3.2DME Mystery Solved - Almost...
 

Thanks to a really helpful post from FR Wilke, I think we are close to solving this. First, FR is a 944 guy who really understands the ML3.1 Motronic. Check out his web site at http://944dme.tripod.com

It appears that there are three fuel/ignition maps in the DME chip - CA, 48 States, & ROW. The active map is determined by an external jumper plug in the DME wiring harness [!!]. Section 28 of the 944 workshop manual has the details of this jumper, so I'm hoping that the same section of the 911 3.2 Carrera shop manual has the details for our cars.

Now we need someone with access to the WS manual to look up the info and post the details - and the mystery will be solved forever!

So......do we have a volunteer, with workshop manual at the ready??


FWIW I ran the Autothority at Watkins Glen last weekend, and saw oil temps a good 20 to 30 dgrees higher than before, which I'm fairly sure was due in part to the more agressive ignition advance in that chip. No other changes since Mid Ohio, and ambient temps were close to the same. Will be moving back to the stock chip if we can get the info above confirmed.

Tony,
I will look when I get home. I know my 3.2 manuals say something about which ones have the jumper. I was at WG last weekend too for the Zone 1 48 hrs and CVR Monday and Tuesday. It would have been cool to meet.
-Chris
'86 Carrera

Thanks Chris.
The CVR event was before the 4th, right? You might have met Eric Stratton [beautiful red 964 RS] from Potomac? We arrived on the 4th, and ran through Sunday. My first WG run - those bloody blue bushes grow too close to the track! What a great track - definitely a fully commited high speed facility. We had a few non-contact spins, and one contact damaged car. There is very little room for error at the Glen, right? Will be back there at every opportunity.


Thanks for you help with the info - looking forward to it.

Yeah, those blue bushes have left their mark on my car before.:rolleyes: (My codriver did that.) It was raining and he spun my car right after the off camber left and backed into the guard rail. The run before his, I spun in the same spot (2 rotations!) but I didn't hit anything. Often immitated, never duplicated...
I was there for June 29-July 2 so duh! I guess that makes it the weekend before last doesn't it? There was a 962 in my run group. I didn't feel bad when he passed me! I'm going back at the end of the month July 31,Aug 1 with NER. I think there might still be openings... :D
-Chris

Sorry about the quality.
-Chris
http://forums.pelicanparts.com/uploads/DMECoding.jpg

Finally.........This is IT!

Chris, I cant quite make out the pic - are they harness connectors in the engine bay or jumpers/connectors on the DME board?Maybe you could fax the copy to me at 703 288 3476??

Many thanks,

Sorry Tony, I'll try again tomorrow. The jumpers are in the harness under the seat. Connector 10 is near the front of the seat and 28 is near the rear.
-Chris

OK - understand that its the main DME connector that terminal 10 refers to. Should be easy from here.
Thanks Chris!

Okay Tony,

You're going to have to help me out here!

First of all, it seems that there are two DME units from the part numbers; one for ROW and Sweden, and one for USA and everything else. So first I have to determine which DME part I have - presumably it is the latter (.16 and .20).

Next, my car is a CA car. So the factory setup would have terminal 10 connected and Jumper 28 not installed. In order to convert to the standard US configuration, I would have to disconnect terminal 10. Is that what I am seeing?

So how does one do that? Is it on the plug or on the DME unit itself? Please explain.

I will try to get some better pictures tonight. I have a spare 3.2 harness that I can take a picture of. The jumpers are located under the seat on the harness not the DME. They come off as wires with connectors.
-Chris

THE ANSWER
 

The definitive solution to the 3.2 DME chip mystery has been found!!

3.2 Carreras delivered in CA run a lower performance fuel/ignition map than NA cars delivered in other states.

The DME stores all three maps [CA, 48 states, ROW] on the same stock chip. Map selection is achieved through external jumpers on the DME harness.

The selection of CA / 48 states map is made by a single quick disconnect jumper located near the main DME connector under the front of the front seat.

Pull the seat fully rearwards, and lift the carpet to expose the DME unit and harness. Attached to the main DME connector harness is a brown wire loop fitted with a quick disconnect - this is the Pin 10 jumper that makes the selection.
For CA map the loop should be connected.
For the higher performance 48 states map this loop should be disconnected.

MANY thanks to FR Wilke - the 944 DME guru who provided the key to this puzzle, and Chris Bennet - the famous 911 racer and master of Thunder Road, NY who provided the workshop manual research.

any ideas on what kind of performance gains will be noticed by switching from ca to 48? am i correct in assuming that this will only affect the timing and fuel/air ratio? maybe 5 - 10 HP?

this sounds like a easy, inexpensive mod...just my type.

APKhaos, thanks for doing all the research.

Yep - approx 10HP and a better torque curve for $0.00!
Can't beat that, and its a simple step to reinstate the CA map when visiting the smog nazis.

CA delivered 3.2 owners rejoice :D

What exactly do these different jumper configurations do? I've never seen three different map configurations in the eprom dump of a stock 84-86 911 DME. Though I'm not 100% sure of the 87 to 89 dumps, I have not spotted any alternate maps in their dumps either. Maps are usually easily spotted in the dump because because they are groups/arrays of similar sequential numbers.

I am with you Steve. I have not seen more than 1 map in any of the chips we were hacking. That included 24 pin/28 pin oem, Autothority and Weltmeister. There may have been multiple maps in a 944 chip though....

One thing I might add though, I've owned two '84 Carreras, one California car, and one 49 state car, both at the same time. The California car was definitely a bit slower with it's stock chip. When I pulled out the stock chips in both those boxes, they had different numbers stamped on top of those chips. I don't have my California car anymore, but the chip on my non-Cali car reads 1267355102. When I chipped those cars back then with the same aftermarket performance chip, they ended up having the same identical performance, ruling out the possibility the California car had some kind of mechanical problem.

If someone has the stock chip out of their California car that they can send to me on loan, I'll map the the fuel and ignition tables, compare them to the 49 state maps, post the results here on this board, and end this mystery once and for all.

Steve/Thomas,

Firstly, lets get a baseline:
1. The same stock chip is used in CA, 48 states, and ROW cars
2. The workshop manual show how external jumpers are used to select which configuration is active for a given car - see Chris' post above.
3. The same is true for the 944s of this vintage, which use the same DME controllers.

So, what do these jumpers control? I agree with you that there may not be discreet maps for each configuration. Its more likely that the jumpers define parameters used in the DME processor which effectively modify the way the base map is applied to each of the three configurations.

It would be interesting to find some definitive info on this aspect of the DME operation, particularly when you are hacking the map.

However, I think there is sufficient evidence and documentation to be certain that the jumpers actually modify the map as stated above.

Interesting. I wonder what changing the chip to an aftermarket really does then?

Hey Kurt,

Do you still have that G-tech? Interested in trying out a couple of performance chips and letting us know how they compare to your Weltmeister?

There is actually only 1 input to the DME that is jumper controlled and that is pin #10. Designated on the DME pinout as 'Bridge to GND, California'. On the 944 pinout pin #10 again is the only input that can be jumpered and it is labeled 'Perf. data change for CH,S,AUS,NZL,Cal.,Japan'.
My guess would be that grounding pin #10 controls timing in some way.
Let us know of any other discoveries, and I will do the same.
Regards,