|
|
Post by coulomb on Jun 20, 2021 20:44:35 GMT 11
My 2012, bought second hand 2 years ago, had a warranty replacement battery installed about a month before I bought it. It was in great condition, so I got a near-new car for about half the original sale price.
All went well until a month or two ago. It dropped to 11 health bars, and suddenly the estimated remaining range (GOM) was dropping ridiculously fast, often 10 km at once. After charging, the balance is good (e.g. 16 mV), and even under hard acceleration, all the cells appear to be at roughly the same voltage; no cells standing out from the rest. After charging to 80%, I lose 2 regen bubbles; one comes back after roughly 10 km, the last after very roughly 30 km. I don't know what my true range is, since the GOM drops so fast. I tried driving it to a low SOC and charged it back to 100%; no improvement. I had to charge it part way through that exercise, so I didn't get a reading of true range.
I've never noticed losing any power bubbles (at the right end of the power meter).
I wondered whether the HV fuse might have gone high resistance. I took out the disconnect today, and the copper on the fuse was somewhat tarnished, as if by heat. But I think copper does that over the years, unless coated with something. I put it on my current limited power supply at 3 A, and measured a solid, consistent 1.0 mV across the fuse with my Fluke 87 III. So that's about 0.33 mΩ, ±10%. That seems to be about right for a 225 A EV fuse when cold. I was hoping it might have read high, say several mΩ; that would possibly have explained the sudden drop in performance while the cell voltages remain reasonably constant.
My other faint hope was a dirty high current connection somewhere, perhaps a corroded plug or socket. I'll have a poke around to see if anything is obvious, but I don't hold much hope.
When I took out the disconnecter with the fuse, there was a moderate smell of battery electrolyte. I can still smell it faintly in my office. Lithium batteries do tend to emit that smell, but I suspect that it indicates at least some bad cells.
I may take up an offer to get the battery down and try to test some cells for internal resistance. Maybe some are high resistance, and somehow Leafspy isn't catching them (it does seem to have a fair lag when recording so many cells).
Suggestions welcome.
|
|
|
|
Post by EVangelist on Jun 20, 2021 22:15:13 GMT 11
My 2012, bought second hand 2 years ago, had a warranty replacement battery installed about a month before I bought it. It was in great condition, so I got a near-new car for about half the original sale price. All went well until a month or two ago. It dropped to 11 health bars, and suddenly the estimated remaining range (GOM) was dropping ridiculously fast, often 10 km at once. After charging, the balance is good (e.g. 16 mV), and even under hard acceleration, all the cells appear to be at roughly the same voltage; no cells standing out from the rest. After charging to 80%, I lose 2 regen bubbles; one comes back after roughly 10 km, the last after very roughly 30 km. I don't know what my true range is, since the GOM drops so fast. I tried driving it to a low SOC and charged it back to 100%; no improvement. I had to charge it part way through that exercise, so I didn't get a reading of true range. Well, the GOM is called that for a reason. Based on what you’ve written, I would not be concerned. Our GOM has always exhibited that kind of behavior - drops overly quickly at high charge states, and very slowly at low charge states. I rely more on how many bars of charge remain. That seems to be more predictable. |
|
|
|
Post by coulomb on Jun 21, 2021 11:30:51 GMT 11
Well, the GOM is called that for a reason. Based on what you’ve written, I would not be concerned. Our GOM has always exhibited that kind of behavior - drops overly quickly at high charge states, and very slowly at low charge states. Mine was generally pretty good, before the change. It's radically worse now. I agree with that, but the bars are dropping far too quickly now as well. Often now two bars in a short (5 km) run to the shops. Usually I get 10 km per bar. |
|
|
|
Post by rusdy on Jun 21, 2021 11:55:08 GMT 11
... After charging to 80%, I lose 2 regen bubbles; one comes back after roughly 10 km, the last after very roughly 30 km. Hi coulomb , That is normal as the battery ages (yes, our gen-1 battery ages way too quickly, even the 'new' replaced ones). My guess, it's due to the rapidly declining Hx (not SoH). I've noticed this when I first had firmware update in 2016 (the blip in my SoH chart), the Hx tanked by 15%, and I've lost regen bubble immediately. Same with my 'new' replacement battery, my regen bubble gone after a few months only (at 80% charged). My Hx already dropped below 80% at the moment (less than a year of newly replaced battery). ... I don't know what my true range is, since the GOM drops so fast. I tried driving it to a low SOC ... Did you notice the GoM suddenly stop dropping and slows down significantly at lower SoC (just before LBW, i.e. first low battery warning)? If yes, then this is a typical gen-1 battery issue. I've documented this thoroughly ( click here) and concluded that Leaf's BMS has struggled to accurately measure the SoC throughout (non linear behaviour). If you run this experiment yourself, you'll most likely find the same thing, i.e. discrepancy between kWh consumed, reported by the dash, VS the GoM (which is also what LeafSpy uses). The easiest to run this experiment is to charge your car at home from completely depleted (lower than VLBW) to fully charged while LeafSpyPro running (showing charging details). You'll quickly see the SoC has a weird non-linear behaviour despite being charged by constant power. Conveniently, this method gives you the true kWh capacity of your battery as well. I tried driving it to a low SOC and charged it back to 100%; no improvement. I had to charge it part way through that exercise, so I didn't get a reading of true range. ... Suggestions welcome. The only reliable way I've found so far is to drive it to as low as possible (beyond the VLBW, i.e. Very Low Battery Warning, when the GoM shows no number anymore), without charging. That'll be your true range. The LeafSpy would've reported imbalance greater than 100mV by then. I think you're going too far investigating the fuse  . Rusdy |
