Other Parts Discussed in Thread: INA300-Q1, INA300, , TPS25940-Q1, TPS1H100-Q1
This question regards multiple products in the TPS1H*, TPS1HB*, and TPS2HB* families.
Context: I am considering use of the TPS1HB08-Q1 in an application where the t_OFF parameter is a critical parameter. In this application, the TPS1HB08-Q1 defines the subsystem "fast" current limit (both short circuit protection and via the current limiting mechanism); a separate monitor defines a "slow" current limit across multiple subsystems using an INA300-Q1, with an \ALERT from the INA300 pulling down the TPS1HB* ENABLE. The bulk capacitance needed on the input rail is determined by the sum of the time to detect a violation of the slow current limit, and the time to disable the TSP1HB* as a result.
Observations and questions:
1) The TPS1HB* datasheets specify t_OFF(max) of 235 us. This is (not coincidentally?) the sum of t_DF(max) and V_BB/SR_f(min). There seems to be a potential confusion here; t_DF spans from 100% to 90% V_BB and SR_f is valid from 80% to 10% V_BB, but there's no definition provided on slew rate or delay as V_BB falls from 90% to 80%.
2) The TPS2HB* datasheets specify t_OFF(max) of 147 us, but the values of t_DF(max) and SR_f(min) are the same. Since I assume these are (nearly?) identical silicon, I suspect this is more detailed characterization that is less pessimistic. Is there any unofficial guidance as to whether this value is likely to be applicable to the TPS1HB* parts, and/or when the tests that gave this tighter bound may be performed on the TPS1HB*? [Side note: The "test conditions" column for t_OFF on the TPS2HB* datasheets is wrong; it is copy-pasted from the t_ON row and describes a turn on condition, not a turn off condition. Not reassuring.]
3) The TPS1H100 datasheet does not specify a t_OFF at all, but does specify both t_DF(max) and SR_f(min), so the assumed calculations from the TPS1HB* can be applied here -- but as already discussed, these are pessimistic by some 60%. Is there any chance of getting a reasonable t_OFF value for the TPS1H100? Is it at all reasonable to assume that, since slew rates are similar to the TPS1HB* family and t_DF(max) is 50 us faster, it will be very roughly 50 us faster?
4) In my application, turn-off is happening from 8.5 V initial V_BB, not 13.5 V. I assume that this will be faster, but I assume also that the slewing rate is fastest at high voltage, so the total t_OFF will be somewhere between the 13.5 V characterized t_OFF and (t_DF + (8.5/13.5)*(t_OFF-t_DF)). Is this a reasonable assumption? Is any other guidance available?
5) Are there other parts I should be considering in this application? Going above 50 mOhm would essentially eat remaining power margin, so I'm leaning towards the TPS1HB* over the TPS1H100 for that reason -- but the doubling in bulk capacitance hurts. Looking for a combination of short circuit protection, current clamping or limiting, and fast turn-off. Should I be considering something like a TPS25940-Q1 eFuse instead of a high side switch?