Stuart's Bosch Tassimo Hot Drinks Machine Forum

Previous Topic | Forum Index | Next Topic

Reverse engineering the Tassimo barcodes

Posted by private, Fri 15 Jan 2021 4:10 pm Edit | Delete
Thanks for the Tassimo repair guides! https://www.stuartdalby.co.uk/miscellaneous/tassimo_repair/ I'm reverse engineering the Tassimo barcodes by doing semi-automated testing. Do you have any barcodes or brands of T-Discs that are exclusive to the UK? I'm almost completely positive there is no difference between Canada/NA and European barcodes. We can still get Twinings Earl Grey but it seems to be packaged by JDE and not Kraft Canada.

Andrew
Posted by StuartDalby, Fri 15 Jan 2021 7:49 pm
Hi Andrew,

Thank you for your message. Glad to hear that you found my Tassimo repair guides useful.

I'm intrigued by your reverse engineering of barcodes. If you'd like to share more details I'd be interested to learn more.

When I had my machine I wasn't very adventurous in the range of T-Discs I tried. I don't think any were exclusive to the UK.

Sorry I can't be more helpful,
Stuart
Posted by private, Sat 16 Jan 2021 5:45 pm Edit | Delete
So, from observation of a T20 and reading other people's reports online I have come to these conclusions:

• European and North America products are the same, I have some T-Discs ordered from Spain to test and confirm 100%
o Some products on the Canadian Walmart/Sobeys shelves are made in Europe while some others are made in Canada. McDonalds/Tim Hortons, some other coffees are made in Canada. Earl Gray from Europe and it seems Canada is the only market where it's sold. I'm not sure why they discontinued all teas in the UK, maybe a licensing deal issue? You could get Tetley Orange Peoke here a few years ago.
o T-Discs look and function identically with minor differences that can be attibuted to running the product at a different manufacturing plant. The machines that were originally made and installed in the USA were moved to Kraft Canada in Don Mills a few years ago around the GroceryCo/SnackCo split. Mondelez outsourced their European coffee to JDE/Jacobs, North American coffee is owned by Kraft Foods. Kraft Global Foods owns the Tassimo brands and patents, cross licensed to Mondelez for Europe who get JDE to make the European products.
o Outside package plastic is thicker from Kraft Canada than JDE Europe. Foil on T-Discs are slightly different, Canada Thicker, JDE Thinner. Inner box-board cartons are slightly different, Canada Thicker, JDE thinner and easier to open.
• Brewer sales in Canada are discontinued.
o I'm stocking up on used units, they are functionally identical to Europe except with 120v/60Hz AC parts.
• Barcode last digit is a check digit that is stripped before converting to binary.
o My Excel calculation is this:
o =MOD(-SUM(MID(TEXT(D31,"00000\0"),{1,3,5;2,4,6},1)*{3;1}),10)
• Convert the numeric barcode data into little endian binary. This solves the problem with barcodes > 65535 looking incorrect (no need to invert data)
o I think this is due to the manufacturer ability to add additional bitfields into the barcode and older machines will ignore the lower bits as they will just be ignored when parsing if done from left to right.
o When converted to little endian all barcodes seem to have similar bitfield changes in the same places.
• Brewing is done in a set approach that doesn't seem to change, same as listed in docs
o Pre-wet period the T-Disc
o Pause period for soak
o Brew/mixing
o Purge with steam/air
• There are very notable patterns between similar products and the binary representations of the barcodes
o Brewing data is stored as field length delimited binary bitfields
o I think I have narrowed volume down to a minimum 4 bits, but it looks like some others are 3 bits in width
o I have two products, Twinings Earl Gray (LXB lot) and a Tim Hortons Steeped Tea (made in Canada) which have similar barcodes.
- 33395 - Earl Gray 236ml approx
- 33379 - Tim Hortons 240ml approx
- 110011100100000100000000
110001100100000100000000
- This is interesting as it doesn't look like the volume is different. I think the pre-soak is longer on one than the other. (4ml lost in the Earl Gray T-Disc?) I need to find my k-type thermocouple to see if it's temp or pre-soak that changed.
- Volume bits are both 1000 - Start 10 length 4
- What I believe is the 2nd or 3rd bitfield - Start 4 length 3
- 011 - Earl Gray
- 001 - Tim Hortons
- Is this temp or just pre-soak timing. I'm leaning toward pre-soak as the patents all outline a bunch of parameters that affect brew quality and this one should be finer grain than 'short/long' but less bits than pump time (volume)
- Further testing after I wrote the above seems to be correct with it being a pre-soak pause timer.
• Barcode readers between brewers seem to be interchangeable. I'm waiting to get a logic analyzer and see what protocol it is. Once I determine this, I'll be able to emulate the reader to fully automate the test rig.
• A new logic analyzer should provide cleaner data at a consistent sampling rate unlike my homebrew Arduino data acquisition code.
• It uses an NXP LPC2103 ARM microcontroller
There's a programming header and some sort of test/calibration header
o There's an EEPROM on board probably to store the temperature calibration values
o There's a cold side and hot side thermocouple connected to a voltage reference from the LPC
o Most I/O is only protected by 100 ohm or 330 ohm resistors
o Ground is referenced to line neutral directly. Don't reverse your line/neutral on the plug or it may go bang if you're connecting an external non-isolated device to is (Arduino/PC)
o Flow meter is supplied +12v from the main step down converter
- Flow meter outputs open collector square wave 50% duty cycle
- I tapped into the microcontroller's pin after the voltage divider on board
o Micro and other logic is operated at 3.3v from a separate buck regulator
o Heater is probably supplied PWM via a 12v drive triac through some other components
o Pump is supplied via a small SMD triac with low voltage trigger (3.3v) - not sure if PWM
o The micro firmware file name is on a little pink sticker on the board
o Opening the lid switch while brewing cancels the brew, great for automated testing
o There is no air pump, the brewer just pumps a little tiny extra bit of water into the heating tube, runs the heater watching the hot side to boil the water which is then forced through the T-Disc for final purge.
• The Tassimo machine makes excellent, consistent coffee and tea
o It's a bit sad that Kraft Foods just let it fail and rot as it's got huge potential for other things as well
o Patents are expired or expiring very soon so hopefully this may help create a third-party T-Disc market at least in France/Spain/UK where it's still moderately popular.

If you have time to have a look at my spreadsheet/dataset I can send a link.

Oh here's a couple pics of my test setup: https://imgur.com/a/gBpDar4

Andrew
Posted by private, Mon 18 Jan 2021 3:50 pm Edit | Delete
I've done a bit more research into this than most others ? and I have a logic analyzer on the way to see if I'll be able to emulate the barcode reader. If so, I'll have a fully automated test setup!

Another interesting thing, the Twinings Earl Gray seems to only be available in Canada and I think Germany. Strange that they wouldn't distribute it in the UK but it's probably due to licensing between JDE/Mondelez and Twinings.

Oh one more question for you, do you know if Twinings makes a large loose leaf bulk pack of English Breakfast in the UK (like for food service)? I can only get it in small tins or bags, but thankfully a couple grocers stock the 50ct box.

-Andrew
Posted by StuartDalby, Mon 18 Jan 2021 4:23 pm
Wow! Thanks for that Andrew. There’s a lot of interesting information there.

Best Regards,
Stuart
Posted by StuartDalby, Mon 18 Jan 2021 5:31 pm

Want to comment on this page? Contact Me

Back to Top