Strange issue - odometer stuck at 999,999mi

Hoping to get some help on this from folks smarter than me. Some background.... This is on my 2004 40AE car. No prior issues related to the gauge cluster of any kind. I have been experiencing an intermittent issue of the HVAC control unit not powering on. I decided to plug in my Durametric to see if any codes pop up. I also went ahead and checked all modules for any codes while I was at it. While going through the typical motions, the Durametric app did freeze for a moment and then kept on working. At one point, I looked up and saw 999,999 on my odometer and immediately thought wtf? I had the app hang up like that in the past but never had any issues with the car. The battery is just a few months old with a relatively new alternator and the car did have 12v for the relatively short time when I was scanning for codes. I do have the basic Durametric so I was not monkeying around with any coding. What have I done so far? - drove a bit to see if I end up with a million mile 996 - no changes but the trip odometer works fine. - reached out to our resident expert, Skip. Skip recommended to replace the ignition switch as that powers on a whole bunch of modules and when it starts to fail it can cause electrical gremlins. It's a cheap VAG part used in many other cars so I gave that a go. After some in-car yoga to get it replaced still no luck. Prior to the replacement, that car started fine and key went in and out of ignition without any fuss. - I did a bunch of searching on line and found a thread with someone having a similar issue on their 1999 Boxster. They were able to change the units from kilometers to miles, which would change the 999,999km to 599,999mi. I tried to toggle this but it showed 999,999 no matter what. - I found the following to have worked for someone with a 1999 Boxster - Disconnect the negative cable with the key in the ignition. Turn key to position 1, let the car sit long enough to bleed off any power, then turn key to off position. Then hook up negative cable. I tried this with no luck. Not sure if I did something wrong but seemed like simple directions or maybe the procedure is different for later cars? - I checked all fuses in the footwell. All fuses were fine. I'm pretty puzzled at what happened. The gauge cluster continues to work fine with the trip odometer is recording distance as usual but the main odometer continues to be stuck at 999,999. Is there another fuse somewhere else? Can this be an ECU issue? I can't see how as otherwise the car continues to run perfectly fine. Hope someone can shed some light on how to resolve this. I have put together the DIY PIWIS from the recent thread, however, I have not used it much except for connecting it to the car to see if it works so my experience there is pretty non-existent. I know it's a powerful tool but that also can cause issues if you don't know what you're doing so I'm treading carefully.

Hi Pete, it sounds like the data in the instrument cluster has gotten corrupted......It can happen sometimes when you have strange voltage/power issues... Since you have tried re-setting by removing power, I would use the PIWIS to "Write " the basic data from the PIWIS Tester into the cluster, then do manual codeing... Note : do not " Read" the old data from the cluster, it is corrupted..just "write" in new data from the PIWIS Tester ... Note 2 :: it may or may not show the original actual mileage ... There is a program in the cluster that blocks it from being able to change the mileage except on "NEW" clusters with less than 100 miles on them......Some instrument cluster hackers have found a way around this feature I think, but I have no experience with that... Just fire up the PIWIS and get to your 996 menue, then highlight instrument cluster , then "maintenance/Repairs,, then "Write data" ..... It can't hurt it and may fix it.... you may have to go into codeing and set the Country, OBC, Spoiler , Vin Number, ect.. back like it was, so write that down first before "writting the data record to cluster"... Hope this helps !!

Thank you, Skip! The other day I also tried another potential fix....someone was suggesting, with the battery removed, to touch the positive and negative cables and hold them together for several minutes. This should "reset" all modules and cause them to reboot once everything is reconnected. Unfortunately, no dice. I'm away for Easter but when I'm back on Monday, I will give the PIWIS a try and cross my fingers! Will report back.

Crazy. Good luck with sorting it out Pete.

Quotes source post: 20001811

Is this why Jake says NOT to use a battery charger/tender when using a Durametric?

Quotes source post: 20002843

Its Very likely that this is a Durametric bug because with PIWIS Porsche is very adamant that you MUST keep the voltage from going low with a high quality 40 amp charger that is putting out "clean" voltage....

Quotes source post: 20002519

Electrical engineer here. That's an old wive's tale.

Quotes source post: 20002938

ooooh really? Can you elaborate? I�ve been under the assumption that it works.

