Dumb**s alert: am I the first to drain it??

[AndyH]

2. Do not plan a trip where you will dip below 20 miles left. Besides being good for your stress level, this fits with the recommended Depth of Discharge of 20%. We really should not be driving these cars into the "dash"-mode or turtle mode on a regular basis.

Josh, will you please help me understand what you mean when you recommend one not drives to turtle or why you think there's significance to a 20% DoD?

Thanks!

 

[garygid]

Is the LEAF's SOC estimate derived from the battery PACK voltage,
or from the measured voltage of the LOWEST-voltage cell-pair?

 

[AndyH]

Is the LEAF's SOC estimate derived from the battery PACK voltage,
or from the measured voltage of the LOWEST-voltage cell-pair?

You've asked that before, Gary, and the question doesn't make sense to me.

The car measures current in and out so can track SOC directly. Voltage is second order and can be a useful cross-check, sure. We know the computers monitor cell voltages and keep us from going outside the safe zone (not the cell limits but the allowable zone!)

If the car top balances the cells, and if the vertical bars are cell capacity, isn't this what we've got? And since lithium can be top balanced, bottom balanced, or balanced anywhere in the middle, as long as the car keeps us away from the ragged ends, why do we care? Especially since we remember from the service manual that the #1 job of the management controller is to protect the cells - not to get us safely home but to protect the battery from US!

{PS - the known info is in the service manual, section EVB, EVC, and the CAN chapter... }

 

[garygid]

I am aware that many things CAN be done, and MIGHT be done.
However, I am more interested in what IS DONE by the systems IN the LEAF.

If current-counting is done to estimate SOC, then one cell with (for example) a high self-discharge would get to the low-voltage limit first, and (perhaps rather unexpectedly) stop the car, even when the car's SOC "calculation" was estimating sufficient remaining charge to go another 10 or 20 miles (under the present driving conditions).

 

[AndyH]

I am aware that many things CAN be done, and MIGHT be done.
However, I am more interested in what IS DONE by the systems IN the LEAF.

If current-counting is done to estimate SOC, then one cell with (for example) a high self-discharge would get to the low-voltage limit first, and (perhaps rather unexpectedly) stop the car, even when the car's SOC "calculation" was estimating sufficient remaining charge to go another 10 or 20 miles (under the present driving conditions).

Ok. Thanks for the help here.

If I understand your proposed scenario correctly, you're suggesting that if the car only uses current counting to estimate state of charge (SOC) that it could miss a significantly failed cell which in turn could result in an erroneous range display and potentially leave a driver stranded on the side of the road?

A number of members have talked about current counting, cross checking with cell voltage, pack models, and sloped discharge curves. It's pretty clear that the car does not rely on 'ONLY' current counting. That appears to stop your chain of events before it starts. Moving on, though...

We know from the tech info direct from Nissan that the car directly measures cell voltage. We know it can do that 24/7 in all driving, parking, or charging conditions. We also know that the car does a diagnostic to measure cell internal resistance - which it stores in the battery controller in the pack. We also know that data from the battery controller travels via CAN bus to the main vehicle controller - and that the vehicle controller splats data on the driver's displays and also decides if it can honor a driver's request to go or stop or charge.

So far so good.

We also know that the car uses the cell voltages to decide if the battery is in trouble. If any cell (parallel group - two per module) rises above or falls below the safe zone (not the absolute cell limit - but the operator safe zone!), it triggers a "no charge/no drive" hard fault and the car will not move.

(Important points here - the car will not allow one to over charge or over discharge any cell. It could happen but it would take a controller failure - and the controllers are monitored as well - and a controller failure stops the car - which should also prevent over discharge or over charge. Belt and suspenders. Redundancy. Good Things.)

Your proposed chain of events breaks a second time because a 'high self discharge' is a direct indication of a failed cell. A cell with a high self discharge has an internal short or some other significant damage - and its internal resistance and thus cell degradation value will change significantly. This doesn't happen quickly or instantly or catastrophically unless someone shoots the battery box, so we'll ignore that situation for the moment. The car tracks internal resistance at the cell level. Well before a cell gets bad enough to leave a driver stranded, the computer will set error codes and alert the driver of the battery failure.

Therefore no - it's my opinion that the chain you propose cannot happen in the real world.

 

[Ingineer]

I agree with Andy, there's no way something like that could "slip by".

The internal resistance calculations are done by watching dV (voltage change) under a load. The battery ECU would see the internal resistance go high as the first indication of a fault and you'd get christmas on your dash.

There well may be bugs in the range calculation, and I for one wish they'd simply ditch that and show you SOC, or at the very least at least let you see SOC as well.

-Phil

 

[garygid]

So, in your view(s) there is no way the SOC calculation, as ACTUALLY DONE in the LEAF, could create a misleading too-high SOC number?

If I KNEW (but I do not) that the SOC calculation actually uses the lowest cell-pair voltage properly, I would be tempted to agree.

But, if the SOC calculation uses only the Pack charge-counting, I will have to disagree until I learn more.

Also, if there is no SOC-calculation problem, what is the Nissan software update (Range & SOC related?) likely to "fix" or improve?

