Honda Insight Forum banner

1 - 20 of 924 Posts

·
Linsight Designer
Joined
·
1,774 Posts
Discussion Starter #1 (Edited)
Howdy Insighters!
As y'all may know, I'm working on a drop-in G1 Lithium IMA. I had intended to use cells from a donor Chevy Volt, as it's:
-cheap (~1600 for a 6 kWh battery) <clarification edit: this is only part of a Volt battery>,
-plentiful (51,693 Volt's sold in the past three years; 74 presently wrecked on iaai.com), and
-has a built-in BMS harness (easy installation).

Alas, only two 48 V packs fit in the OEM battery compartment... at least three packs are required; ideally, four packs would fit. The Volt pack's dimensions leave much empty space:
VoltCells2.JPG
<clarification edit: the Volt's packs don't fit without complicated modifications to said packs>

This project requires widely available donor vehicles, so I found a list showing lithium car sales by year. The top five:
Screen Shot 2015-11-27 at 6.30.23 PM.jpg
Let me know if I missed another top seller.

Fortunately, 18 Leaf packs (2 cells per pack) fit in the OEM battery bay <clarification edit: there are 4 cells per Leaf module, arranged 2S2P>:
LeafCells2.JPG
**Note: The bracket that attaches to the switch that turns off the pack must be removed (spot welded to cover underside)

Unfortunately, the Leaf packs are more expensive, particularly the 2013+ packs (which last longer). Buying packs <clarification edit: enough packs for one G1 conversion> from a reseller (e.g. hybridautocenter.com) costs ~2300 plus freight. Buying a complete Leaf pack from a wrecked car costs ~3000 plus freight <clarification edit: If you buy a wrecked Leaf and remove the packs yourself, and then dispose of the car yourself, that'll cost you 3000; none of the other parts are particularly valuable>, and yields enough cells for 2.5 insight conversions.

...

It's possible to hack the OEM BCM & MCM to work with a larger lithium battery, but that solution is kludgy and not for most users. Thus, I'm designing a new PCB with the following features:
-BMS supports up to 48 lithium cells (36 Leaf cells fit in OEM enclosure)
-Use whatever lithium cells you want (programmable voltage set-points)
-OEM battery meter gauge displays ACTUAL state-of-charge
-Bluetooth iPhone/Android app displays battery/engine parameters
-Hardwired &/or Bluetooth input devices control system
-OBDII plug to get engine parameters (e.g. clutch/throttle/etc)
-IMA cruise control
-No CEL/IMA light
-Open Source, allowing full IMA control (no more MCM/BCM hacks)
(optional) Universal J1772 charging connector (expensive charger)
-USB port to reprogram, monitor system & retrieve logged data

The OEM wiring harness remains uncut; my design plugs directly into the existing connectors. Thus, the entire conversion is reversible, except for the metal bracket I noted above (see '**').

I intend to ship units Feb/March 2016.

...

So now the focus group questions:
0: What additional features do y'all want?
1: Would you pay 1000 for the PCB described above?
2: Would you pay 900 for a PCB that only supports 36 cells? The BMS components are expensive.
3: Advanced users: Is (only) 36S support (133:151 V) - instead of 48S (177:201 V) - a deal breaker?
4: Any alternate lithium cells I missed?

FYI: I'm making the PCB regardless. Just trying to see how much effort I should put into 'mass' manufacturing the design. If only the devout (or nobody) are/is interested, I'll skimp on the usability.
 

·
Registered
Joined
·
5,912 Posts
I'm down probably somewhere around $600 max... I take it the device would be able to over ride the stock programming in some manner, right? All I'd be looking for is a BMS for my choice of lithium cells (at least 40 cells), that can basically commandeer the functions the stock battery management does for NiMH, i.e. I want to replace my NiMH pack with a comparably sized lithium pack, I want a device that can allow me to do that...
 

·
Premium Member
Joined
·
750 Posts
No working prototype yet and Feb/March projected release date? Good luck. ;) In my experience projects of this scope end up taking at least 2-3x as long as one thinks they will, but we'll see how it goes! I have faith you can do it, I just think your estimation might end up being a bit off.

I'm not sure I'd pay that much, to be honest, especially when I also need to buy the new pack. I think I'm with eq1 on this one, I'd suppose $600-700 max or so.

