P0B91: block 18 rationality compares both late-pack channels

GM describes P0B91 as block 18 voltage rationality. The generic voltage-sense R range/performance name should not be reduced to a low reading or a battery-capacity judgment. The controller is testing agreement among several measurements.

Reconstruct the agreement test

Twenty times block 18 voltage is compared with pack voltage. Its absolute difference must exceed 70 V, and the corresponding block 19 comparison must exceed that limit as well. Both conditions are joined in the published failure row.

The monitor counts 160 failures among 170 samples, sampled every 100 ms. This is a more sustained observation than the direct low/high range checks on the same page. Simultaneous records of the two blocks and pack voltage are more informative than isolated readings collected at different times.

Do not ignore blocked evaluation

A valid block 18 input is required. RUN mode, the stated low-voltage supply range and more than 200 ms after contactor closure are also prerequisites. Active block-circuit and listed pack-sensing faults inhibit the rationality test.

Those conditions matter when a fault appears to disappear after memory clearing. First establish that the controller was eligible to run the comparison. The source does not identify which signal is wrong or prescribe replacing a battery block. A trained hybrid technician must follow the exact vehicle procedure to investigate the physical cause and confirm the outcome.

Source and applicability

GM 2010 10OBDG10 Hybrid Diagnostics, cited BPCM or hybrid-controller monitor only. Historical calibration values are not universal specifications; no per-code model, engine or hardware fitment has been established. GM PDF, physical page 232.