Honda Insight Forum banner
781 - 800 of 830 Posts

·
Registered
Joined
·
1,289 Posts
I'm excited to see temperature sensors being added. I didn't realize that wasn't a thing before so I apologize for asking about temps so much, I can see why bulldog would think it's not ready for me yet, I push stuff hard and it could have a thermal overload in the mountains. I'll be really excited to see the temperature performance of the pack. Right now even when it's under 80 out and my pack isn't hot if I try and empty it in one climb it will start throttling my current output by about 50% SOC and then eventually trigger a recal because it's asking for so much of the battery when it's already drawn down so much. I'm hoping lithium is more balanced and doesn't have voltage delta or heating issues but time will tell.

All very exciting stuff, keep up the good work. I definitely want to order it as soon as it's available.

You mentioned that you drained and recharged the battery several times, was this a full drain to where it wouldn't let you take any more out? Were there any issues with doing this?
 

·
Linsight Designer
Joined
·
2,677 Posts
Discussion Starter · #782 ·
Clarification: The existing beta (RevB) LiBCM PCB has QTY14 temperature sensors (QTY10 onboard, QTY4 external). The next (RevC) LiBCM PCB has QTY17 temperature sensors (same as above, plus QTY3 that bolt directly to each battery module). On the firmware side, I still haven't programmed any of these temperature sensors. Honestly it's not much work, but I just haven't done it yet.

But @Bull Dog is correct that LiBCM isn't 'ready' for general release yet, as the firmware still needs work. Right now we beta testers are having to manually monitor cell voltage to ensure the pack doesn't go out of bounds (SoC). Natalya and Isaac are working on that firmware, but right now I"m focused on final hardware details. Once I finish all that up, I'll switch over to firmware.

Lithium self heating isn't going to be an issue for a long time. Over time the cell resistance will increase, but it'll be a long time before the lithium cells get above the brand new NiMH ESR.

...

I'm not sure which specific "drain and recharge" you're referring to. LiBCM is already quite good at preventing assist when the battery is empty and/or regen when the battery is full. There are some corner cases we're still working out (e.g. driving down a steep grade for several miles while the battery is initially fully charged). We'll get there in time. I'm not sure what your specific question was, so if I didn't answer it please clarify your question.
 

·
Registered
Joined
·
1,289 Posts
Sounded like on your trip you talked about discharging and charging it a dozen times in the mountains. I wasn't sure if that was full cycles or not and how well it behaved.
 

·
Linsight Designer
Joined
·
2,677 Posts
Discussion Starter · #785 ·
I've got the wiring figured out for all three 48S module configurations:
12S+/18S-/18S-
12S+/18S-/18S+
12S+/18S+/18S+

I also (probably) have a solution to convert an 18S+ module to 12S+. Waiting for off-the-shelf parts to arrive to verify it works.
 

·
Premium Member
Joined
·
110 Posts
I've got the wiring figured out for all three 48S module configurations:
12S+/18S-/18S-
12S+/18S-/18S+
12S+/18S+/18S+

I also (probably) have a solution to convert an 18S+ module to 12S+. Waiting for off-the-shelf parts to arrive to verify it works.
Sure, since I have 12S and 18S :). Seriously though, fantastic news, I haven't even tested my 18S for polarity yet.
 

·
Linsight Designer
Joined
·
2,677 Posts
Discussion Starter · #787 · (Edited)
FYI:
12S+/18S-/18S+ is the ideal 48S pack configuration.
Based on what I've seen, each cardboard box (containing QTY2 18S modules) has either QTY2 18S+ OR QTY2 18S-... so if you didn't luck into QTY1 of each 18S battery type, you'll want to swap with someone local if you want to squeak out every last drop from your LiBCM conversion.

12S+/18S-/18S+12S+/18S-/18S-12S+/18S+/18S+Notes
Total 4 AWG Cable Length (mm)170222672405
Cable Length between 18S modules (mm)212762660Adds resistance...causes voltage error on C18/C19 under heavy assist/regen
Anderson Connector between 18S modules?NoYesYesAdds resistance... causes voltage error on C18/C19 under heavy assist/regen
Total Anderson Connector pairs in pack112'Good' requires 2nd pair to safely isolate modules during pack assembly
Relative performanceBestBetterGood

Note that any configuration will work... but heavy assist as the pack empties will cause LiBCM to tap out sooner with the 'better' and 'good' configurations (due to larger C18/C19 cell voltage measurement error under heavy assist/regen). Most people won't notice a major difference between the configurations, but just figured I'd mention it now for anyone who wants to swap modules.

