On Fri, Apr 30, 2004 at 06:21:42PM -0400, James Nazarian wrote:
> Larry, what about a linear encoder, or rotary encoder with bellcrank? I
> would think that you could get one on the piston and a rotary encoder on the
> throttle and log the two in reference to each other. I can't imagine it
> would be that hard. Being a computer nerd, you could probably even whip up
> an RS232 app that could log it all to a laptop.
There is something odd about this SU such that if the cap is missing,
the car won't run. So, I went through all sorts of gyrations to make a
widget that would measure piston height and not bind:
http://red4est.red4est.com/jasaside/carbpist/
Then there is the issue that multi channel dataloggers are not
cheap. My dmm has rs-232 out (and it was only like $70 at Radio Shack)
but that is only one channel, without being able to compare it against
anything else, it wasn't too useful.
What I finally ended up deciding is that by the time I managed to work
out all the instrumentation I'd need to properly dial in the carb
needle, I'd have about everything I needed to control a fuel injection
system.
As long as what I've got works "well enough", I'd actually be kind of
silly to invest the R&D into getting it to the next level (whether it
be better tuning of the carb or FI) except as a learning project.
If I can find a few other people intrested in investing several times
the value of the car into developing 150-180 hp MGBs, the R&D costs
become more managable.
A couple other sensors I'm curious about:
NOx for on the car, to detect detonation. But I suspect EGT would do
nearly the same.
Mass Airflow for in the exhaust, I'm curious as to how that relates to
power, as well as MAF into the motor.
I'm sure I could have lots of fun if I had access to GM's engine
development lab.
--
Girling is not a verb.
lrc@red4est.com http://www.red4est.com/lrc
|