HighWind, #908:
Hi, I'm an Electronics engineer working in another industry, involving MW size
power electronics, UPS systems, and IEC61508 P/P/E.
Very welcome and judging by the name, location and work, wind energy systems
perhaps ?.
I would prefer two, not only for redundancy, but also for insulation monitoring.
The battery is in a conductive casing, and is expected to be insulated from the
power circuits. But one single fault in a battery could connect it to the casing,
thereby creating a circuit that can't be disconnected. I would insert a HAL
sensor in both the positive and the negative connection, for insulation detection.
Very good point, since any differential / imbalance between the two current sensors
are an indication of leakage to ground. Ideally, each sensor would be on the busbar
terminations at the power socket, to cover leakages right up to that point.
I would prefer two separate contactors (or at least a two pole), with normally open
contact sets, one installed in the positive and the other in the negative connection.
The contactors and BMS has to be able to detect and break the highest possible short circuit current.
Hadn't thought about about case leakage isolation, which does need contacts in
each line to work. There's stiil the problem of >1 cell shorting to case, but I guess
that scenario could be covered by fusable links between the cells, rather than the
existing copper straps. A two pole contactor would get the job done and should be
more reliable and cost effective as well. However, my point would still be that
if the cells are properly managed and run within data sheet limits, the problem
won't arise in the first place.
This would then create a chicken and egg problem, since the contactors has to be
unpowered, when the aircraft is parked in order not to drain the battery, and the
system needed to wake-up the battery has to be powered from the same battery.
This could be solved by providing power via. the data communication connector to the charger/DC bus controller.
That's just a system design issue and could perhaps be covered via a low power micro,
with uA in sleep mode, always connected to the battery, or even a small disposable
memory style backup cell. The micro is then woken up by the first comms packet down
the line.
Over the past few weeks, many people contributing here have found loads of issues
with the design and while hindsight can be a wonderfull thing, the present design
looks like half the job that should have been done. The introduction of cots tech
into aviation may have saved loads of money, but I wonder if some of the design
rigour and attention to detail have been thrown out of the window at the same time.
Having said that, I'll bet the design teams were really pleased with themselves to
start with
...
Regards,
Chris