...

Also, if you don't have a 12S+ module (I believe all 12S modules are 12S+), note that my 18S+ to 12S+ conversion process requires an 18S+ donor module (i.e. an 18S- module CANNOT be used to create a 12S+ conversion). This is due to how the BMS harness is laid out. Theoretically you could de-pin the BMS connector on an 18S- module, then flip all the cells around, and finally re-pin the connector... but I leave that as an exercise for the reader (it's not an easy process).

Another possibility is that I could create a custom "18S-/12S+" adapter cable (between the OEM 18S- BMS connector and LiBCM's ribbon cable header)... but then if you used said cable anywhere else it would destroy something (probably an LTC6804 on LiBCM, but who knows). I don't think this is a good option.

So maybe one last option for those that don't have 12S modules - and only have 18S- modules - would be to provide a custom ribbon cable with ring terminals that go all the way to each cell's threaded stud. That would prevent issues, and wouldn't be any additional effort (on both my end or yours), because you already have to take the BMS harness off as part of the 18S->12S conversion process. This is probably the correct answer.

Don't worry about committing all this to memory... I'm mainly just writing this for my future reference. I'll make sure the solution is simple.
 

·
Premium Member
Joined
·
1,106 Posts
Mudder, with your recent posts talking about converting the 18S to 12S, are there still plans for a 54S option?
 

·
Premium Member
Joined
·
110 Posts
I have QTY4 18S and QTY1 12S so hopefully I have ideal. Still need to open boxes and test, but building the raceboat and dealing with rheumatoid arthritis so I'm moving much slower than usual.
 
  • Like
Reactions: jime

·
Linsight Designer
Joined
·
2,677 Posts
Discussion Starter · #791 ·
Mudder, with your recent posts talking about converting the 18S to 12S, are there still plans for a 54S option?
54S is still on the docket, but it will require substantially more work, whereas 48S is well underway and mostly sorted out at this point. My #1 goal with my previous post was to make sure people can get 48S working without a 12S module (which as you know is in quite limited supply).

For (at least) the first several months, LiBCM is 48S only. Note that ALL LiBCM PCBs will support up to 60S... just initially the firmware will only support 48S.
 

·
Registered
Joined
·
1,289 Posts
Is the 60s worthwhile? I was under the understanding that 54s already has to be limited to prevent overvoltage? You'd have to run a meanwell and it wouldn't have much more capacity or power than 54s. Is 60 the realistic upper threshold for high power applications?
 

·
Linsight Designer
Joined
·
2,677 Posts
Discussion Starter · #793 ·
Is the 60s worthwhile? I was under the understanding that 54s already has to be limited to prevent overvoltage? You'd have to run a meanwell and it wouldn't have much more capacity or power than 54s. Is 60 the realistic upper threshold for high power applications?
60S is more geared towards lower voltage lithium chemistries (e.g. LiFePO4). Note that the 'extra' 12S bank (i.e. cells "49 to 60") are completely isolated from cells 1 to 48. Therefore, the 'extra' 12S bank could be used to monitor a lithium/NiMH 12 volt battery (i.e. remove the battery under the hood, and replace it with a lighter one in the IMA bay). That's not something I'm focused on now, but I will eventually do that in my car.

...

Increasing from 48S to 54S will increase capacity 13%. Increasing from 48S to 60S will increase capacity 25%. Of course if you're wanting more capacity, a simpler option will be to just put more cells in parallel. Note it is trivially easy to reconfigure these lithium modules (e.g. an 18S pack can easily be converted to 9S2P, 6S3P, 3S6P, etc). For example, you could triple the total capacity by using QTY8 reconfigured 18S modules (each reconfigured to 6S3P)... that would yield a 15Ah 48S module. Note that you'd save quite a bit of space by removing the plastic spacer between each cell (which is used to allow airflow between cells in the OEM configuration (but not in LiBCM). You'd then need to manually connect LiBCM's "DIY BMS ribbon cable" to the modules... but that's just a mechanical implementation... LiBCM can monitor any lithium battery chemistry you want... up to 60S.

...

Certainly there will be more power at higher voltages... but not as much as you might guess. The primary limitation is going to be that LiBCM can only spoof the voltage down to 67% of whatever it actually is... and given that the OEM MCM throttles down current proportionally with voltage increase (above 120 VDC), it won't be as much as you'd otherwise expect. 60S with standard lithium is ~240 volts, which LiBCM can spoof down to 161 volts... so you should expect about 8 kW less than the "on paper" power calculation... which would still be around 26 kW or so, but given that I'm already getting 23 kW with 48S, you can see that there aren't huge additional gains by adding additional cells.