36S doesn't sound like a dealbreaker, since the batteries would support a higher discharge rate, so we could still get the same watts into the IMA motor. Then again, I think 48S support would help future-proof the board a bit. Better too much than too little, right?

This sounds like an awesome project, by the way! I wish you luck!
 

·
Premium Member
Joined
·
2,818 Posts
Let me know if I missed another top seller.
Tesla ModelS from June 2012 to June 2015 sold a total cumulative 3 year total of 49,720 units in the U.S. alone .. 78,359 Globally.

4: Any alternate lithium cells I missed?
A123 cells ;)

So now the focus group questions:
I've enjoyed my own tinkering project .. soo this is not for me in particular .. but I do like/support the idea of a more 'drop-in' option to appeal to a larger % of Insighters.

3: Advanced users: Is (only) 36S support (133:151 V) - instead of 48S (177:201 V) - a deal breaker?
I don't think it should be a deal breaker .. As long as the cells used can keep their loaded terminal voltage .. above ~120v under a ~100A discharge load ... and bellow ~180v under ~50A charge load... It should be fine.

Lower battery nominal voltage pack would need proportionally lower internal resistance to be able to do the above terminal voltage under the IMA loads ... but given the OEM NiMH OEM internal resistance (see attached) .. getting ~10% less pack internal resistance for a ~10% lower nominal voltage should not be all that difficult.

- - - - -
Note: Although all the cells have their pros and cons .. one not widely publicized about the C-Max cells is their fairly linear voltage charge / discharge curve .. which makes identifying (and or managing) pack imbalances no longer require the cells to be at top or bottom SoC/SoE.

(See example attached of some testing I did a little while ago on a few)
 

Attachments

·
Premium Member
Joined
·
1,751 Posts
I'm with the other two. For 600 I would definitely be interested in one. But I couldn't justify 1,000$.
 

·
Super Moderator
Joined
·
4,690 Posts
We've been talking about Lithium conversions here for years. I think Peter had one running, but I'm not sure anybody else has. When we have these discussions it all becomes very complicated electrically and mechanically. Now Mudder is offering us a plug and play solution and all we're doing is worrying about a few hundred dollars? Sheesh. This is a huge thing for us, and Mudder needs to be supported. It costs what it costs.

It looks like the Leaf batteries are the best option, at least for pnp. It sounds like the whole thing could pretty much dropped in. We could get together and "share" the Leaf packs. Five people could share two packs, making the cost $1200 per car, plus Mudder's gadget. Maybe, with this knowledge, the question should be whether some of you would be willing to pay more for the "bigger" BMS. To support this decision it would be good to know what modification would be required to install the bigger, cheaper Volt cells. Isn't somebody already working on this? It's just sheet metal back there and there is a lot of air space.

What is the difference between a stock pack, a Leaf conversion (36S) and a Volt conversion (48S)? What would be our benefit? Would we still need to or be able to grid charge? Could we use our existing chargers? I don't know much about batteries, but I assume Lithium would give us lighter and more powerful packs that would stay balanced. 48S would be 25% bigger than 36S. What else? How about a simple "Lithium 101" post?

Would it be possible to offer both versions of the gadget?

Lots of us want to convert to Lithium, and this is our chance. Let's do what it takes to get it right and appeal to as many of us as possible.

Mudder, this is huge. Thank you for the effort you're putting into it.

Sam
 

·
Premium Member
Joined
·
1,751 Posts
We've been talking about Lithium conversions here for years. I think Peter had one running, but I'm not sure anybody else has. When we have these discussions it all becomes very complicated electrically and mechanically. Now Mudder is offering us a plug and play solution and all we're doing is worrying about a few hundred dollars? Sheesh. This is a huge thing for us, and Mudder needs to be supported. It costs what it costs.

It looks like the Leaf batteries are the best option, at least for pnp. It sounds like the whole thing could pretty much dropped in. We could get together and "share" the Leaf packs. Five people could share two packs, making the cost $1200 per car, plus Mudder's gadget. Maybe, with this knowledge, the question should be whether some of you would be willing to pay more for the "bigger" BMS. To support this decision it would be good to know what modification would be required to install the bigger, cheaper Volt cells. Isn't somebody already working on this? It's just sheet metal back there and there is a lot of air space.

What is the difference between a stock pack, a Leaf conversion (36S) and a Volt conversion (48S)? What would be our benefit? Would we still need to or be able to grid charge? Could we use our existing chargers? I don't know much about batteries, but I assume Lithium would give us lighter and more powerful packs that would stay balanced. 48S would be 25% bigger than 36S. What else? How about a simple "Lithium 101" post?