 

[Ingineer]

So, in your view(s) there is no way the SOC calculation, as ACTUALLY DONE in the LEAF, could create a misleading too-high SOC number?

If I KNEW (but I do not) that the SOC calculation actually uses the lowest cell-pair voltage properly, I would be tempted to agree.

But, if the SOC calculation uses only the Pack charge-counting, I will have to disagree until I learn more.

Also, if there is no SOC-calculation problem, what is the Nissan software update (Range & SOC related?) likely to "fix" or improve?

Since you don't have a copy of the source code for all ECU's, you will never know or be able to confirm any of this.

You also have no idea what the real reported SOC is unless you use Consult III+.

It's clear to me that the SOC calculation is based on the calculated amp-hour capacity, but I cannot PROVE anything until they hand me the source code. The calculated AH is not likely to change abruptly unless there is a fault, so this number will be very slow to change in normal use.

You can disagree all you want, and I'll be happy to disagree with your disagreement! =)

-Phil

 

[AndyH]

So, in your view(s) there is no way the SOC calculation, as ACTUALLY DONE in the LEAF, could create a misleading too-high SOC number?

no.

If I KNEW (but I do not) that the SOC calculation actually uses the lowest cell-pair voltage properly, I would be tempted to agree.

Use your mad skillz and contact Japan. Or open the service manual. Or both.

But, if the SOC calculation uses only the Pack charge-counting, I will have to disagree until I learn more.

Why do you even suggest that it would?!

Also, if there is no SOC-calculation problem, what is the Nissan software update (Range & SOC related?) likely to "fix" or improve?

Because the range display 'problem' isn't about SOC - it's about using the SOC to make a range estimate. The SOC number is an input to the rang guestimate, but it also feeds the 'fuel gauge' and that's working fine, right? That might be a clue...

Gary - I feel very very good about the way the car manages the pack because: 1. I've read the service manuals for the Leaf (as well as the books for the S10EV and the Ranger EV) 2. I've got some time with battery test equipment on the bench and on the road, and 3. I've experienced lithium cell degradation on the road.

I ran for 11 months with a failing cell on my bike - I felt the slow decay in top speed and performance, and measured the slow decrease in range. And I was able to nurse the failing cell because I had a BMS on board that protected the cells! Once it hit my 'I gotta do something about this' threshold, I pulled the cells and tested them on the bench to see what happened to the bad cell.

The Leaf will not be as affected by a single bad cell as there are two in parallel in the car's pack - I have only a series string in my bike. The Leaf has two probably 38-40Ah cells in parallel. If one starts to fail (internal shorts, damage, overheat, gunshot wound) the other will pick up the slack. Yes - range will drop but it'll only go down about 1/4 rather than the 1/2 if there was only one cell in the string. And yes - the dash will light up in a big-dog way!

We have to stop thinking about the car from a DIY lead-acid perspective, and from a no-BMS Chinese import perspective, and from a DIY electric car perspective. We simply do not have to worry about or second guess the car's systems. This car is a fully engineered electric vehicle that has fully redundant systems designed to protect the battery pack from the nut behind the wheel. Period. Thank God. I'm getting sick of temporary add-on BMSs and tangles of wires all over my garage when I want to balance my pack. I am SO looking forward to 1. Park 2. Plug-in and 3. Walk away. :lol:

 

[AndyH]

I am aware that many things CAN be done, and MIGHT be done...

Here's an example of what's done today. Since this is an LG/Chem presentation that describes how their current cells are managed, and since it's got the CPI tag - the Volt supplier...you can do the math, I'll bet.

I would be very, very surprised if NEC/Nissan didn't have their own version of this.

In the same way multispectral imaging provides much more info than radar or IR or photo provides by itself, here's a look at 'multispectral' current determination...sorta.

Originally posted 26 May 2010.
http://www.mynissanleaf.com/viewtopic.php?f=9&t=456&start=0
http://www1.eere.energy.gov/vehiclesandfuels/pdfs/merit_review_2009/energy_storage/es_07_alamgir.pdf

 

[jstack6]

I just got my new LEAF 2 days ago, I drove 88 miles Sunday 3-20 and still had 41 miles after. That would be about 120-130 mile range. No Air or heater was needed, that can take about 10% . It all depnds on the loose nut behind the wheel. I like to pay attention and drive like I care. In a prius I have even had 84 mpg on a round trip.

Be sure you always include a full round trip so you don't have a bad up hill result or super mostly downhill result. Make sure you know what mods the owner has done. A local pirus owner got new wide low profile tires and his MPG droped below 40. After we convinced him to switch back it jumped to 45+.

 

[JRP3]

Congrats on the LEAF, and the range. What sort of speeds were you doing?

 

[jstack6]

I drove on all city streets right with traffic from 30-50 mph. Lots of lights and pretty level. It was round trip so any up or downhills along with extra wind is averaged out. If I really hypermiled I could even get more.

Today my wife drive 12 miles home from work. It showed 88 mile range when she started and had 86 when she pulled in at home. Similar on the wayo to work.That's also city streets with average traffic and very level areas in Tempe, Chandler AZ