[BUG] M-2000C: PCN outputs not working

[charliefoxtwo]

Seems like device 10 was modified at some point

[charliefoxtwo]

Both devices 9 and 10 were modified in 2.9.2.49940, and I'm a bit stumped as to how they're encoding data now. e.g. for 42.54.5 we used to be able to read a variable that had the value 42545. Now we get this:

PCN_UL_SEG0:  i   
PCN_UL_SEG1: j jjb
PCN_UL_SEG2:  kk k
PCN_UL_SEG3: ld l 
PCN_UL_SEG4: m mmm
PCN_UL_SEG5:  nn n
PCN_UL_SEG6: goooo
PCN_UL_SEG7:   p p 

[charliefoxtwo]

Upon further investigation, it seems each SEG# corresponds to a segment on a 7-segment display (0-6, +7 for the decimal).

Lowercase letters map left-to-right, top-to-bottom on the image below.

this maps roughly like so (with the top row being a-h and the bottom row being q-x)

 k     2
j l   1 3
 o     6 
i m   0 4
pn    75

It appears different characters are used for different brightness levels. This would be fine, however it also appears sometimes segments are intentionally missing from the display. In these cases, it could be virtually impossible to use any heuristic to determine what number something should be (e.g. consider a 9, which may be a 9 or may also be an 8 missing the 0/i bar).

[charliefoxtwo]

Individual segments of the PCN displays are now exported - this is the only data we get from the module. As mentioned in the PR:

Output values are 0-3 depending on brightness.

Digits are left to right.

Output segment indices follow this chart (7 is the decimal, which is on the left):

 2
1 3
 6
0 4
75

The outputs are formatted as:

PCN_DISP_<DISPLAY>_<DIGIT>_<SEGMENT>

e.g. PCN_DISP_L_4_7.