@Hargobind: Brilliant! By coincidence, I was also working on a theory of one of the digits being a checksum - I had a hunch that the six digit code was actually a form of zero-compressed UPC-E that could be expanded into a standard 12 digit UPC.[1]
I'm going to work on getting new measurements to better-align with the five parameters - I'm most certain about the volume measurements, however. For example, the Columbian and Crema programs each dispense 250ml of water - exactly. Using Table 3 from the patent, this should correspond to 13 or 1101 in binary. I've tried to see that pattern from a number of angles - and it's not readily apparent.
That said, I think your observation of the 16 bits is the most likely solution.
Here's how the comparison breaks down when I run it.[2]
As we can see, the volume is really off - by as much as 120ml. And if anything, they should be identical. This is what is making me suspect that the tables and bit orders in the firmware are significantly different from the patent.
Also, the Charge cycle bits indicate two different "no soak" settings - both programs have the same soak time before beginning to brew.
Flow is a bit subjective to me - 30% of what? If I assume that it's a very constricted flow, that doesn't seem to match what the machine does with both programs - although I would surmise that the Crema program is more constricted than the Columbian, so as to produce the frothy head.
So, we need to divine a new set of tables, methinks...
[1] - en.wikipedia.org/wiki/Universal_Product_Code
This comment has not been screened by an external service.
[2] - blog.chapmanconsulting.ca/.../...n_colum_crema.png
I'm going to work on getting new measurements to better-align with the five parameters - I'm most certain about the volume measurements, however. For example, the Columbian and Crema programs each dispense 250ml of water - exactly. Using Table 3 from the patent, this should correspond to 13 or 1101 in binary. I've tried to see that pattern from a number of angles - and it's not readily apparent.
That said, I think your observation of the 16 bits is the most likely solution.
Here's how the comparison breaks down when I run it.[2]
As we can see, the volume is really off - by as much as 120ml. And if anything, they should be identical. This is what is making me suspect that the tables and bit orders in the firmware are significantly different from the patent.
Also, the Charge cycle bits indicate two different "no soak" settings - both programs have the same soak time before beginning to brew.
Flow is a bit subjective to me - 30% of what? If I assume that it's a very constricted flow, that doesn't seem to match what the machine does with both programs - although I would surmise that the Crema program is more constricted than the Columbian, so as to produce the frothy head.
So, we need to divine a new set of tables, methinks...
[1] - en.wikipedia.org/wiki/Universal_Product_Code
This comment has not been screened by an external service.
[2] - blog.chapmanconsulting.ca/.../...n_colum_crema.png
