Honda Insight Forum banner
501 - 520 of 567 Posts

·
Registered
Joined
·
135 Posts
Reading all of the posts has been an education in itself and well worth the time invested. I just put a new battery in mine (Thanks Eli!), just so I can get used to OBDIIC&C/IMAC&C and the diagnostics these give, as well as another pack and new cells to play with for some +40% fun. Hope to be fairly skilled by the time this is released.
And why would you quote me on that?
 

·
Premium Member
Joined
·
105 Posts
And why would you quote me on that?
was my opinion that reading through the entire thread is worth the time investment. Mudder covered it in his later post more concisely. I'm old ans ramble.
 

·
Linsight Designer
Joined
·
2,457 Posts
Discussion Starter · #503 ·
It would be cool if when bumblebee does their drop ins if they offer a battery trade in for people with current bumblebee batteries.
Bumblebee will almost certainly have to offer this option... otherwise where will the IMA battery enclosures (reused in LiBCM) come from?
 

·
Registered
Joined
·
61 Posts
I just checked Jag35.com to see what's new. It appears that the 2x18s modules are now only available for local pickup. But now he's selling the 18s modules individually for $90 and they will ship. Also, a new item has been added. A "compression kit" for G3 Insight modules. Maybe this would allow a 18s module to be converted to 12s.

 

·
Registered
Joined
·
210 Posts
Also, a new item has been added. A "compression kit" for G3 Insight modules. Maybe this would allow a 18s module to be converted to 12s.
The person in the video explained why the batttery prismatic cells need to be (slightly) compressed. It's so that in case something goes wrong, the pressure relief valves on the cells could rupture and do their job. If not restrained, the prismatic cells would deform and the pressure relief valves may not rupture.

I agree this "compression kit" may be a suitable option for converting 18S to 12S. The long bolts of the "compression kit" will slightly increase the width of the 12S module -- but IIRC in one of Mudder's videos there still was some free play when the 18S / 12S battery modules were inserted into the IMA battery case, so hopefully this won't cause an issue.

 

·
Linsight Designer
Joined
·
2,457 Posts
Discussion Starter · #507 ·
This compression kit won't work as LiBCM is designed at present... there's no way to mechanically connect the cells to the pack. Rest assured that if the 54S configuration doesn't work out (TBD, pending engineering effort), I'll design and have manufactured custom Honda-like side plates that work with the existing end plates.

Note that compression is important on pouch cells because it prevents dendrite growth from puncturing the insulator between the anode and cathode. Without compression, the dendrites can grow when the pack 'deflates' (i.e. low SoC), and can then puncture as the matrix expands in volume (i.e. high SoC).
 

·
Registered
Joined
·
7,745 Posts
^ These cells aren't pouch cells, they're fairly rigid prismatics. My understanding is that compression is mainly about reducing mechanical deformation under charge and discharge, without which the cells would expand and contract a lot more. That expansion and contraction can physically damage the electrodes, as the active material adhered to the plates moves a lot more, can crack and fall apart... This would be a major issue for non-rigid, flimsy pouch cells, obviously, but not so much for rigid prismatics. The electrodes are tightly compacted into the aluminum case, and the aluminum case resists major deformation. They aren't perfectly rigid, however, which is where extra compression comes into play... I don't think the cells would or should expand and contract (much) under normal operation, but they would/should if they're abused or faulty...
 

·
Registered
Joined
·
1,107 Posts
I don't think the cells would or should expand and contract (much) under normal operation
This could be read as "it's ok to run these uncompressed in a bind". There's a video someone here made showing the uncompressed LTO cells expanding (or was it these cells?) just in the short time they were removed from the holder compressing them. Since failure can be devastating (fire) I plan to never run these uncompressed.
 

·
Linsight Designer
Joined
·
2,457 Posts
Discussion Starter · #510 · (Edited)
There's a video someone here made showing the uncompressed LTO cells expanding (or was it these cells?) just in the short time they were removed from the holder compressing them. Since failure can be devastating (fire) I plan to never run these uncompressed.
That's my video, which was filmed with these cells. These cells absolutely expand and aren't nearly as rigid as they first appear.

My understanding is that compression is mainly about reducing mechanical deformation under charge and discharge, without which the cells would expand and contract a lot more.
That is correct, but note that expansion results in multiple failure modes (discussed below).

That expansion and contraction can physically damage the electrodes, as the active material adhered to the plates moves a lot more, can crack and fall apart...
This is certainly a valid concern when the cells aren't compressed. However, this specific result only reduces cell life... it's not dangerous by itself; de-lamination itself isn't going to result in a fire, but rather will cause the cells to fail prematurely. The increased ESR exhibited by a de-laminated cell will cause excess heat to build up in the cell, but the BMS would be able to see the resulting voltage drop and either throttle the current back, or disable the IMA system.

