I would encourage everyone to re-read @Eli's post on battery voltages at the beginning of this thread. The proposed A123 pack is incompatible with the Insight because the car will push its cells beyond their normal operating range.
Even a slight imbalance can cause individual cell voltages to reach damaging levels because the voltages are beyond the knees in the charge curve, where a small change in charge can lead to a large change in voltage; a cell that's an outlier even slightly could reach a damaging voltage level before the rest of the pack.
In other words, an explosion is not out of the question if the cells are not handled properly.
The impact of such a failure is not necessarily limited to the experimenter. It can impact a passenger, a family member who borrows the car, or someone else on the road if a significant event causes the driver to lose control.
It's my oversight that I forgot to add a balance wire for each cell.
(That can be corrected easily enough in the schematic and why I said it needed checking!)
This pack of course like all Lithium IMHO should be used with a BMS, hopefully people know that.
The car will not automatically and inevitably destroy the cells (if we have a BMS)
The voltage range 50S I have suggested is a natural sweet spot for them when in balance.
One of the reasons for the wide variation in voltage when using Nimh is higher internal resistance.
A123 will be many times better in this respect when faced with assist/regen.
They don't dramatically sag or spike in use like ageing Nimh when in the normal SOC range.
When they are exhausted the car naturally cuts assist as the pack voltage is too low.
Ditto it cuts regen as the resting pack voltage rises to the top end.
We also have a whole host of gadgets to monitor them and manipulate the cars behaviour.
No modification is 100% risk free. It's a delicate balance of course and a discussion worth having.
Everyone's diverse Lithium opinions can be freely expressed on here.
^ I was writing the stuff below before you posted, Peter, so some of my points are the same as yours...
A 50 cell LFP would probably 'work' at the top in terms of voltage, with an allowed max of about 180V at 3.6V per cell vis-a-vis the car's absolute max of 192V. Even with a 120 cell pack of old NiMH cells, you really have to try hard to hit 192V. With only 50 cells, probably lower internal resistance, too, I can't see an appropriately sized LFP pack having an issue, even if you don't have extra BMS functionality, such as cell balancing and top and bottom voltage cutoff control...
The more important issue I'm seeing when skimming over what y'all have said, is the cell capacity. For example, OP has suggested 5Ah LFP cells that have only a 5C charge rate - 25 amp's of charging won't cut it. I suspect that's the way it will be with LFP cells in general - low charge rate capability, high discharge capability. So you either have to consider parallel cells or much larger capacity, like 10-20Ah cells...
One other concern has to do with trying to utilize the stock taps or generally, what extra 'wires' might be added to a pack or a pack's two modules... Off the top of my head, LFP cells have a minimum V of 2V, a nominal of 3.2V, and a max of 3,6V. A 5 cell 'tap' would thus have a min. of 10V, nominal of 16V, and a max of 18V. How would that square with the stock management were you using stock taps? That's something that'd need to be hashed out in a little more detail.**
Personally, I think there should be cell ports added to each 5-cell...module. Basically, every busbar should at least have a place where either a tapped (threaded) hole can be added, if not the threaded hole itself, so one could attach a ring terminal that belongs to a harness with a JST connector at the end (or something like that) and/or attach a stock voltage tap wire... I think the easiest thing to have a manufacturer do would be to simply add that threaded hole on each busbar, before they weld it to the cells.
I also think it might be prudent to make 10 5-cell modules per pack rather than two 20 and 30 cell modules. This would allow more flexibility for placement/fitment, in lieu of not having an exact schematic for how these things will be be installed into the existing pack space... 5-cell modules such as this would need to be connected together in a string, so the end busbars would have to have a hole for a bolt and nut or something like that.
