Re set car computer?

[Nick Triesch]

By unhooking the negative battery cable, does it reset the computer and motronic system in a 1985 911? I know it works on new systems but I know 1984 was the first year for a computer. Thanks.

[Techno Duck]

It wont reset anything in the DME.

[steely]

I agree with TD, and I have never noticed or experienced any problems after reconnecting my battery cable.

However, I was always intrigued by something I read in the Bentley manual (a great book but with some minor boo-boos here and there), where it states (under the Engine Management section WRT the DME) that the DME "features advanced adaptive control circuitry" that updates performance info into memory. (Bentley 200-2).

I also thought I had read once that this info resets and gets updated sometime after re-connection and driving (how long?), but 1) I can't find it, 2) my own headspace memory maybe faulty, and 3) I don't know why this would be any different from turning the ignition key OFF.

I've disconnected my Batt' may times for safety when working on the car, and have never experienced any noticeable or odd effects.

Sorry for the long no-value-added reply, I just thought I'd mention this curiosity.

[uwanna]

The 964/993 DMEs do some amount of adaption after battery power is removed and restored according to maintenence manual.

[911obgyn]

No adaptive memory on earlier motronic brains, I think that came in with OBD electronics.

[Bill Verburg]

Originally Posted by uwanna

The 964/993 DMEs do some amount of adaption after battery power is removed and restored according to maintenence manual.

Gen1 Motronic used on '84-89 911 has nothing to set or reset when the + or - terminals are disconnented

964 has a gen 2 version w/ some adaptive features, mostly related to idle speed

'94/95 993 has motronic 2.10.1 gen 3(OBD1) where other adaptive features and volatile fault memory have been added.

US '96 993 has Motronic 5.2(OBD2) and is even more adaptive w/ even more memory

You can do adaption on 964 or OBD1 993 by starting and letting the engine idle for 10min

[Quicksilver]

Originally Posted by 911obgyn

No adaptive memory on earlier motronic brains, I think that came in with OBD electronics.

Bingo!
The 3.2 DMEs simply read sensors and look up those values in the tables stored in "the chip" and do all the ignition and fuel control based on those tables. As mentioned before there is no learned adaptation. It simply looks it up and spits out the result.

[proporsche]

Nick..the only thing you can help your self in the future by disconnecting the battery is.When your handbrake warning light stays on .resetting help to reset the warning light so it goes off when you release the hand brake....or if it stays on due to low brake fluid and does not go off after refilling the brake fluid reservoir.

One thing i did to my DME comp is, i made a short cut to the lambda circuit with external switch....so i can shut it off and on.

Ivan

[ischmitz]

Actually, both the 964 and the early 993 OBD I DME have volatile fault memory. OBD-II mandated non-volatile memory and freeze-frames for faults.

As others said the 3.2 doesn't retain anything in its volatile memory

Originally Posted by Bill Verburg

Gen1 Motronic used on '84-89 911 has nothing to set or reset when the + or - terminals are disconnented

964 has a gen 2 version w/ some adaptive features, mostly related to idle speed

'94/95 993 has motronic 2.10.1 gen 3(OBD1) where other adaptive features and volatile fault memory have been added.

US '96 993 has Motronic 5.2(OBD2) and is even more adaptive w/ even more memory

You can do adaption on 964 or OBD1 993 by starting and letting the engine idle for 10min

[steely]

Many thanks for the responses - sorry for the diversion Nick.
So the Bentley is wrong here? - No wonder I couldn't find anything to confirm this.

[Duc Hunter]

I wil say that I have had massive starting problems on my 964 powered Rat Rod. In great part in turned out to be no constant power to the DME. As soon as I got that straight, and let the DME learn the car it starts like a champ. Especially when cold (i.e. winter here in Florida, temps in the 40's) the car WOULD NOT START.

After battery has been pulled, or the DME unplugged you must start the car, and once started let it idle for 10 minutes. After this the DME learns the car enough that it starts great. Pull the DME, and it starts like crap until you do this procedure.

As Bill noted above....my 89 Carrera 3.2 has no such issue when you unplug the DME.

[ischmitz]

No the Bentley is not wrong here. Even the 3.2 DME has volatile memory and optimizes things during run time. It doesn’t retain it through a re-start though.

[steely]

Thanks Ingo

This is the other paragraph that got me thinking about this in the first place (from 240-4 of the 84-89 Bentley)
This one has to be incorrect.

[steely]

bump

[ischmitz]

Yes that paragraph is complete nonsense related to the 3.2 DME. The 3.2 doesn't learn anything in the first 10 minutes. It only changes mixture as the engine warms up (given the CHT works properly)

The paragraph applies to 964 and 993 (OBD-I) where the DME learns some VE-related coefficients and idle values.

[steely]

Many thanks Ingo

[Nick Triesch]

The reason for the post is that I wanted to know if the DME ever got stuck in a too rich or too lean mixture and if taking the battery off its post would reset the computer. But I know now that I just do not have a clean connection from the green wire to the 02 sensor. That will make the car run very rich.

[steely]

Originally Posted by Nick Triesch

..... But I know now that I just do not have a clean connection from the green wire to the 02 sensor. That will make the car run very rich.

Nick, I believe you need to wait the 10 minutes or so after start b/c the O2 is ignored until the engine reaches temp according to the CHT before you evaluate the O2/lamba. Disregard me if this was already clear.

[Bill Verburg]

Originally Posted by ischmitz

Actually, both the 964 and the early 993 OBD I DME have volatile fault memory. OBD-II mandated non-volatile memory and freeze-frames for faults.

As others said the 3.2 doesn't retain anything in its volatile memory

I believe that's what I said, M2.10 just expands on that over the 964 version, both have permanent + and - pin assignments to accommodate this.

[Bill Verburg]

Per Bosch that is correct

adaptive closed loop A/F control was introduced w/ the 1989 964 Motronic

In this system deviations from setpoint value when encountered cause the control range of the Lambda control system will adapt by moving the map in the control unit. Due to the fact hat this relocated map must be stored, the engine must be run for ~10 min. if the power supply is interrupted, causing the stored data to be lost., Positive adaption is only possible w/ the engine running.


Prior to the '89 964 this was not a feature on Motronic used by Porsche 911