DIY Electric Car Forums banner
1 - 5 of 5 Posts

·
Registered
Joined
·
8,093 Posts
If you're breaking down an OEM battery to the original module level, and you are going to assemble these modules to reach a reasonable voltage level, then you will have a large battery. Most OEM batteries for current EVs and plug-in hybrids run close to 400 volts, so - for instance - if you want to run half of that voltage you'll be using half of the full OEM pack, which is a big pile of battery when it is half of a Tesla Model S pack. This might be a challenge to fit in the Spyder, especially with an older motor and original Toyota transaxle taking much of the engine compartment space.

I suggest a look at specific module sizes and voltages, keeping in mind the voltage for which your motor and controller are suited.
 

·
Registered
Joined
·
8,093 Posts
I believe the Spark EV uses three 48v packs. I don't know how they are wired in the Spark, but I could wire them in series and get 144v. That would be a perfect match for my current motor and controller.
The Spark is uncommon - it was essentially GM's "trial project" before the Bolt so there aren't many of them, and they tried at least one unusual feature... the relatively low-speed/high-torque motor with only a single reduction gearing stage. Despite the odd motor, it would be strange if they used only 3x48 V nominal voltage; that seems at least a factor of two too low.

The Spark battery pack apparently changed significantly between the first version of the Spark (2014: cells by A123) and the second (2015: cells by LG Chem) The LG pack apparently has 192 cells - could that really be four in parallel and so only 48 cells in series? It's more likely that it resembles a second-generation Volt pack. GM's published 2016 specs say "p296s", which was presumably meant to be 2p96s (for a nominal pack voltage of 360 V, typical of recent EVs especially from GM)... so if there are only three modules they must each be 2p32s, and so 120 V (nominal) each, not 48 V. If they run around 48 volts per module, there are likely more modules. If the earlier pack with cells from A123 really ran at much lower voltage, there would have been motor and inverter changes between 20114 and 2015 as well... unless the inverter went from voltage-doubling to not doubling. :confused:
 

·
Registered
Joined
·
8,093 Posts
A Tesla battery pack is a bit more complicated, but I think it would work. Tesla uses 3.6v cells and they combine 74 cells in parallel into a module which should be ~21.6v. They have about 20 of these modules in series to get to ~450v. Getting this level of battery detail is convoluted as Tesla doesn't post the exact pack details. I think the information is generated by enthusiasts that have disassembled them personally. If the modules are even 25v each (hot off the charger), then I could run 5 of them and should have enough space in the MR2 for that. I guess I'd have a lot of spare modules. Maybe I could put a few packs in my Gem. :p
Although Tesla details can be hard to find, a lot of people (not me!) have put a lot of effort into literally tearing the components apart to understand them. Yes, my understanding is that the Model S/X have many 18650 cells in parallel, then six in series per module; the 16 modules are identical, and all connected in series. This means that several modules are need to hit a desirable voltage, but on the other hand some people are piling that amount of battery into original VW Beetles (apparently enough for a low-voltage motor can stack up behind the rear seat, under the window).
 

·
Registered
Joined
·
8,093 Posts
As a technically oriented auto enthusiast, I find it hard to see functional and nicely balanced sports cars turned into front-heavy things with no trunk space, so the typical mid-engine conversion with modules filling the front, engine, and rear compartments is what I would watch out for.

I assume that the Spyder continues the tunnel-mounted fuel tank of the first two generations of MR2; it would be nice if some modules could fit in that space, but it's pretty tight for modules not designed to fit there. Ideally, a low complete drive unit (almost any of them mount the motor around axle height, with only inverter and sometimes charger stacked on top) would be nice, allowing modules to be stacked on top without being excessively high. For a bizarre alternative, if the inverter/controller and motor could fit in the tunnel and be connected by a shaft to a final drive (differential) from front-engine/rear-drive car (a single-ratio system), with the charger in the front compartment, perhaps all of the required battery could fit in the original engine space.

Since no one has the development budget of the auto manufacturers, the best setup (if the wiring and software issues can be managed) would be the complete powertrain (battery to final gearing) of a production EV transplanted into the desired body and chassis, but that's not practical for many DIY builders.
 

·
Registered
Joined
·
8,093 Posts
The facts I read were on the A123 pack. I didn't know they changed packs in mid-run. Coincidently, one of the models on auction is a 2014.
An article about the change (GM Shifts 2015 Chevy Spark EV Battery Production In-House – Pushes A123 Out) showed the Spark battery, and it has six modules, not three... but that is apparently the 2015 pack with LG cells. These modules look very much like Volt modules - with the coolant manifolds low on each side - because they are GM's in-house assembly with the same or similar cells.

An earlier article (Aug 2013, so it must be for the A123 battery) compares the Spark EV and Volt: Spark EV versus Volt Battery. This early Spark appears to have four modules, each running the width of the box (almost the width of the car). This article says there are 336 cells, in a 112s(3p) configuration, so the modules are 28s(3p) each.

A classified ad offered A123 Spark modules:
I have 4 A123 5.37KwH battery packs. These are made with genuine A123 20Ah batteries. The spec sheet says they can do 18C. The modules are 97V and 60Ah.
That means that a roughly 48 V (nominal) group of cells would be only half a module. If you can physically arrange that, you could series three groups of 14(3p) each, for a 3s(14s(3p)) or 42s3p arrangement of 146 V and 60 Ah nominal capacity... or 8.7 kWh (nominal). Detailed specs for this pack give a 21.4 kWh capacity (rated), and you would be using 3/8ths of it, so the rated capacity appears to be 8.0 kWh; the pack usable capacity says 17.3 kWh, so 3/8ths of that would be 6.5 kWh. If you could use 192 V (nominal) from two very long modules, it would be a lot easier to keep the modules intact.

Of course GM would not have chosen to change battery suppliers after only a year, especially to a substantially different cell design; they went with A123 and that fell apart as A123 failed. They even ended up with six Volt-like modules a bit awkwardly packaged (extra space in the box), because they're in a box sized for a different configuration of differently-sized cells. I think I would be hesitant to put a lot of work into accommodating the A123 modules for the Spark, knowing that that they will always be rare, even compared to the LG modules for the Spark. I can only guess that the LG modules are six identical units (unlike the Volt, which as two different lengths and voltages of module).

The A123 and LG Chem cells have different chemistry, so charging and management details will be different.

I suppose I shouldn't be surprised, but for a relatively obscure EV there's a substantial amount of detailed discussion of the A123 Spark EV in this forum... just do a Google search for "A123 Spark" in DIYElectricCar.
 
1 - 5 of 5 Posts
This is an older thread, you may not receive a response, and could be reviving an old thread. Please consider creating a new thread.
Top