787 Batteries and Chargers - Part 1
Joined: Jun 2009
Posts: 1,330
Likes: 0
From: NNW of Antipodes
A very good question.
All the talk has been around balancing each of the 8 cells within the battery, but we seem to have within each of those 8 cells another 6 sub cells, but they are in parallel and provide the low impedance required for high current discharge.
There doesn't seem to be a practical way of balancing the sub cells - a crowbar comes to mind.
All the talk has been around balancing each of the 8 cells within the battery, but we seem to have within each of those 8 cells another 6 sub cells, but they are in parallel and provide the low impedance required for high current discharge.
There doesn't seem to be a practical way of balancing the sub cells - a crowbar comes to mind.
Joined: Mar 2006
Posts: 1,974
Likes: 0
From: Choroni, sometimes
To put it simple...
There are 48 cells in a 787 battery, right?
Eight "master cells" are balanced/monitored, right?
The sub cells are NOT BALANCED / NOT MONITORED, right?
There are 48 cells in a 787 battery, right?
Eight "master cells" are balanced/monitored, right?
The sub cells are NOT BALANCED / NOT MONITORED, right?
Last edited by hetfield; 3rd February 2013 at 20:29.
Joined: Jun 2009
Posts: 1,330
Likes: 0
From: NNW of Antipodes
Originally Posted by hetfield
The sub cells are NOT BALANCED / NOT MONITORED, right?
Joined: Aug 2007
Posts: 3,609
Likes: 0
From: London, New York, Paris, Moscow.
@ TURIN
No offence but could we please speak plainly for the sake of us simple folk.
Discount hard won experience at your peril.
etc,etc, as explained earlier in the thread re-hiring new "grads" and bumming out the older, expensive guys.
thought it was pretty straightforward really.
OOOO the rounders bit.
Sorry.
AND. Whilst the detective work in following posts is to be highly commended,
, the point isn't in the detail, it's WHY and HOW.keep at it guys.
Last edited by glad rag; 3rd February 2013 at 20:59.
Joined: Jul 2002
Posts: 596
Likes: 0
From: Ireland
hetfield
The sub cells are NOT BALANCED / NOT MONITORED, right?
If you look at the multi-pin connectors on the edge of the PCBs they look to be carrying 16-18 connections each, therefore no more than 36 'data' lines to the monitoring system.
Also if you look at the actual connections to each cell, the only connections are to +ve, -ve and the shunts between cells.
Joined: Aug 2011
Posts: 2,074
Likes: 0
From: Grassy Valley
I am not trying to be funny. Depending on the random (chaotic) transient (and functional) isolation of the cells, to include metallic dendrites, the actual number of 'cells', depends on characteristics that were not designed. That includes fluctuating polarities, in stress.
64? 137? xxx?
64? 137? xxx?
Joined: May 2009
Posts: 317
Likes: 4
From: sfo
Lyman
My concern is that someone who, unlike you, I, and others, may not have read the entire thread stumbling in here, sees your post, and believes that at least the battery in the JAL incident was not essentially brand-new.
Anybody remember the post some 300 ago about the article in Flight Global that mentioned that Boeing had changed the architecture of the battery, and was going trying to get the new version in the first delivered aircraft? Is that where they decided to go with the 6x8 configuration? Has this thing had ANY flight test?
Anybody remember the post some 300 ago about the article in Flight Global that mentioned that Boeing had changed the architecture of the battery, and was going trying to get the new version in the first delivered aircraft? Is that where they decided to go with the 6x8 configuration? Has this thing had ANY flight test?
Joined: Aug 2011
Posts: 2,074
Likes: 0
From: Grassy Valley
sb sfo
I had not considered that. We are on exactly the same page, except for one thing.
I do not think the "proposed change" was architectural, but chemical.
They reported they were switching to Manganese Oxides prior to delivery of the first 787 to ANA, its debut client.....as I recall.
So you got my point re "new". Yep.
A "fresh" unit of an "old" technology, etc.
I had not considered that. We are on exactly the same page, except for one thing.
I do not think the "proposed change" was architectural, but chemical.
They reported they were switching to Manganese Oxides prior to delivery of the first 787 to ANA, its debut client.....as I recall.
So you got my point re "new". Yep.
A "fresh" unit of an "old" technology, etc.
Joined: Jun 2009
Posts: 335
Likes: 57
From: Oxford, England
A few days away and there's pages of it to catch up with 
I don't know where the "subcells" is coming from, but there are 8 cells only. If it looks
like more, that's just the manufacturing process, or a trick of the light ?..
From what we know, all 8 cells are individually monitored and each cell has current balancing during charging...
I don't know where the "subcells" is coming from, but there are 8 cells only. If it looks
like more, that's just the manufacturing process, or a trick of the light ?..
From what we know, all 8 cells are individually monitored and each cell has current balancing during charging...
Joined: Mar 2006
Posts: 1,974
Likes: 0
From: Choroni, sometimes
I don't know where the "subcells" is coming from, but there are 8 cells only.
Eight packs a 65Ah, each consistent of 6 parallel connected cells.
To make clear, one single cell can't hold 65 Ah!
