Adverse Effects wrote: have you/anyone looked in to bottom balancing and is there a BMS that dose it?
Yes, we've looked into it. Anyone with suitable programming skills can implement it with our BMS since it is open source. Nevilleh has implemented it on his open source BMS (whose daisy-chained opto-coupled comms was inspired by ours, which was itself suggested by Tritium_James). Use the "Advanced Search" option to search on the whole phrase "bottom balancing" in all AEVA forums for all dates.
Bottom balancing is probably the best way to go if your system controller cannot automatically back off both your motor controller and your charger in response to signals from the BMS (or if you don't have
a system controller). By "system controller" I mean something like Tritium's Driver Controls unit (which we use), or the Zilla "Hairball". Something that can control and read-status-from (a) your BMS, (b) your motor controller and (c) your charger.
But bottom balancing is less convenient than top balancing because you have to run the car flat (or nearly so) before you can do it, and it has to be manually initiated, and so may get forgotten. Whereas top balancing can be done automatically at the end of any charge.
In regard to Kurt's comment on top balancing at 3.45 to 3.50 V versus 3.60 V to 3.65 V: Again I think the difference is between systems that can throttle the charger current back to match the bypass current versus systems that can only turn the charger on or off. At the cell terminals there is a voltage rise above the rested voltage that is proportional to the charge current. If you're putting in C/10h up to the last second and then shutting off, then you will need to go to 3.60 to 3.65 V to actually achieve any balancing at all, and you will have to put up with the little bit of cell damage that occurs each time you do this. But if your system controller can throttle your charger back to C/100h when the first cell is full, then 3.45 to 3.50 V is a great place to balance.
Adverse Effects wrote:just been watching jack over at EVTV.me and it makes sence to me to have all the battery's alined voltage wise at the bottom of the SOC not at the top because the bottom is where it matters when stressing the battery's and a lot less chance of damaging/reversing a cell
Yes. I watched one of Jack's videos once. I find that it is a more enjoyable experience if your player can play them at twice normal speed. Jack's voice is quite deep so they are still quite intelligible.
It is good that you are seeking other opinions on the things Jack says.
There remains the question raised by Renard and Kurt (offgridQLD) as to why we didn't get the expected range. Kurt is right that we have gone for performance over efficiency, and it wasn't exactly a controlled experiment in regard to driving speeds or styles. There were a couple of laps of Mt Coot-tha in there and maybe 30 km with the top down, and maybe I did a few too many bursts of maximum acceleration just to see what she would do.
I had to laugh when, on the way to the restaurant, my wife informed me that if I went slower I'd get a lot more range. As I explained to her, I was laughing because I majored in Physics and am intimately familiar with the cube law of aerodynamic drag, so she was unknowingly teaching her grandmother to suck eggs, but for her it was (wonderfully) a matter of practical experience with the range-meter in the Nissan Leaf. But I wanted to find out what the range was if I drove "normally".
I note that we do not believe in undervoltage cutouts
as such (or any other kind of automatic cutout while driving). Any sudden cutoff is too dangerous. Our system gives us 8 levels of warning on the cell stress gauge (based on the most stressed cell) before
it starts to smoothly limit the power, as required to keep the stress below 8. Below 8 is stress but not dis-stress. But if you keep going, as I did, it will not wind the power down to zero, it will leave you with enough power to crawl up a 1 in 10 slope at 15 km/h (so long as the power is there to be had) and it will then display the distress levels from 8 to 15. And once it goes past 10 you will know that you are choosing to damage the battery because the battery will be screaming
at you (audible alarm).
I note that cell stress, in our terms, may be due to overvoltage, undervoltage, overtemperature and when charging or regen braking, undertemperature. And possibly in future, due to excessive voltage between a copper link and its bolts (high-R joints).
This run-flat experience (with individual cell data-logging) was extremely valuable in telling us where we should set these stress levels, and in encouraging us to get amp-hour counting going. [Edit:]
And indeed in proving that the whole scheme works, and gives a good balance between human protection and battery protection.
[Edit: Spelling and punctuation]
One of the fathers of MeXy the electric MX-5, along with Coulomb and Newton (Jeff Owen).