|
|
|
Post by coulomb on Jun 28, 2021 21:38:59 GMT 11
The only reliable way I've found so far is to drive it to as low as possible (beyond the VLBW, i.e. Very Low Battery Warning, when the GoM shows no number anymore), without charging. That'll be your true range. Good idea. I've done that now; 77 km (starting with an "80%" charge). I can live with that for now. Well, 102 mV  I've done some logs and graphs; I've decided that it's a pretty standard distribution of capacities and just suddenly high internal resistance. Interesting that there's a group of lower voltage cells from about 10-46. As far as I can tell, these are all in one row at the back, towards the left side. Maybe they got hotter there? It seems that module m contains cells 2m-1 and 2m. Layout diagram from Dala, this post. Edit: I note that the back seat stack of cells is close to the battery fuse/disconnect, which had the strong smell of electrolyte solvent when I removed the fuse. Hardly strong evidence, but I read somewhere that the large stack in the rear usually has the worst cells, since the heat there can't dissipate as well; the middle cells are surrounded by many other cells. _  Attachments:
|
|
|
|
Post by rusdy on Jun 30, 2021 15:40:11 GMT 11
Good idea. I've done that now; 77 km (starting with an "80%" charge). I can live with that for now. Hmmm... from extremely limited samples (yours and mine), looks like the replacement battery is indeed a 'canary' type, which has ~4% degradation per year for Australian condition. Early warranty replacement (such as Feng 's one) must be the lucky newer chemistry types. Me sad... Interesting that there's a group of lower voltage cells from about 10-46. As far as I can tell, these are all in one row at the back, towards the left side. Maybe they got hotter there? Yours have same pattern like mine at VLBW. Must be a standard new battery replacement profile! |
|
|
|
Post by 4wardthinking on Aug 23, 2021 12:00:43 GMT 11
Bear in mind that the policy of cobbling together used cells is a problem.
My guess, looking at the symptoms is there is a few cells that passed certain tests, but the true test is capacity.
Voltage is relatively irrelevant in a large, high capacity battery, it’s only capacity that counts here.
There are financial reasons etc for real-using cells that can be used, and it’s looking like you have well used cells that passed certain criteria, unfortunately, capacity after reconstruction of the battery has degraded, or it was hoped it may be good enough.
If you can smell the sweet smelling electrolyte, there is a pressure leak in there.
Once this is open, the cell will loose it’s ability, for many reasons.
My advice, bounce it back, tell them to replace, not refurb.
I believe that if you stand your ground, and clatter a few sabres, maybe prod Nissan outside of Au, imply that it’s ‘good, that the world doesn’t know that refurbishment doesn’t actually replace a good, battery with original capacity, etc’, you may open their eyes.
Quite clearly, the capacity of your battery is way below production standards, and that if the vehicles specification.
I do know that the previously called tech will invariably not be aware of how the battery works that carried out the refurbishment, only how to do from a paper sense.
He/she will be convinced that they did a good job, and charged accordingly, sadly, the way a battery actually works is still a probable mystery.
|
|
|
|
Post by coulomb on Dec 4, 2021 19:13:02 GMT 11
I'm starting to wonder if there are loose wires in the BMS. Based on how the estimated range can go down quickly, but go up just as quickly as well. I also find that the cell voltages jump around a bit in Leafspy. This idea is slightly strengthened by by wife's anecdote that the trouble started after a bad scrape over a traffic island in a car park.
I have vague ideas about dropping the battery and inspecting the wiring, but that would require me to find space in the garage for extended work on it. The car currently lives in a carport, and junk lives in the garage. I suspect I'm not the only one with that arrangement.
|
|
|
|
Post by rusdy on Dec 4, 2021 23:36:30 GMT 11
" ... Based on how the estimated range can go down quickly, but go up just as quickly as well. I also find that the cell voltages jump around a bit in Leafspy. How bad is the jumping? from my experience, when the Hx goes down (like 60% and lower), the estimated range does drop quickly (and up) at higher SoC (like 50% and above). Then it slows down after. Also, I'm guessing you don't use your leaf that much? What SoC when you leave it parked? |
|
|
|
Post by coulomb on Dec 5, 2021 11:33:56 GMT 11
Often 10, sometimes 20 km jumps. It's just under 67% Hx now. So I guess it's getting pretty close to that range. Not for longer journeys, that's for sure. But it's still the #1 vehicle for most shopping, appointments, etc. So twice a day for 3-20 km runs. Sometimes with charging in between. I always charge it to 80%, once every few months to 100% and driven soon after. But with the wet weather lately, it's not been charged as often, so it is more likely to sit at 6-8 bars of SoC. |
|
|
|
Post by rusdy on Dec 5, 2021 21:12:38 GMT 11
Often 10, sometimes 20 km jumps. Ah, got you now, I thought the 'jumping' was related to the cells imbalance. I always have a 'range jump' far worse than that at higher end (i.e. at 70 ~ 80%SoC) even with my 'new' battery. Considering you have a regular small run, the GoM is doing pretty well . So, that's normal GoM behaviour in my opinion.
Maybe the Leaf GoM is by far the worst compared to all other EVs... |
|
.
