maxed out at 621358 miles, 1 gigameter limit?

[jetta 97]

How can I reprogram the speedo to read in furlongs per fortnight?

What do you mean?Speedo can be converted form miles to KM and that's all.

 

[jcrews]

NOPE, two bites can record 999999km.
65535 is just max that you can put in adaptation of cluster with VCDS.
Read post #39 again.
On Eeprom address line 00F0 two bites storing millage not KM.

Let say two bites in dump are 460D. When you swap the two bytes, so 460D becomes 0D46.Then subtract that value from FFFF.
So FFFF - 0D46 = 2FB9. Now convert 2FB9 to decimal which is 62137, now multiply that number by 10 and we have 621370 which you can multiply 1.609 344(1 mile =1.609 344 KM)

621370 miles x 1.609344=999 998 KM

My evidence is contrary to that (along with common sense). Racertodd had 411048km (41104 in NVRAM), so when the battery was reconnected, the value was re-read into the counter program, resulting in 411040km being displayed to the user as 255408 miles. Therefore, the explanation you are citing is simply incorrect. Not accepting that fact is venturing into the domain of pseudoscience.

If I had an EEPROM flasher, I'd just grab the application from the cluster and disassemble it. Then I'd be able to tell you exactly what each NVRAM storage area is for.

 

[jetta 97]

My evidence is contrary to that (along with common sense). Racertodd had 411048km (41104 in NVRAM), so when the battery was reconnected, the value was re-read into the counter program, resulting in 411040km being displayed to the user as 255408 miles. Therefore, the explanation you are citing is simply incorrect. Not accepting that fact is venturing into the domain of pseudoscience.

If I had an EEPROM flasher, I'd just grab the application from the cluster and disassemble it. Then I'd be able to tell you exactly what each NVRAM storage area is for.

Above info what you just said has nothing to do with how much cluster can go.
The thing that cluster goes form let say 104566 miles to 104560 after disconnecting battery is because the store value is in cluster is only 5 digits multiply by 10.
Any time you adopting cluster with VCDS ,it will never show last number for old cluster. It will be every time 0.And I just explain why is this like that.Example if you have 135987 on your cluster, after putting new cluster and adopting millage in new on it will show 135980.
Like I said read post #39 again.
Stave explain everything very clearly.

 

[jcrews]

Ugh, I just said that. Please read carefully.

 

[jetta 97]

NOPE, two bites can record 999999km.
65535 is just max that you can put in adaptation of cluster with VCDS.
Read post #39 again.
On Eeprom address line 00F0 two bites storing millage not KM.

Let say two bites in dump are 460D. When you swap the two bytes, so 460D becomes 0D46.Then subtract that value from FFFF.
So FFFF - 0D46 = 2FB9. Now convert 2FB9 to decimal which is 62137, now multiply that number by 10 and we have 621370 which you can multiply 1.609 344(1 mile =1.609 344 KM)

621370 miles x 1.609344=999 998 KM

I was explaining to you how much cluster can go , not why is resetting last number to 0.You the one who said in post above that mu evidence in not correct ,by saying that cluster reset last number to 0 after battery being disconnect.What that has to do with how much cluster can go.
AND I KNOW CLUSTER can GO 999 999KM. I just saw it at Friday on Brain's car.
You need to read carefully.

 

[jcrews]

Yes, I understand what you're saying. However, it doesn't validate the hypothesis you cited, so some information is missing. Just because my contrary evidence doesn't explain how the cluster can display > 65535*10km, doesn't allow you to exclude it.

I might run an experiment with a spare cluster to see what the behavior is in the corner case of overflow. I will also see if I can get access to the flash machine required to extract the entire cluster memory, which isn't possible over the diagnostic interface.

 

[jetta 97]

Like I said 65535 is the value that is maximum that you can put with VCDS . Witch does not mean it is maximum value for cluster.
VCDS make Ross-Tech, and cluster on Jetta in made by VDO.

 

[DaveLinger]

What do you mean?Speedo can be converted form miles to KM and that's all.

