brake pedal travel).

2. The MR2 pedal spacing is pretty far in stock form. I would personally push it 5 mm TOWARDS the brake pedal. See E36/E46 M3 pedal spacing for reference (great heel and toe spacing):

Moving the accelerator 5 mm towards the brake gives about the same pedal spacing there as those cars, and they're great (size 12 shoes here).


I can measure pedal travel on a '93 MR2 with the stock '93 TB if you'd want. Where were you basing your linear measurement from, the top or middle or bottom of the pedal? It's going to have a bit of an arc, so might as well get similar data.

 

[Gouky]

I'm measuring pedal movement about 1" up from the bottom of the pedal and measuring to the seat cushion to orient it in the correct direction. And yes, getting the measurement from a stock MR2 would be great, thank you!

is that bimmer yours? if so can you take a picture with a tape measure across the center horizontally?

I'm glad you agree that the stock MR2 pedal is a bit low.

 

[DefSport]

Also S2000 pedal spacing, which is good, but honestly a bit far of a reach. They're probably 5-6 mm closer together than MR2 pedals going off foot feel.

 

[DefSport]

Nope, not my current M3 - just found an internet pic. I've owned 2 over the years though, so the pedal spacing stands out as "good" in my mind. I personally think most German cars (good height correlation at ~0.8-1 G decel, not too far apart) get it much better than most Japanese cars (gas too far down at ~0.8-1 G of decel, a bit far apart generally).

The S2000 is pretty good though, I think they took typical Honda pedal spacing and tightened them up a hair and got the height right. I've moved pedals around before and found that 5-6 mm of movement up/down or left/right can completely change how things feel when trying to heel toe.

 

[Gouky]

I agree that just a few mm can make a big difference. Thankfully the up and down can easily be solved with a small spacer. There's enough extra travel that i could sell an optional 2mm shim that you could stack a few of to get your ideal pedal height. Or course it would increase the dead zone at the bottom when you do that. But left/right alignment is a bit more permanent unless i get a bit more crazy with the design.

 

[DefSport]

Stock '93 3S-GTE TB, Stock carpet, MR2 heaven floormat (cheap, thin). Tape measure resting on top of seat and I picked a consistent place on the firewall to make linear measurements from.

Pedal resting position about 4.25" up:



Pedal at WOT, about 1.25-1.375" from carpet (hard to get a head on view that far down):




So I'd call it about 2.9-3.0" of pedal travel on a stock '93 MR2.

I think I'd be fine with 2.9-3.25" of pedal travel in my car if I had to put a number on it (stock '93 SW20 is as short as I'd go personally, I'd be fine with a hair more but it all works as you stated).

 

[Gouky]

ok, i'm measuring a bit further up on the pedal so to compare to my measurements it looks like it would be closer to 2.7-2.9" or so. So it looks like the rubber over-limit element on the pedal does get used a fair bit if you smash the pedal to the floor.

so i think my travel target is correct then. with a 16 degree sensor i can get that with about a 10" pedal length.

 

[DefSport]

I was drawing a line up to a point about where the wider pedal part merges to the main pedal body. I was keeping the tape measure out of the way to clearly show the angle of the pedal. Exactly where the tape measure is things were probably more like 3.1-3.2”. About 4.5” to 1.3” ish.

It was tough taking pics in the footwell, but that at least gives perspective without tape measure blocking stuff.

It does look like the pedal has hit the carpet some, so I imagine the rubber stopper gets some usage.

 

[Alex W]

Issues of pedal spacing and other physical aspects aside, I think the biggest thing to make sure is that 100% =100% throttle. That was the only complaint I had with mine. The throttle reached 100% at about 75% pedal, and the last 25% was a dead zone. So even if the pedal has 3" of travel, all of your throttle modulation has to happen in the first 2.25" of that.

What might be an easier project than to build a completely new pedal is to build a plug-and-play electronic adjuster that can plug into the pedal and allow the output to be fine tuned. Or maybe it's not an either-or question? I considered doing something like this when I built the adjustment circuit for my car, but designing and manufacturing electronics (even simple ones) is a bit outside of my area of expertise.

 

[Gouky]

Issues of pedal spacing and other physical aspects aside, I think the biggest thing to make sure is that 100% =100% throttle. That was the only complaint I had with mine. The throttle reached 100% at about 75% pedal, and the last 25% was a dead zone. So even if the pedal has 3" of travel, all of your throttle modulation has to happen in the first 2.25" of that.