** [edit, I think after Peter's 'thumbs up'] Again off the top off my head: I don't really see 5-cell LFP modules working with the stock 'taps'. For instance, as I recall, LFP cells are virtually empty at their nominal - 3.2V, or 16V at tap level. The analogous NiMH/BCM voltage would be 14.4V. So a 5-cell LFP module would be empty, but the stock management wouldn't see it that way, at least not right away. It would take an LFP cell truly reaching empty and the corresponding steep voltage drop to trigger the stock BCM 'neg recal' empty signal...
Normally, I recommend 3.65V max high and 2.3V lower limit for LFP cells. The factory curves are attached for your review. This LFP 20Ah cell has a max cont. charge rate of 4.5C (90A), max cont. discharge rate of 15C (300A), and a peak discharge rate of 30C (600A). Factory wants to know if you'd settle for all black wires for the taps.
Are they going to do them in 5 or 10 cell groups as has been suggested?
I don't like all black tap wires.
The negative end tap wire of a block should be black and the positive end red.
In between IMHO they should be different colours, brown, green, blue, whatever.
That colour code must be consistent between blocks.
Since there are multiple ideas on module size, how about just ordering individual cells and connector kits for each person to assemble their own modules. I can get the factory to punch holes in the tabs for the connectors. This should save some labor costs on module assembly.
They would like to see a photo of the "Sil 0.1" Female plug" if you have one. So, the quantity will be 100 5s modules and 10-sets of the 20s and 30s modules. Correct? If not, please clarify how many units of each module (5s, 20s, 30s).
Have you had that many people interested in the idea and willing to commit $$$?
I thought you were ordering some samples to try in your own vehicle.
Depending on price/delivery I would take 12 of the 5S modules to help support the project.
Basically enough to build one pack with a couple of spares for testing and evaluation.
Just thinking out loud.
If you/they did 4S A123 20ah cell modules as well they would also serve as lighter replacements for the 12V battery up front..
There has been a lot of talk about that recently and I'm using a 4 cell A123 packs as the 12V battery in my car.
I've not tried to market the drop-in Li-ion pack idea for the Insight-G1, because I got messages on this forum that the LTO packs recovered from other EVs were working fine and inexpensive, so run along and go do your own thing. I've been busy with other projects, so I let this one go. When I was asked if I can get some prices for 500 LFP cells, I asked my supplier, and they are willing to consider small orders from me since I've helped them with some of their customers' inquiries in the U.S. on EV and renewable power packs. If I had the funds, I'd buy 5,000 cells to make them available to others (like me) whose NMH packs threw the towel in.
Since the lead-acid battery doesn't last long, I put four spare LFP cells (80Ah/cell) together that I had from another project years ago to run the car until I have time to figure out how to use the little board and resistors I got from you. I haven't ordered any cells for my car until my learning curve on the car's electronics is less steep. I'm happy to get LFP or LTO cells for the group, but at the moment I don't have the time to tinker on my own car. Fortunately, I have a couple of other old cars that still run fine to move me around for the longer drives.
What you got on the LTO packs earlier has changed somewhat by the exhaustion of the supply. In addition, the LTO approach was never going to acceptable to a large percentage of Insight owners because of the conversion complexity. The idea of a drop-in pack using the stock mounting points remains a very attractive idea.
I'll play around some with the dimensions and use the various cell count configurations. Maybe there is an easy way
For my own pack, I plan to build my own frame to accommodate the cells for best maintenance access and air cooling/heating to match OEM mounting points. It is not essential for me to have the new pack to be the exact same dimensions or appearance as the OEM NMH pack. I guess function over form is more important for me. The very minor CG difference will not affect vehicle performance other than having greater power and capacity from a Li-ion pack. The LTO can handle this vehicle's charge/discharge current at a greater C-rate and at a lower Ah count than the LFP. The LTO will also have a greater service life by about 4-8X. The LTO will cost a bit more, but the much longer service life, higher C-rate tolerance, and broader operating temperature range outweigh the lower LFP price, in my view. I'm happy to go with the LFP pack if the group decides on the 20Ah LFP cell, so I can to follow your connection methods. The 500 cell quantity is not large but will save a few dollars over retail price.