Quotes source post: 20002938

I don't know about that ... These super capacitors can hold a charge for hours/days/weeks up to months....I have witnessed this first hand several times when the memory data was corrupted and couldn't be cleared/erased by just disconnecting the battery/power supply...... The only thing that cleared the Memory was to touch the Positive to the Negative to drain the capacitors....Known as a "Hard Reset " ...It's rarely needed , but I still know what I have experienced.... I'm willing to listen to any Technical Reason why this is not plausible...., I remember one in particular was a 1999 996 3.4 with code 1341 ..and actual values were showing cam was ~200* out of time but engine was running fine...I double checked cam timing, replaced cam sensor, replaced cam actuator , cleared fault codes ect...but it would come right back ... Note : on the 3.4 the bank 1 and bank 2 cam signals are sent on the same wire and the DME decides which is which except on ROW cars where they only use one cam sensor, the DME doesn't need to figure anything out... I had done everything possible, erased memory, disconnected battery for an hour, checked and double checked cam timing, replaced the whole system..ect....Then out of desperation I did the "Hard Reset" because I figured, why not, I'm, out of idea's,and then voila !! All fixed !! It had to be that the corrupt data in the NVRAM or ROM as staying alive ( from capacitor charged) no matter what I did to kill the power and erase the data ....

I mainly tried the "hard reset" as I was running out of ideas. I'm still going to try the PIWIS recommendation from Skip but now I'm wondering if I should have held the two battery wires for longer? To Skip's point, maybe I should hold them together with a metal object to ensure good connection and leave it be for at least a day (instead of minutes) to allow the super capacitors to discharge...

Quotes source post: 20002896

If I recall correctly their reasoning for not using a charger or maintainer was that they can introduce voltage fluctuations, so it sounds like the same thing. No charger unless it's a really good one.

When I got home this morning, I plugged in my PIWIS 3 set up to see if I can follow Skips steps to try and WRITE and data to the instrument cluster. First I went through the steps in the Simulation Mode to make sure I know where to go in the UI. Once I got PIWIS to connect I went to the Instrument Cluster module and then Maintenance Repairs where the Read / Write functions were located under Control Unit Replacement. Unfortunately, after several tries I was not able to get the communication to establish and do any Writing. The below error came up every time. I'm not sure if this is an issue with the cluster or with PIWIS and/or user error, which I wouldn't discount with me being a noob with this. I tried going to other sections of the Instrument Cluster tabs and got the following: I tried doing this with the engine running ( to make sure there was ample power ) as well with engine off. The voltmeter was reading slightly over 12V the whole time and I did this after driving the car for a while so I doubt this was a power issue. Any thoughts @Porschetech3 ? Seems odd that the gauge cluster seems to function perfectly fine, except for the odometer issue and I'm still getting these connection issues. Is there some sort of a fuse in the gauge cluster that might be blown or is there a different way to reset things?

Think about it. They intentionally make it nearly impossible to change the mileage. Why? Lawsuits! A manufacturer that made mileage changes easy would likely get sued...a lot. I think the only place that may help you is Palo Alto Speeddometer. http://www.paspeedo.com/

Quotes source post: 20008748

I did think about it and of course I'm not trying to change my mileage - I want to restore what was there prior to whatever caused the 999,999 to display. Their website has been down for a while too so I will have to give them a call if I don't end up finding a solution.

Quotes source post: 20003376

I understand what you are saying, but all automotive control systems have their own power management systems and that alone would isolate any capacitors on the circuit board from direct connection to the battery. Beside keeping the chips isolated and controlling when and how they power up and power down, microprocessors were developed to run on 5V. They were optimized and standardized decades before they found use in the automotive industry. I'm not saying its impossible to have some impact by grounding the positive battery cable, but it isn't likely due to the design.

Hi Pete , Make sure you can communicate with instrument cluster via PIWIS Go to main menu , then highlight instrument cluster , then select "drive links" ... In Drive links you should be able to raise spoiler up and down, turn ON/OFF any warning lights, and also make the Tacho read any RPM you select in the bottom slide bar... If you can perform these functions it means you are commumicating with the instrument cluster via PIWIS . PIWIS cannot Write in the mileage EXCEPT on NEW instrument clusters with less than 100 miles on them......But it can read/write in the programs ... You can also try to change Country to ROW or something else..... \You can also "Read" from another car just like yours, then "Write" to your instrument cluster from another car.....( I have done this on specific issues) Just try many things since your cluster already appears to be corrupted ....If PIWIS can't save it then you are looking at replacing it or maybe a specialist instrument repair facility can repair it...

Quotes source post: 20008986

I know that NVRAM and ROM will stay alive even without battery power, and so will other types of RAM if they are non-volatile .. I admit I don';t know all the means that can keep them alive that retains the data ( and corrupt data) whether it is capacitor charge , magnetism, electro-magnetism, or permanent magnetism.... I just know that in rare cases when nothing else can erase corrupted data that is disrupting the intended proper function of the data program that grounding the positive battery terminal will erase the corrupted data memory and allow the "learned data" to re-learn correctly...