3.6 hall sensor fault fixed

[ischmitz]

When I flashed out the stored fault codes I got a 1-1-3-4.

It was quite simple in the end: The three leads where ripped off clean inside the rubber boot of the connector. I should have checked that first especially since the fan belt minder connector was missing all together when I got the harness. Maybe the donor car has been hit in the rear and both connectors got challenged in the process.

After fixing it the code changed to 1-2-3-4 (intermittant fault) and after resetting the DME there are no more fault codes. Just a happy 1-5

What's funny is that the car ran fine without the hall sensor. No check engine light came on either. But it must have been in some sort of limb-mode (and I still smoked MikeZ cab on Cliff Drive the other day....)

I hope someone can explain what this hall sensor signal is needed for. I still don't understand.

Ingo

[Vipergrün]

Ingo, your car is sweet and sounds killer. I'm the guy who ran up and said *hi* in the Costco parking lot.......

Man, what a sleeper you have. Those 'tangs riced out Civics won't know what hit them

[ischmitz]

Brad,

we missed you today on the Solvang stomp. It was lots of fun in the back twisties. I had a hard time keeping up with Dan from TRE and his 3.0 914. Watch the board for some pictures from Dan or Dave tomorrow. I hope to see you for BBR III.

When I stepped on it today you could certainly hear the exhaust. But I like the agressive roar around 5000RPM a lot. Sort of reminds me of that nasty 3.8 TT that passed me during the Santa Barbara run. I'll never forget that sound.

ingo

[k9handler]

pics Ingo.....Pics!

Good to hear ya got her all worked out. What exhaust do you have?

[ischmitz]

I did not bother taking my camera out since there were at least four others snapping pics all the time. I had my hands full keeping the beast on the road

I still run Paul's old setup: 993 HE's into two independant high-flow glass packs. That's it.

ingo

[Joe Bob]

Originally posted by ischmitz

What's funny is that the car ran fine without the hall sensor. No check engine light came on either. But it must have been in some sort of limb-mode (and I still smoked MikeZ cab on Cliff Drive the other day....)

Ingo

Bullpucky.....the only smoking that you do is after sex with yerself....or when the car gets hot and the oil leaks on the heat exchangers.....

[k9handler]

Ha HA....I want video proof!
Mike and Ingo need to run the hills and video tape it...that would be cool!

[JR Indy]

I'm not an expert on the Bosch system, but I have designed a few engine control systems, so this is my read on it. Also, not sure of what the specifics of your car were.

The hall effect sensor is also known as the camshaft position sensor. So, as you can now guess, it is used to determine the camshaft position. Another sensor, the crankshaft position sensor, is a variable reluctance type that is used to determine the -- you guessed it -- crankshaft position. Without the camshaft position sensor the DME can not provide sequential fuel injection and will instead use the crankshaft position sensor to provide fuel in a pseudo port fuel injection mode. Basically, it doesn't provide the most optimal timing for fuel delivery.

For spark, it will provide what is sometimes called waste spark (assumming it is not coil at plug - I don't know the 3.6 system) which delivers spark to the two oposing cylinders at a time. Some systems do this all of the time, tradeoffs are not important here.

So, yes a system will run without the "hall effect" sensor, or camshaft position sensor but with less than optimim performance. I may be off a little depending on the specifics of the 3.6 system.

[ischmitz]

John,

the camshaft sensor (dizzy sensor) runs at half crank speed and allows the DME to know which cylinder is firing. For the Bosch M2.1 motronic that info is needed to individually retard the spark of the cylinder where knocking is detected.

The 964 has dual coils and dual distributors, so waste spark delivery is not necesarry.

Good call on the sequential fuel injection. I don't know if the Motronic 2.1 used in the 964 uses sequential of multipoint injection. The factory WS manual states that you can activate each injector independently when you use a hammer (Bosch system tester) However, I think I heard somewhere the 964 uses multipoint mode when the engine runs. Not sure, maybe someone else can chime in here.

What's funny is that my car behaves better on decel while the engine is cold. So maybe it does use sequential injection now that I fixed the hall sensor lead.

Maybe I'll just hook up a dual channel scope and check if any two injectors fire in sync or not.

ingo

[JR Indy]

I don't want to get too carried away on this, but knowing which cylinder is firing (Camshaft position) as apposed to cylinder pairs (Crankshaft position) offers other advantages for fuel and spark control.

Knock control as you mentioned is one, each cylinder typically has a "learned" spark retard value that is applied to the final spark timing for that cylinder. Each cylinder can then have slightly different spark timing based on it's own propensity to knock and can therfore be adjusted individually. The learned value is based on the previous knock indications and the DME algorithm attempting to decay the retard out until knock again occurrs on the given cylinder. I wasn't aware that the 964 used dual distributors. Are the coils one per cylinder? They still may be one coil per two cylinders, coil at plug and two coils does not necessarily mean the same thing. Knock control is possible with waste spark.