Would it be possible to offer both versions of the gadget?

Lots of us want to convert to Lithium, and this is our chance. Let's do what it takes to get it right and appeal to as many of us as possible.

Mudder, this is huge. Thank you for the effort you're putting into it.

Sam
I see what your saying and +1. I was thinking 3,000 for a pack and 600 for a module, not 1200 and 1000. That sounds much better.

What AH would half of a leaf pack be? Also, with a leaf pack, wouldn't we be able to charge it nightly?
 

·
Premium Member
Joined
·
2,818 Posts
When we have these discussions it all becomes very complicated electrically and mechanically.
Over the years and all those discussions .. and work of many people (including Peter of course) .. we have solutions for all those issues .. electrical / mechanical / etc... there really is allot documented and shared on this site .. a wealth of information.

There will be some choices to be made .. one can not always have their cake and eat it too... That choice part is what I find becomes the thing that tends to make many people squeamish .. It's getting into the fine details that can get complicated .. but those details are just about making those choices .. people need to either be comfortable with getting complicated to be involved in those choices .. or they just have to give generic general information to those who are getting complicated .. and those who didn't .. just trust and accept the choices that other person makes.

It's just sheet metal back there and there is a lot of air space.
Yup .. but any such modifications would probably scare away many .. thus the 'drop-in' type direction , forces it to have a more narrow form factor.

Would we still need to or be able to grid charge?
My guess is 'able to' .. not 'need to'.

Could we use our existing chargers?
Only if the BMS can disconnect them .. a simple relay could be triggered to do that.
Again choices.
Design BMS to be able to disconnect random type of grid charger ... or design BMS to communicate with a specific known grid charger?

Remember most of the NiMH chargers are very low power chargers ... the higher the battery capacity the longer it will take to charge... but more powerful charges cost more$... more choices. ;)

How about a simple "Lithium 101" post?
There are numerous different 'flavors' .. each slightly different from each other .. but in simplest generic over generalized terms.

1>
Lithium will NOT tolerate sustained trickle over charging like NiMH will ... thus the need to be disconnected .. and a greater benefit to cell level BMS .. and a need to prevent the OEM system from trying to warm up the pack (in cold weather) by force charging it.

2>
Lithium have a lower rate of self discharge than NiMH.

3>
Lithium have a higher cycle efficiency than NiMH.

4>
Lithium should drift apart in balance much slower than NiMH... due to 2 and 3 above.

5>
Lithium will be lighter for a given power and/or energy.
5A>
Soo a system of the same power and energy could be made smaller and lighter.
or
5B>
With the same size and weight have more power and energy.
or
5C>
Some middle between A & B above.

Of course other cells could be used .. bellow is just an example of how choices go.

For example .. a very smallest lightest .. with OEM power .. from something like 48s A123 26650 cells ... just ~8Lbs of batteries ... but it would only have about ~1/3 the OEM energy capacity.

Or ... for example .. to get full OEM power .. and about the same as OEM usable energy .. from something like A123 Automotive 32113 .. just ~20 Lbs of Batteries.

Or .. for example .. to get roughly the same usable energy as the newer ~8Ah NiMH BetterBatteries ... and could also handle the more power mods Peter did .. could be done with about ~27 Lbs of A123 32157 cells.

Or .. for roughly about the same space and weight as OEM NiMH .. all the way up to about ~4x the OEM usable Energy storage capacity .. and able to do about ~3x the OEM power levels .. with 48-50s of the A123 20Ah pouch cells.

Would it be possible to offer both versions of the gadget?
Options cost time and $ .. only worth that additional development cost if there is a significant market for A that B will not fit .. otherwise .. just spend the time and $ on developing one product.

Lots of us want to convert to Lithium, and this is our chance. Let's do what it takes to get it right and appeal to as many of us as possible.
:plusone:
 

·
Premium Member
Joined
·
2,818 Posts
What AH would half of a leaf pack be? Also, with a leaf pack, wouldn't we be able to charge it nightly?
I'm thinking 1/2 a leaf pack would be way too much.

I think Mudder was proposing much less .. 36 Leaf Cells ... Full Leaf pack is 192 cells.