Whoosh

http://en.wikipedia.org/wiki/FFF_system "a set of units that uses impractical measurements not seriously deemed suitable for scientific or engineering use"

"a humorous system of units and is not used in practice"

 

[jetta 97]

Whoosh

http://en.wikipedia.org/wiki/FFF_system "a set of units that uses impractical measurements not seriously deemed suitable for scientific or engineering use"

"a humorous system of units and is not used in practice"

I think it is not possible.

 

[1854sailor]

Rocket Science?

I think it is not possible.

Do the arithmetic. 1 mph = 2,688 furlongs/fortnight. You figure out the code. Just pointing out how ridiculous this thread has become.

 

[Steve99]

Yes, I understand what you're saying. However, it doesn't validate the hypothesis you cited, so some information is missing. Just because my contrary evidence doesn't explain how the cluster can display > 65535*10km, doesn't allow you to exclude it.

I might run an experiment with a spare cluster to see what the behavior is in the corner case of overflow. I will also see if I can get access to the flash machine required to extract the entire cluster memory, which isn't possible over the diagnostic interface.

What other memory are you hoping to read from a regular (non MFA/FIS) cluster?
You're going to have to find a non VDO cluster if you want it to display higher mileage.

 

[alphaseinor]

As I said earlier, the highest adaptable limit is a short int (16 bit). There's gotta be something that switches somewhere to allow it to reset the eight bits to 0, and in that logic somewhere there is a 999,999km limit, without a rollover to 0

There is no logic to the possibility to it just being a sixteen bit int
65,535km^10 = 1111111111111111 <-- the highest adaptable short int, 16 bits
99,999km^10 = 11000011010011111 <-- The closest km available to store internally

so, something at 65536 km ^10 has reset the short int to 0000000000000000. I'll bet there's something written to NVRAM that we're just not seeing (or looking for) that allows it to go past that number, and start counting again; without being a long int.

Inside the ECU is an unsigned long (32 bit) so the highest possible km tracked by the ECU is 4,294,967,295... don't think it'll be on the original rings...

Can you post the read you did from my cluster Jetta ,97?

 

[jetta 97]

As I said earlier, the highest adaptable limit is a short int (16 bit). There's gotta be something that switches somewhere to allow it to reset the eight bits to 0, and in that logic somewhere there is a 999,999km limit, without a rollover to 0

There is no logic to the possibility to it just being a sixteen bit int
65,535km^10 = 1111111111111111 <-- the highest adaptable short int, 16 bits
99,999km^10 = 11000011010011111 <-- The closest km available to store internally

so, something at 65536 km ^10 has reset the short int to 0000000000000000. I'll bet there's something written to NVRAM that we're just not seeing (or looking for) that allows it to go past that number, and start counting again; without being a long int.

Inside the ECU is an unsigned long (32 bit) so the highest possible km tracked by the ECU is 4,294,967,295... don't think it'll be on the original rings...

Can you post the read you did from my cluster Jetta ,97?

I will e-mail you dump.

 

[jetta 97]

Brian did you receive e-mail form me?
Also when you can post that picture from ECU with millage?

 

[turbocharged798]

I just realized that the title of the thread is not technically correct. I should be 1 megameter and not 1 gigmeter. A gigmeter is not a valid unit of measure.

1 megameter = 1,000,000 meters.

 

[alphaseinor]

I'll see if I can post it tomorrow sometime.

I got the email, but unfortunately I've been busy working on other things...

 

[jetta 97]

Any free time for picture?

 

[alphaseinor]

Posted photo...

 

[manual_tranny]

alphaseinor, your VAG-COM owners list link is down- also I would like to be added to this list! (For anyone reading, I am not trying to get on the "trusted mechanic" list, I need more time and experience.)

 

[NarfBLAST]

I just realized that the title of the thread is not technically correct. I should be 1 megameter and not 1 gigmeter. A gigmeter is not a valid unit of measure.

1 megameter = 1,000,000 meters.

I do a megameter every week (1000km).

1 gigameter (Gm) is perfectly valid = 1,000,000km

Great thread! I think if my car ever reached 999,999 km I would be happy that it did not roll over. And it is re-assuring that that the ECU will keep counting. This is cool!