Of course you could add external resistors to apply an additional voltage spoof, which would entirely overcome LiBCM's hardware limitation... but that would no longer be "drop in", and would add quite a few caveats to an otherwise clean install. As you pointed out, you'd (probably**) need to replace the OEM DCDC converter as soon as HVDC exceeds 217-218 volts.

**I say "probably" because the voltage cutoff is controlled by a digital circuit inside the OEM DCDC... I'm almost certain this can be bypassed easily, and am 100% certain there's no hardware risk to do so (all internal parts are rated to much higher voltages, and the non-Honda base part number can handle around 300 VDC).

...

The real long term answer for "I want as much power as possible" will be to replace the MCM (e.g. with LiMCM, release late a LONG time from now).
 

·
Linsight Designer
Joined
·
2,677 Posts
Discussion Starter · #794 · (Edited)
Just wanted to add another complaint about how difficult it is to source parts right now... I'm scrounging up the various nuts and bolts wherever I can source them. I feel especially bad for our supply chain guy (who isn't sourcing the parts for this pet project), since he's been doing this full time throughout the pandemic; we aren't paying him enough.

If you're familiar with any of the various open source hardware websites - where people post their hardware designs - you've probably noticed all the comments stating:
"hello OP, I noticed you used part number 12345xyz on your PCB. Do you have leftover parts? If so, please email me at [email protected] and I will buy them from you. Even if it's just a dozen parts, I still want them."

That's how desperate people are.

Right now it's looking like I'll run out of the following parts first (in the order listed):
-EPM1210SJ (I've got QTY200 backordered 2022MAR, and just ordered similar part EPM1510SJA, which will probably work)
-AP63205WU (I've got QTY250 backordered 2022APR, and with a board spin can use AP6215)
-LTC6820 (I've got ~QTY300 backordered)

STFP!

...

I'll certainly be shamelessly pulling certain parts back off the prototype PCBs (when I get them all back, after sending out RevC units). I (finally) ordered a pair of hot tweezers for precisely this purpose... I've always wanted a pair for my home lab, and boy are they fun! I can now remove parts DIP/TSSOP parts up to 25 mm wide in just a few seconds... they're magical.
 

·
Registered
Joined
·
1,289 Posts
The peak voltage of 48s and the cutoff voltage of the dc-dc aren't all that far apart are they? I'd probably end up doing something with resistors in mine and keeping it 48s for quite a while unless there was a really compelling reason to go 60s. It sounds like until the dc-dc is figured out that I wouldn't gain anything besides capacity from going higher than 48s.

You're going to be releasing the first for sale version as 48s so if the voltage difference between that and 54s within the dc dc limits isn't very high I'd think most original adopters would stay 48s until another big piece of the puzzle is solved
 

·
Premium Member
Joined
·
2,036 Posts
Just wanted to add another complaint about how difficult it is to source parts right now... -snip-
So how many Linsight boards do you think you will have from the get go? I know a bunch are going to builders like Eli, but I know I'll be wanting QTY2(and definitely have the money for it) as soon as they are available. I have a failing pack in my turbo CVT and no pack in my blue car.
 

·
Linsight Designer
Joined
·
2,677 Posts
Discussion Starter · #797 · (Edited)
The peak voltage of 48s and the cutoff voltage of the dc-dc aren't all that far apart are they? I'd probably end up doing something with resistors in mine and keeping it 48s for quite a while unless there was a really compelling reason to go 60s. It sounds like until the dc-dc is figured out that I wouldn't gain anything besides capacity from going higher than 48s.

You're going to be releasing the first for sale version as 48s so if the voltage difference between that and 54s within the dc dc limits isn't very high I'd think most original adopters would stay 48s until another big piece of the puzzle is solved
Peak 48S voltage is 197 volts, which is 20 volts below DCDC cutoff.
LiBCM tries to keep the pack charged to 3.9 volts per cell (187 volts).
Peak 54S voltage is 221 volts, which is 4 volts above DCDC cutoff.
At 3.9 volts/cell, 54S is 211 volts, which won't trip DCDC off.

48S is the only initially supported configuration. This is almost entirely a firmware limitation.