Depending on the condition of course ... new Leaf full pack is rated for 24kwh Nissan with a usable ~21.3kwh ... 21.3 kwh / 192 = 111 wh per cell x 36 cells = ~4 kwh battery pack .. about ~4x more energy capacity than OEM NiMH (144x6.5=936wh) .. but will weigh a little bit more than OEM .. depending on build details (of course) ... my guess would be around ~20 Lbs more than OEM ... 36 cells can easily handle the full IMA 10kw of power ... Nissan rates the 192 cells as a combined 90kw ... 90/192 = ~468w per cell x 36 cells = ~16.8 kw .. the Leaf uses 80kw/192 = ~416w per cell x 36 cells = ~15 kw .. which I think is about as high as Peter went with his more power mods.

- - - - - -

I think the 2.5 packs from each Leaf pack .. was including the assumed weeding out of bad cells .. matching cells ... etc.
192 cells / 36 cells = 5.3 packs worth of 36 leaf cells each .. but from those 192 some might be bad .. or some might not match well ... soo ... I suspect that is why the 2.5 packs of 36 cells figure .. ie a conservative 90 usable and matched cells out of every 192 cells from a wrecked Leaf.

- - - - - -

From a straight $/Wh point of view the previous makes it look like the leaf cells are a better deal.

Leaf ~$3,000 for ~24kwh = $125 per kwh

Volt @ $1,600 for 6kwh = $266 per kwh

I suspect both with have the same comparative loss rate of balanced and bad cells.
 

·
Linsight Designer
Joined
·
1,774 Posts
Discussion Starter #11
I take it the device would be able to over ride the stock programming in some manner, right?
Yes, completely. The entire MCM & BCM are replaced, so the motor profile is 100% programmable... assist-only, charge-only, on-demand charge/assist, etc. Anything the community wants. I'm personally excited about long road trips, where I can command a constant 2 kW boost for over 4 hours - getting well over 100 mpg - and power up hills with an occasional 10 kW boost, then plugging her in overnight to recharge.

All I'd be looking for is a BMS for my choice of lithium cells (at least 40 cells), that can basically commandeer the functions the stock battery management does for NiMH, i.e. I want to replace my NiMH pack with a comparably sized lithium pack, I want a device that can allow me to do that...
I began my project expecting to only replace the BMS - leaving the MCM intact - but lithium discharge curves are so different from NiMH that the MCM would limit current long before the lithium cells were anywhere empty. As I progressed down the rabbit hole, I realized that the kludges IC members had previously done weren't going to cut it for a drop-in solution; the entire IMA control system needs replacement for what I have in mind: 100% motor behavior control at all times.
 

·
Linsight Designer
Joined
·
1,774 Posts
Discussion Starter #14 (Edited)
No working prototype yet and Feb/March projected release date? Good luck. ;) In my experience projects of this scope end up taking at least 2-3x as long as one thinks they will, but we'll see how it goes! I have faith you can do it, I just think your estimation might end up being a bit off.
Mario, I definitely understand your concern and appreciate the advice. I certainly can't promise a delivery date, but this ain't my first rodeo. I'm an EE specializing in point-of-load power conversion (i.e. DC->AC->DC conversion), and worked at National Instruments for nearly a decade before leaving to consult full time. I love what I do and am probably borderline Asperger's.

In regard to not having a PCB yet, I model entire systems on the computer prior to ordering a PCB. For reference, I spent less than 24 hours creating the schematic for my last project, which controlled four stepper motors, a closed-loop BLDC spindle, limit switches, progressive current limiting, temperature monitoring, a probe and also connected to a USB host. I then spent 36 hours laying the board out and the RevA has now shipped in over 1000 desktop CNC mills, called "Ghost Gunners" (see: ghostgunner.net). As the sole engineer on that project, I also designed the entire machine, the manufacturing line, the actual cutting code, and nearly everything else, including the manual... in just four months start to finish.
-Of the last ten PCBs I've worked on, 7 shipped RevA, 2 shipped RevB and 1 shipped RevC; in my defense, that RevC board was a 24 layer design created by me and two other engineers with 128 DDR3 modules, more than 50 power supplies, two FPGAs, and a custom ASIC that piped two DC-26.5 GHz 8 bit streams (streaming over 400 Gb/s). The opamps on the input stream were $500 each!

...