Another concern is dendrite growth, which I mentioned previously. Dendrites are dangerous in uncompressed packs because the expansion/contraction allows them to grow into a larger volume, which then allows them to pierce through the insulator between the anode/cathode when the volume later decreases. Dendrite growth is incredibly dangerous when a pack is left uncompressed, primarily because there's no external electrical 'signal' that the BMS can see which would indicate a dendrite is about to pierce the insulator... we can only see it as a rapid voltage drop that occurs as the module catches on fire... certainly too late to do anything about that once it's on fire.

This would be a major issue for non-rigid, flimsy pouch cells, obviously, but not so much for rigid prismatics. The electrodes are tightly compacted into the aluminum case, and the aluminum case resists major deformation. They aren't perfectly rigid, however, which is where extra compression comes into play...
These cells are not as rigid as you might think. Earlier in this thread, I took a module apart for just a few minutes and it was quite difficult to reassemble due to expansion.


I don't think the cells would or should expand and contract (much) under normal operation, but they would/should if they're abused or faulty...
They do, even under resting conditions.

Obviously many people have run many more lithium cells without compressing them... you might be ok for many years without compressing these cells, but I'm not going to risk it because the result is quite unpleasant: the energy contained in an LiBCM pack is roughly equivalent to just over three sticks of dynamite.

Put another way:
Each 18S module contains more energy than a stick of dynamite. It behooves us to incorporate every available engineering effort to prevent these angry joules from rapidly decomposing into a ball of fire in your IMA bay.
 

·
Registered
Joined
·
7,745 Posts
^ If you've actually seen them expand and contract a lot firsthand, can't argue with that. In general, I would expect y'all would be implementing the same or similar compression/assembly methods used in the original packs...
 

·
Linsight Designer
Joined
·
2,457 Posts
Discussion Starter · #512 ·
Correct: Right now we're using the OEM modules exactly as they're shipped. If we end up needing to modify 18S modules to 12S, then I'll reproduce the manufacturer's 12S end plates exactly.
 

·
Administrator
Joined
·
12,985 Posts

·
Linsight Designer
Joined
·
2,457 Posts
Discussion Starter · #515 ·
The huge hit of the meet was mudder's new LiBCM work. Sean is a beta tester and had just received the system. Beautiful and first rate professional work on mudder's part. I know we are all hoping for development to work out with not a lot of bugs. I don't think it will be pure drop-in, but with installation directions, I think it will be doable for many or most. Highly impressive.
Thanks for the praise, Jime! I've put a bunch more effort into this project than I had initially planned.

Installing the LiBCM kit into a pack looks like it will take a technically competent first-timer about four hours. At this point I'm doing the bench work start to finish in just over two hours. However, those that don't want to do the actual conversion work will be able to buy a drop-in LiBCM module from Bumblebee Batteries; Eli's crew does the pack conversion for you, then sends you the complete pack. From there, it'll be just like any other battery swap.

Thus far I believe Peter is the only beta user to have actually performed the pack conversion... he had some good feedback, which I will be testing out on my next conversion. I've been taking the time to iron out the final hardware & also building a manufacturing line in my guest bedroom. I just finished building a custom reflow oven good enough to reflow a board this large. Eventually I'll need to switch over to a better firmware architecture... that'll happen in due time.
 

·
Premium Member
Chicago & Detroit
Joined
·
1,244 Posts
Can you show a pic of your custom reflow oven with a brief description of how it works?
 

·
Administrator
Joined
·
12,985 Posts

·
Administrator
Joined
·
12,985 Posts
Now in the car and working. Excellent. :) Very early days of course.
I now have my OBDIIC&C and IMAC&C sorted to monitor/control things.


Done about 30 miles now, no significant issues so far.
 

·
Linsight Designer
Joined
·
2,457 Posts
Discussion Starter · #520 ·
Grid charging on the bench after flipping the charger anderson connection pins.
Plus a few other pack assembly thoughts.
Great to see someone else fiddling with LiBCM.
Notes on your notes:

In regards to balancing:
-You are correct that right now the beta testers should do their best to balance each pack prior to placing in the car.
-Eventually the firmware will support balancing, at which point pre-assembly balancing won't be required.
-The final firmware will have zero problem balancing even severely mismatched modules. For example, if one module is installed at 30% SoC, and another is 80% SoC, then it will take 48 hours for LiBCM to balance. Not too bad ;).

The cell voltage does rise quite a bit on regen... at 3.8 volts resting, hard regen will bring the cell voltage to 4.1 volts or so.

Good idea on heat shrinking the lugs.

In regards to requiring 5 volts on the first start after installation:
-This almost certainly won't be required on the final version.
-The issue right now is that the MVP code is too slow to boot, which causes the MCM to throw a P-code if LiBCM isn't already on when the key is turned on.

I didn't even think about the OEM polystyrene foam... I haven't had that in my car in years.

...

All good feedback, Peter. Again, it's great to see it in the wild!
 
501 - 520 of 567 Posts
Top