So how many Linsight boards do you think you will have from the get go? I know a bunch are going to builders like Eli, but I know I'll be wanting QTY2(and definitely have the money for it) as soon as they are available. I have a failing pack in my turbo CVT and no pack in my blue car.
I'll be selling LiBCM in small batches... probably QTY10 kits per batch. Initially I suspect I'll ship one batch per week, and might get up to two or three batches per week. Right now it's just me in the garage, and there's quite a bit of labor, particularly making the cables. People have previously offered to help build subassemblies, and I will eventually figure out how to farm out those tasks... but initially it'll just be me catching up on TV shows while assembling the cables in my garage.

Eli (Bumblebee) won't be selling the initial batches, as I plan to sell them through an "Early Adopter Program". I'll basically be selling them to technical people who like to tinker. The rationale here is that the firmware isn't completely polished yet, and still requires a small amount of hand-holding to make sure the operating parameters are good. As soon as I'm confident LiBCM can handle everything without any user interaction, I'll roll Eli into the fold.

FYI: The "Early Adopter" LiBCMs will ship without firmware installed. Rationale: force customers to figure out how to upload new firmware (via USB), so that as new firmware is released, the early adopters will already know how to upload it. LiBCM units will (initially) also erase their firmware every 50 days or so... again, to force people to update the firmware in a timely fashion. Eventually this 'feature' will go away (probably in a few months). Specifically, a future firmware update will disable this "delete myself" code. It's all open source, so if you want to figure out how to disable it, it won't be hard... but anyone who can figure it out won't have any difficulty figuring out how to upload firmware, so I'm not worried about that group of people. I'm doing this so that old firmware versions don't linger around in cars for years.
 

·
Premium Member
Joined
·
2,036 Posts
I'll be selling LiBCM in small batches... probably QTY10 kits per batch. Initially I suspect I'll ship one batch per week, and might get up to two or three batches per week. Right now it's just me in the garage, and there's quite a bit of labor, particularly making the cables. People have previously offered to help build subassemblies, and I will eventually figure out how to farm out those tasks... but initially it'll just be me catching up on TV shows while assembling the cables in my garage.

Eli (Bumblebee) won't be selling the initial batches, as I plan to sell them through an "Early Adopter Program". I'll basically be selling them to technical people who like to tinker. The rationale here is that the firmware isn't completely polished yet, and still requires a small amount of hand-holding to make sure the operating parameters are good. As soon as I'm confident LiBCM can handle everything without any user interaction, I'll roll Eli into the fold.

FYI: The "Early Adopter" LiBCMs will ship without firmware installed. Rationale: force customers to figure out how to upload new firmware (via USB), so that as new firmware is released, the early adopters will already know how to upload it. LiBCM units will (initially) also erase their firmware every 50 days or so... again, to force people to update the firmware in a timely fashion. Eventually this 'feature' will go away (probably in a few months). Specifically, a future firmware update will disable this "delete myself" code. It's all open source, so if you want to figure out how to disable it, it won't be hard... but anyone who can figure it out won't have any difficulty figuring out how to upload firmware, so I'm not worried about that group of people. I'm doing this so that old firmware versions don't linger around in cars for years.
Hahaha that last part sounds like so much fun. I love figuring stuff out (I played around a lot with the LB bypass code) just because it gives me a deeper understanding on how things work. Especially when I get a eureka moment and it all makes sense. Anyway, I can't wait, and hope I get lucky enough to be one of the early adopters.
 

·
Linsight Designer
Joined
·
2,677 Posts
Discussion Starter · #799 ·
Hahaha that last part sounds like so much fun. I love figuring stuff out (I played around a lot with the LB bypass code) just because it gives me a deeper understanding on how things work. Especially when I get a eureka moment and it all makes sense.
You'll be able to figure it out. It'll even be commented... I certainly won't obfuscate it in any way. Again, anyone who can update firmware via USB isn't going to have any trouble figuring it out. I'm just trying to force people to spend 20 minutes to learn how to update LiBCM's firmware... can't think of a better way to force it than to send out non-functional hardware (until the firmware is loaded).

Anyway, I can't wait, and hope I get lucky enough to be one of the early adopters.
The Early Adopter Program will be open to anyone that can pass the 'quiz' I send them. It will primarily be focused on safety. Anyone that fails the quiz has no business poking around in the IMA bay.
 

·
Registered
Joined
·
1,289 Posts
I think I could pass a safety quiz, I was a meter monkey in a past life (electronics and instrumentation apprentice). I'd most likely be having Scott and bulldog install it. Any ideas on a date you're targeting for that? My car has to go back to them at some point for the cracked windshield anyways.
 
781 - 800 of 830 Posts
Top