As it relates to the Insight, I just finished a 7 month trip and didn't have access to my G1. Thus, I've been forced into the theoretical:
-I started pouring over the G1 schematic right after the Wisconsin meetup in July.
-I condensed the entire IMA system into a single-paged schematic, and even discovered some service manual errors Honda made in the process ;).
-I read nearly every post relevant to this project:
Carduino - 1 http://www.insightcentral.net/forums/modifications-technical-issues/19401-carduino.html#post193742
OBDIIC&C Gauge - 1 http://www.insightcentral.net/forums/modifications-technical-issues/20488-obdiic-c-gauge.html
IMAC&C aka IMA Control & BCM Gauge Combo - 4 http://www.insightcentral.net/forums/modifications-technical-issues/19501-imac-c-aka-ima-control-bcm-gauge-combo-4.html
MIMA original (99mpg) signals and ToO: http://99mpg.com/mima/theoryofoperation
http://www.insightcentral.net/forums/modifications-technical-issues/18809-mima-code-discussion.html
http://www.insightcentral.net/forums/modifications-technical-issues/13041-new-mima-thread-33.html
http://www.insightcentral.net/forums/modifications-technical-issues/23415-bcm-master-board-pp50-a123-project.html
http://www.insightcentral.net/forums/modifications-technical-issues/16928-a123-20ah-lithium-install.html
Sanwiching 2x IMA motors together for EV - done
Parallel IGBT or IPU for more power? - done
Inside the G1 MDM & more power! - done
http://www.insightcentral.net/forums/modifications-technical-issues/22532-building-grid-charged-a123-20ah-pack-pp50-project.html
Mini BCM Interceptor - done
BCM Project - done
Analysing the BCM/MCM Signals - done
-I'm absolutely in love with my G1 "Squib".

So yeah, I can't foresee the future, but I don't have any concerns that would prevent me from having complete PCBs Feb-March. Not a promise, but this fun obsession is also my profession. I'm sure this comes off strong, but I'm trying to show that I'm obsessed with electronics. Probably my Asperger's is shining through a little, too ;).

As a side note, I've reached out to Peter in hopes that we can work together to shave off design time integrating his OBDIIC&C decoding work into this design.

This sounds like an awesome project, by the way! I wish you luck!
Thank you, Mario! I'm excited about the gauge control post you've shown us; if you can make the gauge programmable via the METSCI lines, then my PCB could program the entire screen without any additional modifications (besides your own).

Man it's a great day to be an insight owner!
 

·
Super Moderator
Joined
·
4,690 Posts

·
Linsight Designer
Joined
·
1,774 Posts
Discussion Starter #16
Tesla ModelS from June 2012 to June 2015 sold a total cumulative 3 year total of 49,720 units in the U.S. alone .. 78,359 Globally.
I don't consider the tesla system reconfigurable as a drop-in replacement. Their cells are simply too small to mechanically adapt to the G1. Please correct me if their design has changed, but when I looked at them a couple years back it was quite a chore to build smaller packs.

A123 cells ;)
I know you're using them in your G1 conversion; I've used them on other projects in the past, too (both 36650 and pouch). They're also not a drop-in due to mechanical fabrication requirements. This is why I was initially so excited to use the Volt pack... but the Leaf packs are easily assembled, too, because they thread together and have screw terminals with bus bars.

Note: Although all the cells have their pros and cons .. one not widely publicized about the C-Max cells is their fairly linear voltage charge / discharge curve .. which makes identifying (and or managing) pack imbalances no longer require the cells to be at top or bottom SoC/SoE.
I didn't look too much into the C-Max cells, as Ford has only sold 21,000 vehicles in the past three years and their selling fewer each year... not a large donor pool considering each insight requires the complete 5 Ah pack.

...

Thanks for the suggestions, Ian. If anything else comes to mind, I'm all ears.

My other research:
The CR-Z battery was another candidate, but it's only a couple hundred dollars cheaper than the Leaf cells, yet has significantly less capacity, and Honda's only sold 10,317 in the past three years.

The Prius EV's 23 Ah cells are a better candidate, as a 48S bank easily fits in the G1's battery bay and Toyota has sold 29,477 units in the past three years. Unfortunately, these packs are much harder to find in scrap yards, and they're about the same price as C-Max packs (so still less than half the capacity of a Leaf solution for just a few hundred dollars less).

...

My PCB's BMS uses screw terminals to connect to each cell, so customers can connect whichever lithium battery they want (CV limit programmable), but I'll be marketing the system for the Leaf cells unless someone proposes a better alternative.

