<all> home dyno kit??

Mike trinity at golden.net
Fri Jan 29 00:06:53 GMT 1999


>
>> At 08:24 AM 1/28/99 -0500, Dan Llewellyn2 wrote:
>
>> >level ground.  At least were I live, level stretches of road where
>> >you can get up to speed are rare.  If they incorporated the data
>> >from a G-field measuring device, like an Analog Digital ADXL05,
>> >you could calculate horsepower even if the road was not level.
>
>Yes, but you still need the speed input since the accelerometer
>cannot tell the difference between acceleration and gravity...
>As far as the horsepower calculation is concerned, you would
>use the acceleration from the accelerometer and the speed from
>another sensor...  The calculation (being mass * velocity * acceleration)
>doesn't care whether the acceleration is due to gravity (going
>uphill) or increase in speed.
>

I've got on the backburner a little performance computer project that hasn't
yet come to much (no time of course). The  idea at the time was to use an
ADXL202 in PWM mode.

If the accelerometer is mounted orthogonally to the gravity vector, it's not
an issue since its contribution is zero (cosine of 90). I'd argue that most
of this "on the road" horsepower testing will take place on a remote, level
road with little traffic or hills etc. Once found, this road should made the
standard to avoid adding additional uncertainties. If using an accelerometer
on hilly roads, all bets are off unless one somehow characterizes the road
by driving slowly and steadily along the road, recording the accelerometer
and logging distance and then factoring the results out of acceleration runs
later.

After reading some of the "audio" ideas recently, I thought about a slightly
different way to do performance calculating, using just the primary ignition
coil signal and the statistics of the car in question.

Why not hook (say) an HC11 to the coil primary and time period between
edges. If the car is placed in, say, second gear and floored from 1000 to
7500RPM (my PGT fuel cut), one could use the ignition signal as both an RPM
reference and also as a "delta-RPM" signal to determine how fast the engine
is accelerating. On my PGTs V6, there's 3 pulses per revolution of the
crank...is this enough I wonder? Time to pull out the notebook and start
calculating.

When I was playing with a G-Tech, I did the "2nd-gear flooring" thing and
noted the maximum g-reading during the run. I used this g-value along with
the loaded radius of the front wheels, the mass of the car (as weighed on a
garbage dump scale) I determined the front-wheel torque required to produce
this acceleration. By knowing the transmission and final drive gear ratios,
and figuring in a 15% drivetrain loss fudge-factor, I was able to get pretty
close to the factory torque figures for the car. Knowing the approximate
speed at which this occured I was also able to then calculate approximate
wheel HP and engine HP, which again, were pretty close to the factory numbers.

Of course, one problem with this scheme is that wind resistance isn't
accounted for which tends to reduce the acceleration rate as vehicle speed
increases.

>> One thing I was gonna mention is the Analog Devices ADXL202, a
>> second generation, 2 axis +/- 5g accelerometer. It's even easier
>
>I looked at it when it first came out and the specs didn't look
>very good to me.  It's bad enough getting .01g resolution out
>of an ADXL05 with any reasonable bandwith and the 202 looked worse to me.
>

My biggest problem with these accelerometers is that they have such a huge
dynamic range (like +/- 5g) and my car is capable, in first gear, of maybe
0.6g and about 0.3g maximum in second gear (hey...that's what 2.5L, 164HP
and 3058-lbs of car gets you :). At 12.5%/g, this amounts to a change in
pulse width of 3.75%. Not much to maintain accuracy. At zero g and 1mS
period, the output is 500uS. 3.75% is about 18.75uS or about 38 counts of a
2MHz HC11 timer clock. Not much to work with.

Sorry for rambling ...


--
 Mike




More information about the Diy_efi mailing list