Re Homebrew box

[Alex Samson]

Bruce Plecan was asking some questions about my homebrew box.  

>PIC for idle stabilizer, do tell.

The Subaru 2.2L flat-4 had a very nice air valve that runs across the
throttle valve, to allow idle correction during idle.  The valve has a
small DC motor that could turn clockwise or counter by means of current
flow on its leads to open or close the valve. I need to look at how
Subaru actually operated this thing in one of their cars one day.  When
this project began, I didnt want to implement this idle corrector
because I didnt know how to run it.  I left the engine to manage itself
at idle.  In the course of testing, it became apparent that this engine
needed help at idle.  It had extremely poor low end torque and its idle
was virtually powerless.  Off idle power on the Vanagon produced a
stalled engine in many cases.  Since the valve could just OPEN or CLOSE
and no intermediate positions were possible, the valve apparently had to
be modulated by a duty cycle waveform in order to setup partial
oppenings (or closures) according to the duty cycle of the switch that
operated it.  Since the 80C31's got very busy with their individual
tasks, it became apparent that another micro was the easiest way to
control the idle stabilizer.  One of the 80C31's communicated with the
PIC to tell it if the idle is falling below 770rpm and by how much.
This information allowed the PIC to decide how much opening (and duty
cycle) to open the air valve. I never perfected this but engine stall
off-idle was greatly lessened with this system in operation.  For one,
the valve was kinda noisy whenever it was pulsing (we didnt use the
original Subaru airbox).  There was speculation that since this Subaru
engine came from an automatic, its unique cam timing may have caused the
very weak idle torque we observed. 

>Where would one find out more about a "SRAM".

I used standard 32k byte CMOS Static Ram (SRAM) chips that were memory
mapped with the 80C31's external bus.  I used type 62256 in 28-pin DIP
and required about 10 microamps of 3v battery to maintain its contents.
I installed a write protect hardware switch on the box in order to
eliminate the possibility of a RAM corruption by accident. The chips
were wired as "Write like Xram and read like ROM". There is great
variation in standby current among manufacturers so one has to read the
fine print in their specs.  Mouser sells several equivalents to the
62256. 

>Ahh, 450's, onena my favorite rides.  Makes interstates take on a
>new meaning.

Yes indeed!  Makes me sleepy while driving one though, like when
relaxing on a sofa watching tv...

>OK, what how why where, on the engine simulator.  Just rpm/vss?,
>programmable?.  Just curious how your doing it, I'm still at
>breadboarding mine

My home-made D-jetronic bench has a distributor simulator, a bunch of
old injectors, a vacum sensor, assorted variable resistors for the temp
sensors, switches for the throttle assy, and a 12v power supply.  The
distributor simulator ran off a signal generator to simulate engine RPM.
There is no way one can build a homebrew box without some kind of  bench
simulator. This insures against the box assasinating the target engine.
Email me direct if you want more info.  The D-jetronic is a dinosaur by
todays standards so other guys on this list are gonna get bored.

I am located in the beautiful car capital of the world, Los Angeles. 

Regards to all,
alex samson