[alfa] Re: 86 Spider no start

["Daegrigg (Don)" <daegrigg@xxxxxxxxxxx>]

John,

This is fantastic info.  I did confirm that the sensors are in the right
position - reference (lower) is gray connector, rpm (upper)  is black
connector.  I cleaned one of the connectors but for some reason not the
other.  This is consistent with the car starting a few times after didling
with the connectors in that area.  Also, this probably means that I don't
have a bad fuel pump.  Found out (reading the shop manual) that the 86 has
two pumps - a secondary in the tank and the main under the car.  The
secondary runs in the "on" key position but the main only runs in the crank
and run condition.

Soooo - it back to cleaning the connectors.  Unfortunately all I have is
emery paper and wd 40 but that should surfice.

I'll let you know what happens.

Don

----- Original Message -----
From: John Fielding <johnf@domain.elided>
To: Daegrigg (Don) <daegrigg@domain.elided>
Cc: <alfa-digest@domain.elided>
Sent: Tuesday, October 14, 2003 12:06 PM
Subject: RE: 86 Spider no start


> Hi Don,
>
> Your AR Spider has the Bosch Motronic system, a derivative of the
L-Jetronic used on
> earlier vehicles.  It so happens that I know an awful lot about L-Jetronic
& Motronic
> especially the AR ones.  The only difference between the Motronic used
here and the USA
> version was the inclusion of an O2 sensor in the USA spec vehicles.
>
> You have no fuel injector pulses because you have no spark!
>
> The L-Jetronic used the pulses from the coil -ve terminal to clock the
injector ECU to
> give engine speed information.  If the ignition is bad the fuel will not
be pulsed.
>
> The Motronic design built on the L-Jet design with a number of subtle
changes, the main
> one is that is there is no switch in the AFM to energise the fuel pump.
Bosch changed
> this when the Motronic was designed, instead the ECU has to see valid
crank sensor signals
> before the pump relay is turned on.  The fuel injection side is identical
to the earlier
> L-Jetronic, except the speed signal which used to come from the ignition
coil -ve terminal
> now comes from the coil driver trigger.
>
> The ignition portion of the Motronic ECU needs two signals from the crank
sensors.  One is
> the rpm or speed signal which is the magnetic pickup that sees the starter
teeth on the
> flywheel.  There are 131 teeth on the starter ring gear, so for every
revolution the ECU
> expects to see 131 pulses.  When it has "counted" 131 pulses it knows that
1 complete
> revolution has occured, but it has no idea what the crank shaft angular
position is.
> This is gets from the second magnetic pickup.
>
> The second signal is called the "Reset" transducer and has an identical
> magnetic pickup but this one sees a single pin or lump on the flywheel.
This gives
> crankshaft angular position information.  On the '86 Spider the pin occurs
at 57.5 degrees
> before #1 TDC.  The Motronic computes the ignition advance using this
"epoch pulse" to
> define where #1 TDC is and from the rpm signal to compute how fast
> the engine is turning.  (There is a limited built in default mode where it
can provide a
> fixed idle timing and injector pulse width if the rpm signal is out of
specification but
> this does not allow the engine to be revved up).
>
> In order for the ECU to start looking for pulses several things need to
happen, apart from
> being powered up by the ignition switch being turned on.  There is no
"Keep-Alive" power
> connected to the earlier Motronic ECU's such as yours, this only occurred
with the later
> versions.
>
> 1).    The starter solenoid activating signal is routed to the ECU Pin 4.
When this line
> goes above 6V the ECU switches on and starts looking for rpm signals from
the starter
> ring gear.  (This fixes the ignition timing at a low advance and a fixed
injector pulse
> width - computed from water temeraure - until the engine speed exceeds
approx. 500rpm).
>
> 2).    If the rpm signal is present and above 100rpm the ECU then goes and
looks for
> Reset pulses which occur once per revolution.  If these are both within
given parameters
> the fuel pump relay is pulled in and the ignition coil driver stages are
powered up.
> Note: Cranking with a flat battery sometimes won't boot the ECU/fuel pump
as the starter
> signal is below 6V and the engine is turning too slowly!
>
> 3).    If the rpm signal is too slow the ECU assumes the engine has
stalled and shut down
> until the starter signal is again present.  This kills the spark and the
fuel pump relay.
>
> 4).    There a limited start stategy designed in that allows the vehicle
to be push
> started.  After seeing rpm AND reset signals for longer than 1s the ECU
will start
> computing the spark timing and then fire the coil.  Note: There does not
need to be a
> starter signal present for this.
>
> The most likely cause is the magnetic pickups.  These give endless
problems through poor
> connections and AR is not alone in this.   BMW & Porsche cars use the same
Motronic system
> and 99% of all non-start cases I have worked on the magnetic pickups
connectors are the
> culprit.  The actual transducers rarely give trouble but the 3 pin
connectos are a PITA
> when they become corroded.  (Recently I had to travel to Mozambique to try
and get an old
> Porsche 944 running.  On this car both magnetic sensors were dead, because
the swinging
> casting that sets the correct clearance between the sensors tips and the
flywheel had been
> swung too far and the flywheel nibbled the ends away so damaging the
sesnsors.
> Fortunately AR don't use such a stupid setup!)
>
> Also if you accidentally get the two connectors mixed up the ECU doesn't
understand the
> incorrect signals and will stay in shut-down mode with a resulting
no-start.  A very
> common problem especially if the engine has been pulled and the connectors
are incorrectly
> attached to the wrong sensor.  Unfortunately Bosch didn't use different
connectors for the
> two so it is easy to get them mixed up.
>
> The possibilty of putting the flywheel back in the wrong place will not
result in a
> no-spark condition, you will get spark but it will be in the wrong place.
As you have no
> spark it isn't that.  IIRC the Alfa flywheel can only go on in one
position.
>
> I hope this helps to get your Spider running.
>
> I have the official Bosch/AR service manual on the Motronic system,  like
most service
> manuals, it contains a lot of errors!  I have verified the system on the
bench (I have a
> test jig for trouble shooting the Motronic ECU's I built some years ago)
and the
> measurements I made on a running engine to confirm the system parameters.
>
> Incidentally you may not have noticed this but Motronic actually does
this.
>
> The VVT solenoid on the inlet cam is controlled by the Motronic ECU.  It
switched over
> at approx. 1500rpm.  If the engine speed is increasing it pulls in the
relay that powers
> the solenoid.  If the engine speed is falling at 1500 rpm it drops out the
relay.  But at
> the switching point it switches off the coil driver whilst the VVT relay
is operated.
> This I noticed as a sudden drop in rev counter reading and then a sharp
recovery.  The AR
> rev counter is very damped and so you don't normally see this effect.  On
my race car I
> had an after market rev counter with not much damping and I wondered what
caused the
> needle to twitch, even tho' I was not using the VVT cam as I had Euro cams
fitted to the
> 2L engine. The VVT relay was disconnected but the ECU still cuts the spark
whilst it
> switches over!
>
> If you need pinout details please shout, I have this for the '86 Alfetta
which uses the
> same Motronic ECU, but no Lambda sesnsor.
>
> best regards
>
> John
> Durban
> South Africa
> Alfetta 1.8L turbo
--
to be removed from alfa, see /bin/digest-subs.cgi
or email "unsubscribe alfa" to majordomo@domain.elided