For fuel control, again each cylinder will have its owned "learned" fuel multiplier. This allows slight offsets (Usually a 0 to 2 multiplier) of the calculated fuel amount. These offsets are due to volumetric efficiency, injector tolerances, etc, that are constantly calculated based on the closed loop feedback of the O2 sensor. The learned values can be part of a 3D table based on engine load.

I'll have to dust off my old motronic book and fuel / spark control flow charts if it gets any deeper.

[ischmitz]

JR Indy,

the 964 (and the 993) has exactly two ignitors and two ignition coils total. Each coil goes to one distributor. From there it goes to its six spark plugs. Did I mention that we are talking twin plugged engines here with a total of 12 plugs. No coil on plug though. That did not come until the 996.

I would be very interested to know if the 964 uses sequential injection or just fires all injectors at the same time. I also saw somewhere "sequential multipoint" injection. WTH is that?

It seems there was some evolution between the 964 and the 993 motronics.

What books do you have that talk about the Motronic 2.1. I would like to read myself a little more into that topic.

[JR Indy]

What year is the 3.6 from?

From what I read last night about the L-jetronic system, the 964 and 993 are not sequential injection but instead port injection. I don't have the timeline of when the systems went from port to SFI. Anyone know where I can find this info? So, the injectors will be fired at the same time, twice per crank revolution. I initially thought these were sequential, sorry about that. As you can tell, I'm still very new to the 911s.

I know of sequential and multipoint but not sequential multipoint. Perhaps they were just referring to sequential.

I was not aware of the twin plugs, but it appears that the second distributor is driven directly from the first. I'm curious how the timing is controlled. Both plugs at the same time for each cylinder or each have it's own timing? I can't tell from what I've read.

[JR Indy]

The sequential multipoint may be something we used to called "Group" injection. This is when injectors are controlled in opposing cylinder groups instead of all at once. This is basically a poor mans SFI, and in application is much like waste spark. And as you were saying, with only a single coil there is no waste spark.

Again, sorry for the confusion. I need to learn more about these cars. I'll put a scope on mine, I haven't even done that yet. If anyone wants to see the plots, I can post them here.

[Lorenfb]

Yes, the Hall sensor in the distributor/cam is needed for
sequential fuel injection and for direct ignition systems (DIS).

The Hall is not needed when using an ignition distributor
for spark distribution, but is needed to run in the sequential
fuel injection mode versus bank injection. The sequential
fuel injection for the 964 provided mostly better emissions
control.

You're all assuming the early DMEs (964) used the knock
sensors to control the timing for each cylinder versus all
the cylinders. OBDII ('96) cars with misfire control does have
this capability but they use more than two knock sensors.
The knock sensors in the 964 were used mainly to optimize
performance (run at close to max. advance) and for better
emissions control.

The firmware in the 964 DME was a major advance over
the 3.2 Carrera Motronic, but present day automotive
electronics have advanced significantly since the 964 days.

Loren
'88 3.2
Systemsc.com

[ischmitz]

All,

My motronic is the one used from 92-94. It's the latest revision of 964 motronics. After that the 993 Motronic (OBD I, later OBD II for the U.S.) came about for the model change.

The 964 shop manual talks about the Motronic being able to retard individual cylinders. Why would they say that if it not used?

The timing in the 964 is solely controlled by the flywheel sensor/DME. It resolves the crank position and TDC with it's 60 (-2) teeth on the flywheel. The second distributor is simply belt-driven off the primary one.

[Lorenfb]

Your assumptions were correct. The 964 DME
does retard each cylinder in 3 deg. increments
to a max. of 9 degrees. The is per the Porsche
tech manual WKD 495 121 pg 2-08.

See, everyone learns when challenged!

Loren

[JR Indy]

So, the second set of plugs are fixed timing to the first set? Or is the timing control unique?

A system with cam postion sensor and knock sensor can control spark for each cylinder. Without a cam sensor then the knock sensor input would simply be applied to the spark timing control for all cylinders.

The cam sensor information can be used for better spark control - mainly performance gain. And, it can be used for better fuel control such as SFI - mainly emissions gain.

I don't have the pinout for the 86 DME. I can't find one in any previous post either. Can someone share this information please?

[Lorenfb]

W/O dist/cam all cylinders get retarded.

It is possible to optimize the ignition system for performance given a
cam sensor and good knock sensing using the DIS as with most late
models. A cam sensor and a dist. with knock is still not ideal.

86 DME:
Pin 1 spark Pin 14, 15 fuel injectors
2 idle sw 16, 17 grd
3 WOT 18 +12V
4 starter 35 + 12V
5 grd 34 ISV
6 grd 33 ISV
7 AFM 24 O2
8 speed sensor 25 ref sensor
9 5V ref 26 ref sensor
11 up shift 27 speed sensor
13 temp

These are the key ones.

Good Luck
Loren
Systemsc.com

[Lorenfb]

Forgot Pin 19 - grd & Pin 20 fuel pump relay