Currently I am at 0.434 Gm and counting.

Any updates?

 

[philngrayce]

The computer odometer discussion is way over my head, but I am impressed by the mileage. What is the car up to now and what sort of repairs had it seen along the way? IPs, starters, alternators, camshafts, heads, etc?

 

[smid87]

Has anyone considered the difference that a one inch difference in wheel diameter will make? For example I have an 04 Passat TDI and I have 16" rims for summer and 15" rims for winter (don't ask me, it came with them). So it stands to reason that my mileage will be different if it is based on speed as it will be calculated differently for each wheel size. Correct me if I am wrong.

This winter my girlfriend and I went to SC from NH and we took both cars (7 people) and I noticed during one fill up that my odometer (04 passat with 15" rims and winter tires) was about 11 miles less then her (11 jetta with 16" rims and all season tires). The tire sizes are what the manufacturer recommends.

 

[tadawson]

The rim size really doesn't matter - what's the difference in the diameter of the tires?

- Tim

 

[VeeDubTDI]

Bump for fun! Just in case any of you were wondering how your vehicles keep track of mileage.

 

[alphaseinor]

Yeah, it got t-boned at 780,000 miles... we may never know how many licks it takes to get to the center of a tootsie pop

 

[VeeDubTDI]

D'oh! That's a lot of miles.

 

[TDI smile]

I love those numbers... I have to live in my car for the next few years to even come close. I think I watch every clip on the u tube from Irv Gordon and his VOLVO P1800

 

[NarfBLAST]

Yeah, it got t-boned at 780,000 miles... we may never know how many licks it takes to get to the center of a tootsie pop

Sorry to hear this! I hope everyone was okay.

 

[tonystdi]

So close, hope you can keep it going.

 

[gmenounos]

OK, this isn't going to be the prettiest chunk of data to look at but it will do.
This is part of a typical MKIV VDO cluster eeprom dump
Address in hex, hex data, ascii data
You can see the vehicle VIN and Immobilizer ID(both changed from real in this example)

In green is the immobilizer SKC/PIN, its value is stored three times, take the first pair of bytes and swap them, so 2D0E becomes 0E2D, then convert to decimal - 03629 is the SKC in this dump.

In red is the odometer mileage record. It is also two bytes in length but is stored eight times. The mileage record in this dump is 60D2. To get the mileage that would be displayed on the odometer we first swap the two bytes, so 60D2 becomes D260. We then subtract that value from FFFF.
So FFFF - D260 = 2D9F. Now convert 2D9F to decimal which is 11679, now multiply that number by 10 and we have 116790 which is what the odometer would show in miles. The last digit of the odometer is not stored in the same way and is represented by a different byte, that value I do not know. When you program mileage into a new cluster you don't enter it and generally enter 10% of the mileage you really want. If the odometer is configured to metric it simply multiplies the result by roughly 1.6.
So the stored value in a odometer with 0 miles would be FFFF and it decrements from there.
Its fun having a spare cluster to see how things all work...

00A0 | FFFF56575A375A304134343333313333 |..VWZ7Z0A4433133
00B0 | 56575A375A3041343433333133335657 |VWZ7Z0A4433133VW
00C0 | 5A375A3041343433333333322D0E2D0E |Z7Z0A4433332-.-.
00D0 | 2D0E394257444536314A323234343438 |-.9BWDE61J224448
00E0 | 303538FF585858585858585858585858 |058.XXXXXXXXXXXX
00F0 | 5858585858BCDC90FC8931D860D260D2 |XXXXX.....1.`.`.
0100 | 60D260D260D260D260D260D200000000 |`.`.`.`.`.`.....
0110 | 000000001000000000000000314A3039 |............1J09
0120 | 32303930364A20204134563037FF0C00 |20906J A4V07...
0130 | 0000364E3039303939303120192A1101 |..6N0909901 .*..

Sorry to bump such an old thread but the above info is not quite correct. The full info is here: http://nefariousmotorsports.com/forum/index.php?topic=3504.0 (3rd post is the most relevant)
Even though that thread is discussing MK3 clusters, MK4 work the same way.