My new way of measuring on-road drag

[M_a_t_t]

Originally Posted by JulianEdgar

Interesting! I wonder if most of that 50 ohm variation is noise? You can be sure the ECU filters the TPS input.

At this stage, I'd just go ahead and do some aero testing like windows up/down and see what happens. If you are consistent in the change in speed, and the change in speed makes sense, then do some more subtle aero mod testing.

I was also thinking about how the ECU uses that sensor. It doesn't directly determine anything. So if that sensor has a little bit of variation (not necessarily all of what I am finding) then it won't bother the ECU because it is only looking for major changes. I.e enrichening at WOT.

I need to grease the 2 rear regulators up then hopefully some testing will be done. I still don't have a great road found, but I have 2 that show some promise. The speed limit on one of them is pretty low though. Only 45 mph the other is 55 posted, so "60-65" mph to the general public

[JulianEdgar]

Originally Posted by M_a_t_t

I was also thinking about how the ECU uses that sensor. It doesn't directly determine anything. So if that sensor has a little bit of variation (not necessarily all of what I am finding) then it won't bother the ECU because it is only looking for major changes. I.e enrichening at WOT.

I need to grease the 2 rear regulators up then hopefully some testing will be done. I still don't have a great road found, but I have 2 that show some promise. The speed limit on one of them is pretty low though. Only 45 mph the other is 55 posted, so "60-65" mph to the general public

No, it will also be vital in providing the right acceleration enrichment and transient timing advance during throttle changes.

Probably has (my guess only) 1/10th of a second filter.

What are the rear regulators?

(Addition: In the very old days - like the late 1970s - EFI cars in fact ran throttle position switches rather than sensors - just idle and full throttle contacts. My 1977 BMW 3.0si with Bosch L-Jetronic had that.)

[M_a_t_t]

I meant in this specific case. If I go to pass someone and open the throttle 2.3% its not going to enrich (I have watched the wideband). Which means it should be close.

Rear window regulators

My 86 300zx is like that. For the manual transmission anyway. The auto has a variable rate one.

[Piotrsko]

Had to go back 4 pages to see what the op issue was. Tps are no longer a constant resistance strip like an audio amp potentiometer, they are more like a wire wound resistor such that each decrement is an additional series resistor added. You cant get to 25 more (half spacing) ohms linearly. The steps are close together, sometimes within the width of the wiper so that it is conceivable that .003" wiper arm movement makes or breaks contact with an element. Won't even go into noise and dirt in the measurement system affecting values. The actual values are readable but I will not attempt to tell you where they are PID'ed in Torque. The software types know about this issue so they use a bunch of sensors to decide engine load requirements and adjust accordingly.

Suffice to say: engine load is reasonably close ratio to throttle position, and probably within the users measurement system error bucket. Unfortunately, you might have to live with using engine load instead of throttle position.

[JulianEdgar]

Originally Posted by Piotrsko

Had to go back 4 pages to see what the op issue was. Tps are no longer a constant resistance strip like an audio amp potentiometer, they are more like a wire wound resistor such that each decrement is an additional series resistor added. You cant get to 25 more (half spacing) ohms linearly. The steps are close together, sometimes within the width of the wiper so that it is conceivable that .003" wiper arm movement makes or breaks contact with an element. Won't even go into noise and dirt in the measurement system affecting values. The actual values are readable but I will not attempt to tell you where they are PID'ed in Torque. The software types know about this issue so they use a bunch of sensors to decide engine load requirements and adjust accordingly.

Suffice to say: engine load is reasonably close ratio to throttle position, and probably within the users measurement system error bucket. Unfortunately, you might have to live with using engine load instead of throttle position.

TPS is just being used to check that a constant throttle is being held - no more or less.

[Cd]

What's keeping me from trying this method, is figuring how far to set the screw into the wood block that goes under the gas pedal.
How have you all figured this out ?

I even thought about trying to draw a graph that I can place near the pedal.
I would try and set my phone to record video, and have it placed near the gas pedal.
As I got up to max speed, it would record where the pedal is as shown on the graph.

But this not only seems cumbersome, but very dangerous as well.

BTW, the car is an old 1993 Honda Civic.

Thanks !

[aerohead]

Originally Posted by Cd

What's keeping me from trying this method, is figuring how far to set the screw into the wood block that goes under the gas pedal.
How have you all figured this out ?

I even thought about trying to draw a graph that I can place near the pedal.
I would try and set my phone to record video, and have it placed near the gas pedal.
As I got up to max speed, it would record where the pedal is as shown on the graph.

But this not only seems cumbersome, but very dangerous as well.

BTW, the car is an old 1993 Honda Civic.

Thanks !

In Australia, Julian's limited to 110-km/h ( about 67-mph ) as the upper boundary for velocity.
For Austin, if you could get on the toll-road to San Antonio, 'abandoned,' you could set your upper boundary to 85-mph.
Your baseline would be below that, such that the air drag reduction would create a velocity excursion up to, but no faster than 85. No way to know in advance.
If you had to perform a 'power-on' evasive maneuver you'd be screwed!
Perhaps a length of string from the dash, to the top of the pedal could serve as the detent, and if necessary, simply mashing the pedal would 'snap' the string, allowing emergency acceleration to save your bacon.
Or a split, rigid segment, joined with a magnetic coupling which could be overwhelmed and separated with foot pressure.

[Cd]

I messaged Julian with the question a month or so ago, and he gave me a simple, non complex way to go about it, but I can't recall if it was through his Youtube channel, or here.
I can't recall what the solution was either, because I felt it was beyond my expertise at the time.

I reinterated the question on this post so that others may benefit.

Something that complicates things, is that to merge onto a road, I have to give quite a bit of throttle.
And of course if you suddenly have to dart into the next lane, you need to to use full throttle.
This is an underpowered Civic, and not a superquick Tesla.

The second idea you mentioned seems like it would possibly work for sudden emergency lane changes, but would still deform to more throttle than you need right as you start the test, because the throttle needs to be mashed down quite a ways to even merge onto a road.
The only way it seems like something like that could work, is if I featherfoot the throttle, and build up speed over a long distance.
( And in Austin that is not possible - even on sideroads.
I would have to travel out of town and find a country road )

[M_a_t_t]

What I did for my initial setup was leave it unscrewed and see how fast the car goes and then screw in if it wasnt fast enough.

However, i found that method annoying because it took forever to get up to speed. I talked to Julian about it later and he said he stopped doing that as well and started just watching the tps% on his dash unit. I am now using a parking brake lever and a bicycle cable that is attached to the gas pedal. This way I can give it more gas to get up to speed. To release its just like a regular brake lever, i push the button in and drop it.

[Cd]

Ah yes. That's what it was - Julian had suggested using the throttle position sensor.
My car does not have one of those that I am aware of, so I didn't give the suggestion much thought.

If I understand correctly, if you suddenly needed to quickly pass someone in an emergency situation while testing, you could do so by pressing the button on your parking brake lever and this would release the lever, and give more length to the cable connected to the gas pedal ?
Did you have to disconnect the actual parking brake and make it nonfunctional ?