What might be an easier project than to build a completely new pedal is to build a plug-and-play electronic adjuster that can plug into the pedal and allow the output to be fine tuned. Or maybe it's not an either-or question? I considered doing something like this when I built the adjustment circuit for my car, but designing and manufacturing electronics (even simple ones) is a bit outside of my area of expertise.

A lot of people complain about the look of the plastic pedal and the fact that it's a bit too low so i'm working on fixing that at the same time.

The current pedal has about 2.75" of mechanical travel and about 3/8" needs to be reserved for carpet and firewall flex to avoid stressing the pedal assembly. The current design reserves a bit less of the pedal travel than it really should.

So with all those things i think a redesign is due. If you just want an electrical solution there are some out there already, someone would need to evaluate them to figure out which one is the best but here's an example: Pedal Commander PC55 Bluetooth (Toyota Auris)

Also, manufacturing electronics is something i can do but I'm really concerned about the safety implications. If you alter the output just by scaling the pedal you end up defeating some of the failure modes that the ECU can detect and if you put a microcontroller in the loop you end up with a significantly more complicated solution. So it's perfectly fine for a DIY solution but even though the risk is quite small I'm not willing to take the liability for altering someone's throttle curve and defeating some of the detectable failure modes.

I wish i could figure out how to just tell the software in the ECU that 3.6V is the 100% point because then all the other safeties would still be in place but as much as i try i can't seem to figure that out. It's in the code somewhere for sure and the right person could figure it out but it does not seem to be something for me. I did change the curve on the sienna ECU but that one uses a different pedal control entirely. I can't find the equivalent tables on the rav4/highlander. But even that does not change where 100% is, just that the bottom end is less aggressive.

 

[Alex W]

A lot of people complain about the look of the plastic pedal and the fact that it's a bit too low so i'm working on fixing that at the same time.

The current pedal has about 2.75" of mechanical travel and about 3/8" needs to be reserved for carpet and firewall flex to avoid stressing the pedal assembly. The current design reserves a bit less of the pedal travel than it really should.

So with all those things i think a redesign is due. If you just want an electrical solution there are some out there already, someone would need to evaluate them to figure out which one is the best but here's an example: Pedal Commander PC55 Bluetooth (Toyota Auris)

Also, manufacturing electronics is something i can do but I'm really concerned about the safety implications. If you alter the output just by scaling the pedal you end up defeating some of the failure modes that the ECU can detect and if you put a microcontroller in the loop you end up with a significantly more complicated solution. So it's perfectly fine for a DIY solution but even though the risk is quite small I'm not willing to take the liability for altering someone's throttle curve and defeating some of the detectable failure modes.

I wish i could figure out how to just tell the software in the ECU that 3.6V is the 100% point because then all the other safeties would still be in place but as much as i try i can't seem to figure that out. It's in the code somewhere for sure and the right person could figure it out but it does not seem to be something for me. I did change the curve on the sienna ECU but that one uses a different pedal control entirely. I can't find the equivalent tables on the rav4/highlander. But even that does not change where 100% is, just that the bottom end is less aggressive.

Yeah, the liability issue was the other reason I pretty quickly decided not to pursue it. If it was anything EXCEPT a Toyota DBW throttle system I might not worry about it, but you sell something that somehow causes an "unintended acceleration" issue, with all of the legal precedent there, you would be so screwed if it ever came to court. I put a disclaimer on my blog about my DIY solution for the same reason, for whatever that is worth.

I could be wrong, but all of the stuff already out there seems more focused on increasing the pedal sensitivity even more! Maybe the "eco" mode would fix my complaint, but it's hard to know. Personally I'm happy with my 50 cent resistor solution though.

I guess I don't see the concern over stressing the pedal. Mine bottoms out on the internal pedal stop, not on the floor (carpet, no floor mat). You could raise the whole pedal assembly (something you are talking about anyway) so there would be room for a floor mat, but still have 100% of the pedal travel usable.

I agree, reprogramming in the ECU would be ideal.

Kidding of course...

 

[Gouky]

lol that image looks about right.

what part number is your pedal? my design was intended to bottom out on the floor to not stress the pedal assembly. But there are a ton of pedals that fit on that adapter and i really did not do a good job of documenting the "right" pedals since i did not realize there were so many differences.

