Looking for 2nd gen (2010) Tundra 93C66 Odometer Programming Info

[Zero Cool]

Hello,

I am looking for information on programming Toyota Tundra Odometers that use the 93C66 chip. I am looking to understand just how the programming is written. I have done a LOT of experimentation and have learned quote a bit but i am not quite there yet. many pieces of the puzzle i am still missing. I have learned that The clusters write in a 17 bit wear leveling scheme. I have been able to write 0-8 and in bites of 17, so i can program 17 miles, 170, 1700, 17000 etc and multiples of that. 34, 340, 3400 etc but have not yet figure out how to write 9-16 to the cluster yet. again this is all been hacking so far so i am still learning.

So for my purposes i can rebuild clusters and reset them to zero miles. that was my main goal. but it would be nice to fully understand how these work so that way i could offer an exchange service and be able to program them to match a customers core unit. (all that sounds much more business like then it really is. I really hate the orange look of my cluster and i changed the LEDs to white and people have asked me if i could do the same for them.)

I like puzzles and hacking and I am having a blast digging into this and learning. I see that Mr. Clusters is the resident expert. I saw his posts on another forum as well and hope to have a conversation with him. per his request on his many post's. i am posting here and not sending him a PM. yes sir i can read! and will be happy to give thanks, thumbs up etc.

Some posts here mentioned flashing. I have discovered that if something is not written correctly, or of the eprom contents don't match for some reason, the odometer will flash. for example. 2011-2013 clusters are different then 2008-2010 clusters. In 2011 they added the flex fuel capability to the trucks and i think that may be why the change....there may be more differences i am not sure but...If i take the contents of a 2011 cluster and program it into a 2008 cluster, i get the flashing error...However. If i take the 2008 eprom and hand change the values to the same values in the 2011 cluster. it works and no flashing.

I have acquired a Ch341A and did the 5v mod. Thank you Dr. Shock. I have the ASprogrammer software as well as Tachosoft(really buggy) and Neoprogrammer as well.

I would like to collect some bin files. would be happy to make some donations if i could get bin files for something like 1 mile, 10miles, 100miles etc maybe i can learn what changes with each. and would be Happy to report my findings so all can learn!

Happy to Be Here! let the games begin!


ZC

[clusters]

Read this: https://www.digital-kaos.co.uk/forum...=1#post3903656

[Zero Cool]

Thank you. I am still reading threads on this forum and had not found that one yet. the Yazaki 17 text is a better explanation of a post i found on another forum. and I did learn how to read the mileage in an eprom. by reading the hex value reversed, multiplying by the number of instances etc. but I am still working on the correct way to write.

for an example. if i write 00 00 in the first 17 blocks. and in the next 8 blocks i get 0 on the display. adding a FF FF to block 18 and I get 1 on the display, FF FF FF FF gets me 2 on the display etc up to 8.

or, if i write 01 00 to the first 17 blocks. I get 17 on the display. adding a FF FF i get 18 etc.

So i am still learning. I will experiment some more this evening with the added info you have provided and see where i get to

[Zero Cool]

ok i am trying to write 50000 to the cluster. and the only way i can figure out how to do it is to write 7D 0B to all 17 place holders, which gets me 49997 then add 3 sets of FF FF and then i get 50000 even. but from everything i have read, this does not seem to be the way to do it correctly.

If i write 3124 to 16 places = 49984 and then write 16 to one place (in hex 34 0C and 00 10) with one set of FF FF i get 51308 which does not make sense.

I have also figured out WHERE you put the FF FF makes a difference.

I am very confused. attached should be a copy of my 50000 dump.


Zc

[clusters]

Read the post I linked to again!

If the complement Words are wrong it will freeze the odometer.

The Mileage is all in the first 17 Words.

[bob191919]

Thanks @clusters

@zerocool, i attached some examples:
https://www.calculator.net/hex-calculator.html


file 001
calculate hex to decimal: 2B1C= 11036
Odd= 17 times
11036 x 17 = 187612


file 000
calculate hex to decimal: 4562= 17762
Odd = 16 times, 1 Even
17762 x 16 = 284192
17761 x 1 = 17761
284192+17761=301953


file 002
calculate hex to decimal: 3B78= 15224
Odd= 17 times
15224 x 17 = 258808


your 50000(49996):
16 x 2941 (0B7D) = 47056
1 x 2940 (0B7C)= 2940
47056 - 2940= 49996

[Zero Cool]

I FINALLY understand how this works!

First off, I want to say thank you to everyone that has posted and helped and all the other forum posts on the subject. All have been helpful and especially CLUSTERS man you are out there. I found your name on some of the most OBSCURE websites! LOL you have quite the web presence! Thank you to all!

So while I was able to read and program my Tundra cluster and get it close to whatever mileage I wanted. I didn’t exactly understand how it worked. I googled and googled and was trying to learn about the Yazaki-17 wear leveling algorithm. And found lots of examples using software, or expensive programmers etc etc but nothing that EXPLAINED to me exactly how it worked. Then I found a youtube video in Portuguese (I had to translate it one word at a time using google translate LOL) that sort of gave me the key. In this video he programmed 10,000 KM/Miles and I was able to freeze the screen and copy it into my programmer and duplicate it on my cluster. But it was the WAY that programming was laid out that it finally clicked in my head and I finally understood what was going on and how it worked. Now I get it.

AS It has been explained. There are 34 blocks in total. The first 17 blocks are the mileage and the next 17 blocks are the check sum.

Let’s say we want to program 80,000 Miles/KM. take 80,000 divide by 17 and we get 4705.88235xxx Ignore the change and just take the 4705 and convert to HEX and we get 1261. Now we have to reverse the order so we get 61 12. Program all 17 of the first blocks with 61 12 and the next 17 blocks with all 00 00 and if we program the cluster at that point, we will get 79985 on the display. Which is 4705 times 17 = 79985

If we want to get to 80,000 exactly, we now have to count forward. 80,000 minus 79985 = 15 so, if we advance the count by 15 by changing the first 15 blocks up one digit to 62 12 (remember it is reversed so it’s actually 1262 hex is one more than 1261, reverse it and its 62 12)
and change the first 15 checksum blocks to FF FF we will then get exactly 80,000 on the display.

as it works on blocks of 17 and then rolls around and starts over. If we wanted to keep advancing. One block at a time. Making it 16 blocks of 62 12 and 16 blocks of FF FF we would display 80,001, 17 blocks would be 80,002 AND….if we keep going. And now make the first block 63 12 and the first block of the checksum 00 00 we would then have 80,003!!! And we keep advancing and rolling around in a circular pattern and we’ve got it!

So we can program any mileage that way exactly.

lets say we want 236,546 a number chosen at random. We first divide by 17 and we get 13914 ignoring the change. We then convert to HEX and we get 365A, we reverse it and get 5A 36. We change the first 17 blocks to 5A 36 and change all check sum blocks to 00 00 and we program and we should see 236,538 on the display. So if we change the first block to 5B 36 an the first checksum to FF FF we then should display 236,539, and if we keep going a total of 8 times (236,546 – 236,538 = 8) we then would get the exact number we want 236,546

this now makes perfect sense. It took a minute to click in my head what and how this was actually working but I get it now!

I need to do some real world testing next! Stay tuned for more fun and adventure!

ZC