That does sound simple but without doing the math I'd guess you'd need a pretty large capacitor to absorb the transients that a battery can.

Transient voltage drop out, then back, all within 35 msecs. Some kind of periodic BMS self-test, I believe we were told. The story we (finally) got was that they'd been told avionics couldn't see anything shorter than 50 msecs, which they took to mean that as long as they kept it shorter than that they could do anything they wanted, without documenting it. They may be right about most avionics, but those transients were flagged by our touch screen, which uses hardware from the world of Formula 1 auto racing, causing it to throw a fault and reboot. Not documenting it is bad enough, but repeatedly denying it even after we found it and showed them the data cost us MONTHS of wasted effort.

Transient voltage drop out, then back, all within 35 msecs. Some kind of periodic BMS self-test, I believe we were told. The story we (finally) got was that they'd been told avionics couldn't see anything shorter than 50 msecs, which they took to mean that as long as they kept it shorter than that they could do anything they wanted, without documenting it. They may be right about most avionics, but those transients were flagged by our touch screen, which uses hardware from the world of Formula 1 auto racing, causing it to throw a fault and reboot. Not documenting it is bad enough, but repeatedly denying it even after we found it and showed them the data cost us MONTHS of wasted effort.

What strikes me as even more obnoxious behavior is that their engineers clearly adopted the idea that it was OK to deliberately cause an anomaly that anyone using the spec on the load side would expect to be anything but deliberate.

Transient voltage drop out, then back, all within 35 msecs. Some kind of periodic BMS self-test, I believe we were told. The story we (finally) got was that they'd been told avionics couldn't see anything shorter than 50 msecs, which they took to mean that as long as they kept it shorter than that they could do anything they wanted, without documenting it. They may be right about most avionics, but those transients were flagged by our touch screen, which uses hardware from the world of Formula 1 auto racing, causing it to throw a fault and reboot. Not documenting it is bad enough, but repeatedly denying it even after we found it and showed them the data cost us MONTHS of wasted effort.

Can you share which model battery? Do you know whether they planned to correct it (software update maybe)? Or do the TB batteries all still do this?

It was the TX16. And since they didn't document or admit the behavior on this model, there's no way to know if they do the same on their other models. But I expect they'd use common BMS logic across their entire model line, so it's likely the same for all of them.

It was the TX16. And since they didn't document or admit the behavior on this model, there's no way to know if they do the same on their other models. But I expect they'd use common BMS logic across their entire model line, so it's likely the same for all of them.

OK, you guys made me dig through my records.

We first installed TB's TX16, the one that other Evos had used successfully. After we encountered the conflict and traced it back to the batteries, we swapped them out for TB17-2 batteries. They're TSO'd, and considerably more expensive, but exhibited the same behavior. Which buttresses the suspicion that it is common through their model range.

Yes, I agree it is probably common to all of them then.