Engineering units for Pulse and Glide

[bwilson4web]

Hi,

I've never found more than an 11% improvement in mileage using a reproducible Pulse and Glide (PnG), cruise control based procedure. This technique is based upon using the cruise control, resume speed function, to give a consistent acceleration. Glide is handled by shifting into "N". Sad to say, PnG is usually described with imprecise, manual processes without high-resolution engineering data.

I've found and ordered instrumentation to gather engineering data:

This unit has a high resolution, 320 Hz, sampling interval and built-in, rechargeable battery. They also have a more affordable unit.


This unit has a sampling rate of 160 Hz and uses a removable battery. This is the unit I'll "loan" to those interested in sharing engineering data about vehicle handling. For example, mileage runs could use these to document the driving profile. Combined with a GPS log, complete ephemeris data would be available for post event analysis. Together, these will finally give us engineering units.

Bob Wilson

[RobertSmalls]

I've been lusting after National Instruments' USB-6008 DAQ, which is $170. 8 channels accepting -10V to 10V, 10kHz sample rate, 12b resolution, transmits data to a laptop in the car via USB. NI's LabView software allows you to translate those voltages back into temperatures, speeds, etc. in real time if you like.

Of course, NO sensors are included. However, most of the sensors you might want are already installed in the car. Others can be purchased seperately, such as:
Jameco Electronics Robotics: PARALLAX INC.: 28526 which could be epoxied to the frame of your car.

I would like to log wheel speed (and compute two differentials thereof), fuel injector #1 activity, HV battery current, ambient air temperature, upstream (wideband) O2 sensor output, and any other sesors relevant to whatever I'm working on.

At the very least, this would yield a lean-burn prediction algorithm, and give me a breakdown on where I spend energy on my commute. Not just at which mile-marker, but it should also give me an idea of whether it's due to aero, rolling, or braking.

What I really want is to break MPG into its constituent components: powertrain efficiency (BSFC) and energy consumption per mile (rolling, aero resistance, and braking, predominantly). In an ideal world, you'd have a gauge displaying each in your car.

I haven't done it yet because it's time consuming and expensive. But it would yield top-quality data, and lots of it.

[bwilson4web]

That is a nice $35 part and I agree about the interest in full vehicle data recording. I have a G3 Mac laptop and I've been thinking about a permanent installation in the NHW11 connected to the Graham miniscanner and a GPS recorder. But this 3-axis, USB stick provides a nice, $90, addition, that can easily be used standalone for detailed metrics.

There is very little, hard engineering units on PnG. This little unit along with a GPS track should provide what has been missing, engineering units. Add an ICE rpm track and a complete, detailed, energy model will exist.

Bob Wilson

[orange4boy]

Nice find Bob!

Can't wait to see the graphs you come up with using that. You are the graph king!

I'm loving the Graham scanner and I'm going to be sad to see it go back to the tool library but perhaps I will find another or convince Graham to make another or at least burn a chip for me so I can build my own. SGII would be a close second, I guess.

It's been great to have it. Thanks for making these available. Excellent idea.

[bwilson4web]

Originally Posted by orange4boy

. . .
I'm loving the Graham scanner and I'm going to be sad to see it go back to the tool library but perhaps I will find another or convince Graham to make another or at least burn a chip for me so I can build my own. SGII would be a close second, I guess.
. . .

There is a third possibility, a laptop application and USB based CAN-bus interface. Check the messages over at PriusChat for the last three months. The guy in California who was moving to Florida and had the the problem transaxles was using it to read codes. I'm not sure it can clear codes but it sounds like a dandy reader. For example, being able to read all 19 module pairs. But it needs a laptop.

There was a hack five or more years ago to split the VGA into RGB and map a laptop screen on the navigation inputs. I also understand the touch sensor returns an ASCII string. The most aggressive installation put a headless Mac in the car so the Nav display and touch screen was the interface.

As a loaner, you can see how valuable the Graham scanner is. I believe about 200 were made and I was lucky to snag first one and then a second. I have a standing offer to buy any that come available so we can have a pool.

Bob Wilson

[orange4boy]

Right now I'm watching the injector milliseconds and the #1-O2 sensor for my cold start driveway routine. When #1-O2 starts to fluctuate, I zip up the driveway looking for the lowest microseconds. When the O2 starts to jump around it's out of open loop. the bypass cat section will absorb the nasties till the cat proper can take over if it's not already hot by then.

I'm also watching the INJ MS for the best pulse/accell rpm so far ~ 2200 rpm = ~ 6.125ms and the best accell.

I'd never be able to figure this out without the Graham.

I have a mac powerbookG4 but all the software based programs I have seen are on PC

The autoenginuity is the best one I have found that covers even the hybrid subcodes with the toyota upgrade.

[Frank Lee]

The cc floors it, and that's too much.

[bwilson4web]

Originally Posted by orange4boy

Right now I'm watching the injector milliseconds and the #1-O2 sensor for my cold start driveway routine. When #1-O2 starts to fluctuate, I zip up the driveway looking for the lowest microseconds. When the O2 starts to jump around it's out of open loop. the bypass cat section will absorb the nasties till the cat proper can take over if it's not already hot by then.

You're in an area I can't easily replicate. We live on former river bottom land, mostly flat. The nearest, steep stretch is barely 3 meters in height.

Originally Posted by orange4boy

I'm also watching the INJ MS for the best pulse/accell rpm so far ~ 2200 rpm = ~ 6.125ms and the best accell.

This is a good range. This is my warmed-up, 1.5L, NHW11 BSFC chart:

I used MG1 torque which is 28% of the ICE torque to calculate the ICE power. Either injector timing or for a warmed up ICE, mass flow works. In practice, I use 2,400 rpm as my upper limit but up to 2,600 should be OK. The problem is at 2,600 rpm there can be some overshoot:

• 2,200 -> 25 hp
• 2,400 -> 27 hp
• 2,600 -> 30 hp

Here is another chart:

Again, it shows the optimum ICE rpm ranges.

Originally Posted by orange4boy

. . .
The autoenginuity is the best one I have found that covers even the hybrid subcodes with the toyota upgrade.

That is the one. My wife has a pentium based Mac and we may try it someday.

If you decide to go with some discrete instruments, you can run the cables through the air inlet on the passenger side and drop them down to the cabin filter behind the glove compartment. Drill a hole through the side of replaceable filter and run the wires inside.

Bob Wilson

[luvit]

i, too, have been called graph king. but today my eyes have been opened of how i was never worthy of thy name.
i bow down to you, graph king.

[gone-ot]

"...presentation, presentation, presentation..!"