 

[Alex W]

lol that image looks about right.

what part number is your pedal? my design was intended to bottom out on the floor to not stress the pedal assembly. But there are a ton of pedals that fit on that adapter and i really did not do a good job of documenting the "right" pedals since i did not realize there were so many differences.

I'm pretty sure my CTS pedal was off of your part number list, I'm not sure about the Denso pedal.

According to the confirmation emails from eBay from when I bought them, the Denso pedal is 78110-06010, and the CTS pedal is 78110-07011

EDIT: Just looked at your site, I guess there is no part number list, just that it should be an Avalon pedal. Mine were both sold as Camry pedals, but I thought I read on the forum somewhere that they were the same? I'm not sure.

 

[Gouky]

Yeah, that ambiguity was 100% my fault. I started with an avalon, that car is in storage so i haven't measured it's curve but i bet it's a 16 degree pedal and there's a bunch of other pedals that are mechanically very similar and they output a signal that the ECU was ok with. I never looked into it quite deep enough when i started this stuff. Heck, when i first did the swap in '07 only camry and avalon pedals were available. The rav4 was also available but it was completely different.

Also, i'm not sure if you've opened up the newer pedal (it's been out for a long time but newer than what wasinitially available) it's a much nicer design internally. I plan on showing off the internals of both in a video when i get to that point.

 

[Alex W]

Yeah, that ambiguity was 100% my fault. I started with an avalon, that car is in storage so i haven't measured it's curve but i bet it's a 16 degree pedal and there's a bunch of other pedals that are mechanically very similar and they output a signal that the ECU was ok with. I never looked into it quite deep enough when i started this stuff. Heck, when i first did the swap in '07 only camry and avalon pedals were available. The rav4 was also available but it was completely different.

Also, i'm not sure if you've opened up the newer pedal (it's been out for a long time but newer than what wasinitially available) it's a much nicer design internally. I plan on showing off the internals of both in a video when i get to that point.

Yeah, that makes sense. An extra few degrees of pedal travel would likely cause it to bottom out on the floor instead of the pedal stop, and a lot of the dead zone that I didn't like would be eliminated because you would just never get there. And of course the usable travel would be over a wider range, so modulation would be easier as well.

I have not taken any of them apart, no idea what's going on inside

 

[Gouky]

I've characterized a few of the 16 degree pedals i have here. I just made a rig with a 5V power supply, a digital angle gauge and a volt meter and got some interesting data: (this is only plotting the "A" sensor. The confirmation signal has no additional information here)

While my setup wasn't metrology grade, the differences between the curves above is significantly above the error so even the 16 degree pedals have a different engagement but as we see they are all generally linear. The major difference is right in the first 3 degrees or so. one of the pedals started reporting a higher voltage immediately with very little pressure on the pedal. Most had a soft start and one has a dead zone of about 1.5 degrees.

I'm picking the dark blue line above as the base sensor for my pedal. It has about 1/2 degree of dead zone at the top then about 1 degree of soft start before the linear area. Because the voltage never catches up from the soft start it means that the soft start does not eat into the usable range of the pedal.

As you can see, the pedals are actual capable of sensing 20 to 21 degrees of motion. Mechanically the motion is about 24 degrees max but at the end, this makes a difference between being able to use a 11" pedal length vs a 9" pedal length. So with the custom pedal it really won't make a difference in the end.

 

[Alex W]

It's so odd how Toyota will go decades without changing certain things. Like the E153 bolting up to the 2GR 15 years later, indicating that in all that time they hadn't (completely) changed not only the bell housing pattern, but also things like the offset between the end of the crank and the bell housing flange. Or axle spline geometry that is the same over a 20 year time period.

On the other hand, there are dozens of different pedal assemblies out there, and electrically they seem to change connector shapes and pinouts at the drop of a hat.

 

[Gouky]

Yeah, i honestly don't know why there are that many pedal assemblies. and even if the assemblies are different, why is the sensor different when the ECU could be mapping whatever it wants any way it wants. So it should all just be a very linear transfer function.

 

[DefSport]

I think some of the modern electrical component churn are connector suppliers coming out with new designs frequently, and input device teams likely refining these pieces continually for reliability and lower cost. There’s not much money to save in tweaking a transmission bolt pattern, but if I can save $3 per pedal on every Toyota car with another p/n roll, it’s cake and ice cream day in the accounting offices!