Date: Mon, 22 Feb 2010 19:23:16 -0800
Reply-To: John Bange <jbange@GMAIL.COM>
Sender: Vanagon Mailing List <vanagon@gerry.vanagon.com>
From: John Bange <jbange@GMAIL.COM>
Subject: Re: ISC troubleshooting
In-Reply-To: <de48c0be1002220753x37535905i7270d6a97c59a6d7@mail.gmail.com>
Content-Type: text/plain; charset=ISO-8859-1
> Madness, indeed. I like to imagine some line workers at VDO toiling away,
> hand-soldering every pin...
>
> John, what do you reckon all the inputs to the unit are? There's a
> fair-sized bundle of wire going to it.
>
Well, I've traced them all out in the Bentley and come up with the
following:
R/BK = 12v from STARTER
R/W = 12v from ignition switch ON
W/Y = Power Steering (12v on pressure switch close)
BR = Ground
GY = TEMP2 sensor (thermistor, 150-6500 ohms to ground)
R/BL = Throttle Switch (ground at idle, full throttle; open all other times)
W = Idle Stabilizer Valve +
Y = Idle Stabilizer Valve -
W/G = Air Conditioning (12v when AC is on, open otherwise (presumably))
G = Hall Sensor Pulse (at least 10v)
Looking at the inputs, the functions of the idle stabilizer are pretty clear
(though they raise a few questions):
if (Throttle Switch is not GROUND) then DO NOTHING
if (STARTER input 12v) then open ISV to (some percentage, perhaps determined
by TEMP2 ?)
set IDLE SPEED to 850
if (Power Steering input 12v) then set IDLE SPEED to (something, 1100 or
so?)
if (Air Conditioning input = 12v) then set IDLE SPEED to (something, 1100?)
if (TEMP2 says engine is cold) then set IDLE SPEED to (something, 1100?)
FINALLY
calculate RPM from elapsed time and pulses counted from Hall Sensor
if (RPM < IDLE SPEED) then open ISV (some increment)
if (RPM > IDLE SPEED) then close ISV (some increment)
Since I don't have an oscilloscope, I am limited in my analysis of what
EXACTLY the thing tries to do with the ISV for these various inputs, but I
do know that it generally raises the RPM for all of them. My controller and
valve are old and cranky, so they don't really keep the engine at a steady
speed--- I can only determine "relatively faster", and "relatively slower".
So far I have built a couple annoying prototype digital controllers. I blew
up the first one (5v controllers don't like 12v). The second one is a little
better planned, but I still can't get the timing in the software tuned in. I
get weird oscillations and drop-outs... when I change variables to fix one,
I seem to exacerbate the other. I thought doing it in software would be
easy. It's a lot easier than that crazy OEM analog unit, but still... I have
an even greater respect for the VDO guys now.
Someday I expect to get FILTHY RICH selling these for $249 to people who
don't want to spend $250 for a VDO OEM controller! Actually, I'll be lucky
to get one working at all, and being an eternal pessimist, all I can imagine
is three dozen customers with badly functioning units badgering me for the
rest of my life. No, if I ever get a working design, I'll just write it up
on a web page and let people HAVE it for free. At most I'll make PCBs and
sell kits!
My biggest annoyance is that I can't find a "double relay" connector that
will fit in the existing socket. RIght now I'm using one stolen from a bad
controller I already had...
So many projects, so little time.
--
John Bange