FIS character set
Which byte code shows which character on the cluster's centre display: close to ASCII, but with the lower-case letters a to o moved to 0x01-0x0F and accented letters, arrows and symbols from 0x80. Transcribed from one 2016 table, uncertain cells marked.
The text messages for the cluster display (see Writing text to the cluster display) carry one byte per character. Which character each byte shows is set by the cluster’s own table, which is close to ASCII but not the same.
How it differs from ASCII
Section titled “How it differs from ASCII”The post says the table Audi uses is “more limited than the official ASCII table”, and that “especially the little letters ‘a’ to ‘o’ are in a different position than regular”, so text has to be converted before it is sent to the display. The table shows:
0x20to0x5Ematch ASCII: space, punctuation, digits,@, the upper-case letters and[ \ ] ^.0x5FisÄ, not the underscore.atooare at0x01to0x0F, andpis at0x10. Their ASCII places0x60to0x70hold other characters (Ö,Ü, punctuation) or are empty.qtozstay at their ASCII codes0x71to0x7A, and{ | }at0x7Bto0x7D.0x80to0xBFhold accented letters, arrows and symbols such asß,£,µ,½and§.- Some characters appear at more than one code: the four arrows (
0x18–0x1Band0xAC–0xAF),ñ(0x8A,0x9A),ç(0x8B,0x9B),ş(0x8C,0x9C), the space symbol (0x00,0x17,0x20,0x65), the short upright stroke (0x27,0x62),.(0x2E,0x63),,(0x2C,0x64),_(0x12,0x66),*(0x2A,0x6D),>(0x3E,0x6E),-(0x2D,0x68),%(0x25,0xA3),V(0x56,0x67), andi(0x09) if0x8Eturns out to be anirather than¡. The table does not say whether they differ on the display. - The screenshot shows rows
0x0to0xBonly. Codes0xC0to0xFFare not in it; whether the original table had them is unknown.
So TEST is 54 45 53 54 as in ASCII, but Test is 54 05 73 74: the e moves to 0x05, while s and t stay. The worked example is on Writing text to the cluster display.
The table
Section titled “The table”Row is the high hex digit, column the low one: the cell in row 4x, column x1 is code 0x41. Hover over a cell (or tap it) for the code and any note.
| x0 | x1 | x2 | x3 | x4 | x5 | x6 | x7 | x8 | x9 | xA | xB | xC | xD | xE | xF | |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 0x | ␠ | a | b | c | d | e | f | g | h | i | j | k | l | m | n | o |
| 1x | p | √ | _ | ¤ | ␠ | ↑ | ↓ | → | ← | |||||||
| 2x | ␠ | ! | " | # | $ | % | & | ' | ( | ) | * | + | , | - | . | / |
| 3x | 0 | 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | : | ; | < | = | > | ? |
| 4x | @ | A | B | C | D | E | F | G | H | I | J | K | L | M | N | O |
| 5x | P | Q | R | S | T | U | V | W | X | Y | Z | [ | \ | ] | ^ | Ä |
| 6x | Ö | Ü | ' | . | , | ␠ | _ | V | - | ? | * | > | ||||
| 7x | q | r | s | t | u | v | w | x | y | z | { | | | } | ¯ | ||
| 8x | á | à | é | è | í | ì | ó | ò | ú | ù | ñ | ç | ş | ß | ¡ | IJ |
| 9x | â | ä | ê | ë | î | ï | ô | ö | û | ü | ñ | ç | ş | ĝ | ı | ij |
| Ax | ª | α | © | % | π | ₠ | £ | ← | ↑ | → | ↓ | |||||
| Bx | ¹ | ² | ³ | ± | µ | ¿ | º | ¼ | ½ | ¾ | § | |||||
| Cx | ||||||||||||||||
| Dx | ||||||||||||||||
| Ex | ||||||||||||||||
| Fx |
- A
- reading uncertain
- ?
- cannot be read in the source
- empty in the source (no character)
- not in the data yet
- ␠
- space
Transcribed from a screenshot of an English machine translation of the post 'Audi ASCII table' (janssuuh, 1 April 2016); the original post could not be retrieved, so the screenshot could not be checked against it. The screenshot shows rows 0x0 to 0xB only: codes 0xC0 to 0xFF are not in it. Several characters appear at more than one code (for example the arrows, ñ, ç, ş); which of them the cluster prefers is not known. Nothing here has been checked on a car.
Reading the table
Section titled “Reading the table”Every cell was read from the screenshot, enlarged; cells were checked one by one, and empty cells were confirmed by measuring the ink in each cell. The marks mean:
- Normal cell: the character is clear in the screenshot.
- Uncertain (marked): there is a likely reading, but the glyph could be another character. Nine cells:
0x00,0x17and0x65, which show the same open-box space symbol (⎵) as0x20, but nothing says they act as a space;0x27and0x62, two almost identical short strokes that could be an apostrophe, a prime or an accent;0x8E, which is eitherior¡(its dot is taller than on theiat0x09and its stem reaches one pixel below the baseline, which suggests¡);0x7E, a short bar at the top (a macron or overline, not ASCII’s~);0x9D, agwith an accent that looks like a circumflex (ĝ) but could be a breve (ğ); and0xA9, which looks like₠and may stand for the euro sign. - Cannot be read (
?): one cell,0x6B. The table shows the Unicode replacement character (a black diamond with a question mark), so the character was lost when the table was put on the web. - Empty in the source: 23 cells that the table leaves blank, so no character is known for them.
- Not in the data yet:
0xC0to0xFF, which the screenshot does not show.
In total: 192 codes, of which 159 are read with confidence, 9 are uncertain, 1 cannot be read and 23 are empty.
The data is in site/src/data/protocols/fis-charset.json, written by hand; docs/data-model.md describes the fields, and npm test checks the file.
To verify on the car
Section titled “To verify on the car”- Send each code to the display and photograph the result, starting with the uncertain cells, the empty cells and
0xC0to0xFF. - Check whether the duplicate codes listed above look different on the display.
- Check whether
0x8Eshowsior¡, and what0x27and0x62show. - Check whether a B7 cluster, and a lowline cluster, use the same table.
Related
Section titled “Related”Parts in the wiring data
| Code | Name | Models | Location |
|---|---|---|---|
J285 | Control module with indicator unit in instrument panel insert | B6 and B7 | Behind display unit |
Sources
- Audi ASCII table (janssuuh, 1 April 2016) — janssuuhThe whole table. Original not retrieved (Dutch blog janssuuh.nl, URL unknown); read from a screenshot of an English machine translation supplied by the owner, so the screenshot could not be checked against the original.
- Known CANID's (janssuuh, 18 March 2016) — janssuuhThe example frame 54 45 53 54 20 20 20 20 ("TEST") on 0x261 to 0x267.