E-gas now road tested, mostly successful.
This has been on the back burner for the past several months but I discovered there's a plastic tab on the back of one of the electric TB gears that can be removed so that it closes all the way without the 5 degree "limp home" detent.
That allowed me to reconfigure the the motor drive circuit and I could smoothly open it and close it. I also biased the TB feedback voltage to match the M50 TB voltage at idle.
So, installed it and it idled perfectly. Revved nicely with the electric pedal in the garage so I took it for a drive. Here's what I learned:
1) It's hard to drive with the gas pedal in your hand.
2) E-gas is smooth with great finesse and precision of control.
3) My mapping for pedal position to TB position is perfect.
4) 413 ECU was happy with it.
5) Cat-5 cable: it's not just for your network.
6) You really could be a backseat driver...
The problem I had is also siginificant. I am depending on the spring to close the throttle plate. That takes about 1/2 second.
Not fast enough!
On a hard 1-2 upshift the TB would still be closing when I depressed the clutch and the motor would rev in the middle of the shift.
The M50 TB snaps closed in about .1 second or as fast as you can release the pedal.
So.. I need to go back to full H-bridge so the plate can be driven closed.
It sure was fun!
E-gas
Collapse
X
-
Take 3...
getting better.
I've been wrestling with the stupid detent position on the throttle body trying to finesse a solution that would let it open and close without the detent making it too jerky. I was trying to drive the motor in one direction at a time only.
Finesse is out.
I changed it to complementary PWM where the motor is driven in both directions AT ONCE. (actually, it switches directions at like 500Hz) It's in a constant tug-of-war with itself but the detent basically becomes invisible.
The downside is that the MOSFET gets much hotter because it's delivering 4x the current.
So, engine idles pretty nice and responds to the Egas pedal smartly. The pedal is also mapped so that it's not super sensitive off-idle like it was before.
I'll do some more smoothing on the controller and beef-up the heatsink on the MOSFET before I do the first driving test. It'll be like a handicap car with the gas pedal hand-controlled. Should be fun!
Throttle cables still rule for now.
In the picture I have the M50 TB moved aside and the electronic TB installed. You can also see my brake booster clearnance with a 320i booster.Leave a comment:
-
Dude, this bad ass.
Your skill impresses me.
I'm paying you to come to Renton if I can't get my m52 wiring figured out.Leave a comment:
-
No, physically they have slightly different designs but the wires do the exact same thing.
What's funny is that mine even has a little kick-down botton at the very end of the stroke, even though it's connected to exactly *nothing*.
I guess it makes people feel good to press that button when they floor it. The result of a $2 million dollar BMW marketing research, no doubt.Leave a comment:
-
good info... the e39 m5 pedal looks to have the pot/pots outside the pedal attached to a linkage where your pedal (m54?) looks to have everything in one unit. I would like to have the single unit for ease of install but I was afaid they werent interchangeableIt's not just a potentiometer... it's two and they work in opposition to each other. While one is ramping the voltage up from .5 volts to 5, the other is decreasing from 5 down to .5.
Same thing on the TB. This allows the computer to validate the signal and prevent a short or stray signal from accelerating you unintentionally.
Also, there's no contact in the pedal, it's a rotary Hall sensor. Very
smooth!
I think all current BMW gas pedals work the same, although some are top hinge and some are bottom.
I've got mine working pretty well in prototype form. The issue I'm having now is that my lab power supply trips over 1.5 amps and I need to run it off the battery.Leave a comment:
-
It's not just a potentiometer... it's two and they work in opposition to each other. While one is ramping the voltage up from .5 volts to 5, the other is decreasing from 5 down to .5.question... what model is that pedal from? does it matter what pedal you use( are they interchangeable?) i assume they are just a 0-5v potentiometer. This is for a standalone (ViPEC)and the engine i purchased didnt come with its e-gas pedal. I was hoping I could get one that fits the e30 better
Same thing on the TB. This allows the computer to validate the signal and prevent a short or stray signal from accelerating you unintentionally.
Also, there's no contact in the pedal, it's a rotary Hall sensor. Very
smooth!
I think all current BMW gas pedals work the same, although some are top hinge and some are bottom.
I've got mine working pretty well in prototype form. The issue I'm having now is that my lab power supply trips over 1.5 amps and I need to run it off the battery.Leave a comment:
-
Leave a comment:
-
question... what model is that pedal from? does it matter what pedal you use( are they interchangeable?) i assume they are just a 0-5v potentiometer. This is for a standalone (ViPEC)and the engine i purchased didnt come with its e-gas pedal. I was hoping I could get one that fits the e30 betterLeave a comment:
-
Thanks for the updates. I'm happy to see you are working through some of the problems and figuring out solutions.Leave a comment:
-
-
Second try...
I changed the approach completely. The electric TB has a limp home default position where it goes to 5 degrees open when it loses power, I need to drive it closed AND open.
So, with the new full H bridge driver I installed it and got the car to idle.
Now my task is to clean up the noise to keep the PIC from resetting and to change the TB response to the pedal. I learned today that the linkage on the mechanical TB does a great job at controlling the amount of opening vs pedal position. It doesn't open as quickly for the first bit of pedal travel.
I found that with 100% linear TB movement vs pedal it opens WAY to quickly off idle. I need to make it less sensitive at partial throttle.Leave a comment:
-
Forgot to mention
6) This is my training wheels for a possible N52 project in the future. That motor has the fly-by-wire cylinder head that controls intake valve lift directly.Leave a comment:
-
1) The M54 TB is 20% larger than the M50 TB which was sized for a 2.5l engine. It should theoretically make about a 2-3 hp difference which isn't a big deal, but the fit will also be better.
2) The pedal is smooth as butter. My cable pedal isn't bad either, but the electric one is superb.
3) The responsiveness is a matter of programming. As in everything else with the M54, it has to do with emissions and the "average" driver. I can actually make it much more responsive by reading the derivative of throttle position. A quick stab will open it more than the actual displacement of the pedal.
4) I can change response curves completely.
5) It's fun watching 1 amp chips explode when you put 3 amps through them!
See www.sprintboosterusa.com for a product that restores the responsiveness. I'll be doing the same thing organically.Leave a comment:
-
out of curiosity why did you go with an electric TB?
I came from an e46 with an electric TB and after going into an e30 I don't understand why we decided to move to electronic management as I feel a degradation in responsivenessLeave a comment:

Leave a comment: