@syseng68k
If you look at the schematic in the ntsb report, it's quite clear that the
contactor is drawn with normally closed contacts, which is the contactor state
with the coil / solenoid unenergised. While it may be latching to the open
state, there's no evidence for this and in fact, the contactor looks fairly
standard in that respect, but yes, it could be latching to the contacts open
state.
I originally brought this up in tech log, (#902), which might better explain why this isn't a very good idea..
Well, I did re-read your post of yesterday, and in the clear light of day I can't understand why I thought you had accidentally confused NC with NO. I have to stop posting at 3 in the morning-- time for the red face...
I did go back and read your #902 post again. At the time I read it rapidly, but I recalled then noting that your position was that the coil circuit of this contactor should be carrying current with the aircraft in service. This is the general practice, and there are good reasons for it.
In my comment about the latching (to hold the NC contactor open without continued coil current), I was aware that no one had pointed out this possibility, and wanted to explain that it was possible to operate the contactor arrangement without having it drain the battery. I may have left the impression I thought this was a great idea-- but to the contrary I am just trying to understand the design philosophy operating at Boeing.
At the time I wrote, I had just that day ferretted out information on the size of the two alternator-starters on the APU (not small); and just today I've further determined that there has to be a large DC load on the APU battery bus when the APU is running (not just starting). The contactor coil current was posted today on the Tech board (2.6 amps IIRC, holding I presume), noted as being enough to drain the battery in about a day. I don't know the amps of the large DC load, but it would be rather substantially more, I'd think.
Over on Tech, mm43 today also posted that he thought some form of contactor latching had to be involved, possibly on the overcharge end of the range as well, and perhaps some form of electric unlatching (perhaps if the supervisory board draw bleeds off excess charge?).