Patrick, does that controller have a brake input? Usually most of these type controllers have an input for a brake pedal switch. If so, this could be used with the LVC opto output. The other option, assumng you have either a resistive or hall effect-type throttle input is to connect the LVC lines to the controller ground and the throttle signal line. This will momentarily pull the throttle signal to ground, if one of the LVC circuits trips. The circuit lets go as soon as the load goes away, so the effect is like a big tuna hitting your line (sorry for the sportfishing metaphor, but that's what I'd be doing today if I wasn't stuck here, making BMS kits...
). If you don't let off the throttle, it will "hit" again. It definitely gets your attention, not unlike getting whacked updside the head, as my wife is prone to do to me.
With most LiFePO4 cells I've dealt with, the voltage stays up much better than SLAs, or pretty much anything else. The higher the cell's "C" rating, the more pronounced this is. Actually it is the C-rating times the capacity, minus the load. What this means is that you can get the same "staying power" effect with lower C-rated cells if you have more capacity. Anyway, the point is that for lots of setups, the pack feels as strong 30 seconds from the end, as it does right at the beginning. Without the LVC function looking after each cells voltage, you can easily get to the point that the cells dump, and when that happens, the cell voltage can go from 3V to nothing in about 10-20 seconds. It doesn't really matter what version of the TC54 detector is used, as what you are trying to do is trying to catch that sudden voltage drop, while it is under load, before it drains down to where even the resting voltage is well below 3V. I've tried 2.1V and 2.7V versions, both with a123 and LiFeBatt-based packs, and haven't really seen a difference. Basically, from what I've seen, thee is usually about 10% of capacity left when I get the first "hit". If I back off the throttle a bit, I can go anothr couple miles before it starts hitting at even half-throttle. In short order, it goes into an oscillation, at about a 1/2Hz rate, with even the slightest amount of throttle. At this point the resting voltage of the LiFeBatt packs is about 3.0V per cell, and about 3.1V for my a123-based packs. In both cases, the packs take back in almost exactly their rated capacity.
The other function the LVC circuits are useful for is to detect when you have a bad/weak cell. In this case, the cell's voltage will drop a lot more than the others, when under load. If you detect that this is happening, a quick check of the resting voltage of each cell should show the faulty one. Another indication of a weak cell is during the charge process if one of the orange LEDs repeatedly comes on a lot sooner than the rest. Weak cells lose capacity, so they drain quicker, and get full faster.
-- Gary