As always, thanks for your input, Ian! I totally respect your work and love your contributions to our community.
 

·
Linsight Designer
Joined
·
1,774 Posts
Discussion Starter #17 (Edited)
I'm with the other two. For 600 I would definitely be interested in one. But I couldn't justify 1,000$.
Thanks everyone for your feedback.

Some features won't make it into a $600 board... the BMS costs around $100 per 12 cells. Thus, a 36 cell PCB has $300 in parts (aka "COGS", for "cost of goods sold") just for the BMS, not including anything else (e.g. microcontroller, plug-in harness, OBDII interfacing components, bluetooth transceiver, etc).

Typically the COGS are 25% or less of the final sales price, to cover R&D effort, sales, marketing, support, etc. However, the G1 is a passion of mine and I'm doing the R&D for myself. IC.net covers the marketing and hopefully the support, so I can (and will) certainly charge less than what this system would actually cost if I were designing it to make money.

As with nearly everything I do for the insight, I'm not in this for the money... If I were, those trunk latch replacements I sell would cost around $40, not $15.

Unlike the trunk latch, however, I suspect that some of my customers will incorrectly install my solution - damaging the PCB - even though it will be simple to install. I suspect that those customers will then expect me to replace the PCB free-of-charge, and thus I'm forced to recuperate some of that loss up front by charging more for the PCB.

For example, I previously sold window switch replacement parts, but stopped after the second customer incorrectly installed those parts (breaking them), and then threw a huge stink when I attempted to charge them my actual cost to send out another set. I hardly made any money on them anyways, and lost about $20 when I had to send out a second set. And then those people complained to me that they had to remove the dash a second time. It was an unpleasant experience for all of us, hence I stopped selling them (the parts are still there for anyone who wants to print them).

Long story short, I can't give the PCB away (as I do with the trunk latch parts). I suspect the COGs will be around $400, so I probably can't charge less than $800. Since all my designs are open source, anyone will be able to order their own PCB, populate all the (surface mount) parts, and have their own board, but I imagine their COGs & time spent will be considerably higher, as I have a pick&place machine that places all components with a computer.

...

I've let the cat out of the bag earlier than I would have liked; I typically don't announce a product until it's ready to ship, but in this case I needed some market research after the lithium battery was more expensive than I'd anticipated.

Thanks everyone for your feedback thus far.
 

·
Linsight Designer
Joined
·
1,774 Posts
Discussion Starter #18
We could get together and "share" the Leaf packs. Five people could share two packs, making the cost $1200 per car, plus Mudder's gadget.
Sharing two cars amongst five people would make a drop-in lithium battery (with 8x more capacity) cost the same as an OEM NiMH pack. Seems like a no-brainer to me.

Maybe, with this knowledge, the question should be whether some of you would be willing to pay more for the "bigger" BMS.
I'm fairly certain if the product is less than $1000 it will only support 36 cells. It's possible I could depopulate the components from the last 12 cells, but I'm still paying for the unused PCB real estate, plus I'd need to support both 36S & 48S configurations... not something I want to do if I'm also giving the PCBs away ;)... From the responses I've seen thus far, it seems obvious that I'll only design for 36S packs... maybe I'll throw a header on there to connect directly to an LTC-DC1894 Demo BMS board, and thus those that want the extra 12S can plug-and-play that board directly to my PCB via ISO-SPI. I figure anyone who wants the additional 12S can probably follow any directions I post on how to do that.

To support this decision it would be good to know what modification would be required to install the bigger, cheaper Volt cells. Isn't somebody already working on this? It's just sheet metal back there and there is a lot of air space.
The Volt cells are way too tall for the OEM battery bay, particularly because they have plastic cooling channels on either side. Even if those channels are cut off (not plug&play), the cells are still too tall without either 1). Cutting through the floor (not plug&play) or 2). Placing two 12S packs directly on top of the spare tire (not safe for inexperienced users to do). The Volt cells will work with my PCB for those that don't mind a little metal work, but I think that's beyond what most insight owners want to do.

What is the difference between a stock pack, a Leaf conversion (36S) and a Volt conversion (48S)?
Both the 36S Leaf & 48S Volt packs have ~8x more energy than an OEM pack. A 48S Leaf pack would have ~11x more energy.