Joined: Jun 2009
Posts: 335
Likes: 57
From: Oxford, England
NTSB playing with high tech kit to find out what many suspected
anyway and still no info as to the cause of the problem, in public
at least.
My premise is still that there's basically nothing wrong with the
batteries, despite ntsb playing with them for many, many days, almost
obsessively. Makes one wonder what the real agenda is there.
Assuming that the batteries are ok, this leaves only a few possibilities.
1) Operational conditions far outside those expected, which may have
prompted software bugs in the charger or management boards to become
exposed during fast charging.
2) Again, software bugs or insufficient monitoring of individual cells
under heavy load, say apu starting, which caused one or more cells
to be discharge beyond safe limits.
Any out of band conditions should be logged by the charger, but there
appears to be no evidence of that from what we have been told, so a
hidden / intermittent software bug does start to look likely.
From info available, the chargers and monitoring have been used
elsewhere, but the software would be different and specific for each
application. Bugs can take years to become exposed in some cases and
only under certain sequences of events or timescales of operation. The
battery subsystem is still a very new product.
If the charger hasn't logged the real time sequence of events that led
up to the failures, then that is a deficiency in itself, imho...
anyway and still no info as to the cause of the problem, in public
at least.
My premise is still that there's basically nothing wrong with the
batteries, despite ntsb playing with them for many, many days, almost
obsessively. Makes one wonder what the real agenda is there.
Assuming that the batteries are ok, this leaves only a few possibilities.
1) Operational conditions far outside those expected, which may have
prompted software bugs in the charger or management boards to become
exposed during fast charging.
2) Again, software bugs or insufficient monitoring of individual cells
under heavy load, say apu starting, which caused one or more cells
to be discharge beyond safe limits.
Any out of band conditions should be logged by the charger, but there
appears to be no evidence of that from what we have been told, so a
hidden / intermittent software bug does start to look likely.
From info available, the chargers and monitoring have been used
elsewhere, but the software would be different and specific for each
application. Bugs can take years to become exposed in some cases and
only under certain sequences of events or timescales of operation. The
battery subsystem is still a very new product.
If the charger hasn't logged the real time sequence of events that led
up to the failures, then that is a deficiency in itself, imho...
Last edited by syseng68k; 3rd February 2013 at 22:58.
Joined: Aug 2011
Posts: 2,074
Likes: 0
From: Grassy Valley
syseng68k
The trick is understanding the technique. Yuasa calls the electrodes "wound", yet they present as "folded".
So I repeat my question, syseng68k:
If the structure of each plate is continuous, only round (cylindrical) packing is acceptable.
"folding" a long thin structure concentrates mass under stress at each "return".
This also concentrates heat at the buildup, and traps heat at a center that could be geometrically oriented closer to the outer surface of each cell, if each cell was a cylinder, not a square (rectilinear) shape.
Seems like a major mistake, and Mr. Musk seems to agree.
The trick is understanding the technique. Yuasa calls the electrodes "wound", yet they present as "folded".
So I repeat my question, syseng68k:
If the structure of each plate is continuous, only round (cylindrical) packing is acceptable.
"folding" a long thin structure concentrates mass under stress at each "return".
This also concentrates heat at the buildup, and traps heat at a center that could be geometrically oriented closer to the outer surface of each cell, if each cell was a cylinder, not a square (rectilinear) shape.
Seems like a major mistake, and Mr. Musk seems to agree.
Joined: Jun 2009
Posts: 1,330
Likes: 0
From: NNW of Antipodes
@ hetfield;
Texas Instruments ...
Overall, each group of parallel cells forms one of 8 series cells, and the only monitoring that is done is that of each parallel group as its contribution to the series. The very low impedance of the parallel grouped cells makes balancing of individual cells within the group extremely difficult.
Now, that's my interpretation, and I believe that is what the TI statement implies. I have been known to be wrong - on may occasions.
Texas Instruments ...
A battery pack is constructed from a string of series and parallel cells. Each series cell, or group of parallel cells, requires protection from over-charge, over-discharge and short-circuit conditions.
Now, that's my interpretation, and I believe that is what the TI statement implies. I have been known to be wrong - on may occasions.
Joined: Mar 2006
Posts: 1,974
Likes: 0
From: Choroni, sometimes
@mm43
Thx, I know....
True, but irrelevant, negligible?
Don't think so.
Thx, I know....
The very low impedance of the parallel grouped cells makes balancing of individual cells within the group extremely difficult.
Don't think so.
Joined: Jun 2009
Posts: 335
Likes: 57
From: Oxford, England
I really don't understand all the obsession with the batteries. Though
that is the part that failed, it's also the most low tech part of the
subsystem. Imho, those batteries were operated outside the data
sheet limits, otherwise they would not have failed. (Sticking neck far
out here)
The charger and monitoring systems probably have thousands of lines
of code and a vast array of electronics. Why have there been no questions
about that ?...
that is the part that failed, it's also the most low tech part of the
subsystem. Imho, those batteries were operated outside the data
sheet limits, otherwise they would not have failed. (Sticking neck far
out here)
The charger and monitoring systems probably have thousands of lines
of code and a vast array of electronics. Why have there been no questions
about that ?...