I have a TLV validation routine that works with the standard 5xx/6xx algorithm. It rejects the block on an X430FR4133 Rev B that I have. The TLV area is populated, and the individual structures look plausible.
There's also an anomaly with an invalid REF tag structure at 0x1a1E and missing DCO calibration values that the datasheet section 6.10 says will be into the TLV area without a tag/length field (!) at offset 0x1a22
So it seems pretty clear this is bogus, but the question is: is this because this is experimental silicon and the TLV content is wrong, or will it be that the algorithm changed for the FR4xx/2xx family devices?
Legacy device table:
0xff0 80 7f 24 35 00 1a 00 00 01 00 00 00 00 00 00 00
TLV table starts at 0x1a00 and ends at 0x1b00
TLV table:
0x1a00 06 06 99 4f f0 81 20 10 08 0a a3 51 97 08 1c 00
0x1a10 3e 00 fc fe 11 08 24 80 01 00 8c 02 05 03 12 02
0x1a20 ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff
0x1a30 10 01 1c 96 12 01 00 01 54 01 3b 00 10 01 01 96
0x1a40 1e 01 00 d0 10 01 14 96 12 01 00 00 16 01 00 06
0x1a50 10 01 1c 96 16 01 00 60 10 01 1d 96 12 01 36 00
0x1a60 16 01 5f 04 10 01 02 96 16 01 b8 8c 18 01 30 aa
0x1a70 30 1a 10 00 c7 31 c3 3c ff ff ff ff ff ff 08 ff
0x1a80 bf 59 d6 97 49 de 6f c8 b7 5b 83 02 28 59 3c 53
0x1a90 e6 7e 0b 88 c9 76 d3 f6 9f 00 16 01 9b 03 88 01
0x1aa0 ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff
0x1ab0 0c 96 02 01 0e 96 1e 96 6e 97 45 96 2a 97 7f 00
0x1ac0 0a 47 1a 96 04 47 0e d0 12 07 04 06 16 07 04 00
0x1ad0 0c f7 06 60 0a f7 1a 96 3b 97 34 00 39 97 1f 96
0x1ae0 4b 93 04 96 4d 93 59 04 9e 27 3e aa 90 27 b6 8c
0x1af0 e3 26 34 17 14 0d e7 2b ff ff ff ff ff ff ff ff
BSP430_TLV_TABLE_IS_VALID: 0
CRC for [0x1a04, 0x1b00] expect 4f99 get 7e0c: CORRUPTED DATA
0x1a08 tag 08 len 10
Lot ID: 0x089751a3 ; position 28, 62 ; test results 0xfefc
0x1a14 tag 11 len 8
ADC: Gain factor -32732, offset 1
1.5V T30 652 T85 773
2.0V T30 530 T85 65535
2.5V T30 65535 T85 65535
0x1a1e tag 12 len 2
REF factors: 1.5V -1 ; 2.0V -1 ; 2.5V -1
0x1a22 tag ff len 255
End of TLV structure