What would be our benefit?
-You'd actually have a BMS.
-8x more storage = 8x more assist before grid charging. You could drive for an hour with non-stop full assist, assuming the IMA motor didn't overheat (it would). More practically, you could drive for four hours with 2 kW assist, which would almost certainly put you over 100 mpg, particularly with a "assist-only" mode (i.e. with grid charging).
-IMA cruise control.
-Complete control of the IMA motor personality... step on the gas and immediately get full assist (if that's what you want)... for the modders out there, push as much power as you design your system to handle.

Would we still need to or be able to grid charge? Could we use our existing chargers?
You would almost certainly need to grid charge to take advantage of such a massive battery. Otherwise, you'd probably end up only using a small range of the cells because charging an 8 kWh pack would kill your mileage ;). Any constant current charger will bulk charge lithium cells AS LONG AS YOU CAN DISCONNECT said charger once the first cell hits CV limit (typically 4.2 V, but I'm going to set it to 4.1 V for longer battery life). My PCB will have a FET to automatically disconnect a constant current source as soon as the first cell hits full. Thus, 'dumb' grid chargers will charge a lithium pack most of the way.
 

·
Linsight Designer
Joined
·
1,774 Posts
Discussion Starter #19
What AH would half of a leaf pack be? Also, with a leaf pack, wouldn't we be able to charge it nightly?
The Leaf sub-packs (smallest easily divisible unit, consisting of two cells) are 60-62 Ah, depending on the year. The 2013 and later units have much longer life, as Nissan switched chemistries. There are 48 sub-packs in a leaf battery, which is 96 total cells in series (355-403 V). Total Leaf battery storage is ~22 kWh. 36S insight Leaf pack is ~8.5 kWh (conservatively 7 kWh usable).
 

·
Premium Member
Joined
·
2,818 Posts
I don't consider the tesla system reconfigurable as a drop-in replacement. Their cells are simply too small to mechanically adapt to the G1. Please correct me if their design has changed, but when I looked at them a couple years back it was quite a chore to build smaller packs.
100% agree labor intensive to reconfigure would be the major con.

The only reasons I even mentioned it were:
  • It is another top selling BEV as a potential source of batteries ... used / from a wreck might come at a decent $/wh.
  • At ~265wh/kg the batteries they use are about as good wh/kg as one can get from any of the BEVs that one might be able to find wrecked in a junk yard for Cheap $.
  • Given Tesla has pushed up to ~6.3C discharge rates from their packs .. that's not bad power density either.
I know you're using them in your G1 conversion; I've used them on other projects in the past, too (both 36650 and pouch). They're also not a drop-in due to mechanical fabrication requirements.
100% completely agree .. The pouch and tab format is a significant mechanical fabrication labor time sink.

This is why I was initially so excited to use the Volt pack... but the Leaf packs are easily assembled, too, because they thread together and have screw terminals with bus bars.
100% completely understandably attractive feature.

I didn't look too much into the C-Max cells, as Ford has only sold 21,000 vehicles in the past three years and their selling fewer each year... not a large donor pool considering each insight requires the complete 5 Ah pack.
The metal cell cases are also and electrical path for the cells .. soo if you were tempted to use them .. you would probably need some at least basic electrical isolation around the cells... but they are also another easy square cell with threaded connectors.

My PCB's BMS uses screw terminals to connect to each cell, so customers can connect whichever lithium battery they want (CV limit programmable), but I'll be marketing the system for the Leaf cells unless someone proposes a better alternative.
Given the numbers you've posted above .. I think that looks like possibly the overall best route .. plenty units sold out there as supply .. good $/wh ... plenty of power and plenty of energy for this application (even at 36 cells)... and far easier for assembly than some of the other options.

Depending on how the details all shake out over the next couple months .. It looks like it might put your drop in system in the range of about $2,000-2,500 or so for the whole thing (PCB, Batteries,etc) .. which would be great if that works out.

Thanks for the suggestions, Ian. If anything else comes to mind, I'm all ears.
As always, thanks for your input, Ian! I totally respect your work and love your contributions to our community.
Welcome .. Happy to help.

Hoping to see a drop-in option available for the Insight Community.
 
1 - 20 of 924 Posts
About this Discussion
923 Replies
77 Participants
slociviccoupe
Honda Insight Forum
We’re the ultimate Honda Insight forum to talk about Honda’s hybrid car and its fuel economy and specs!
Full Forum Listing
Top