Better (reverse) SOC meter already in the car?

My Nissan Leaf Forum

Help Support My Nissan Leaf Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

arnolddeleon

Well-known member
Joined
Apr 21, 2010
Messages
125
Location
Cupertino, CA
At end BayLEAFs gathering today someone shared a potentially awesome tip on a better (reverse) SOC meter that is already built-in to the car and can get read from the dash.

The idea is simply use the 110V time to 100% charge display. It is definitely a big improvement over the bars in that you get 1/2 hour steps it there are that are at least 22 steps (he said there are 24). It it turns to be 24 steps that means 4% steps, that would a huge improvement over the bars. I would be quite happy to get 2% steps (of course more steps would be better but the precision in the SOC may not be there anyway).

I have not heard about this tip in any of the discussion on the SOC meter (any my attempts to search for it didn't turn up any matches). I didn't get the person's name to give him proper credit.

arnold
 
I have tried this, but it is not all that easy or it seemed to me accurate. I use L1 to charge most of the time, and the reading seems padded so that the car is charged before the time estimated. Was this someone that knew how the time was determined, or was this just another driver?
 
This has occurred to me also. I figure 120v charging gets you basically 1kw, so the 120v time-to-charge could be a decent indicator of how many kwh you've used. I don't have Gary's SOC kit, but it would be awesome if someone who does could post some data on how the time-to-charge display correlates to GIDs.
 
arnolddeleon said:
you get 1/2 hour steps it there are that are at least 22 steps (he said there are 24)
With 1/2 hour increments and a 20+ hour time to charge from empty shouldn't that be at least 40 steps? Or does it start going up in hour increments after some point? I admit I haven't paid close enough attention to it to notice so far. Even if it does, you still have ~2% steps at first at least. Although I guess it's when you're running close to empty that you really want the finer granularity...
 
fooljoe said:
arnolddeleon said:
you get 1/2 hour steps it there are that are at least 22 steps (he said there are 24)
With 1/2 hour increments and a 20+ hour time to charge from empty shouldn't that be at least 40 steps? Or does it start going up in hour increments after some point? I admit I haven't paid close enough attention to it to notice so far. Even if it does, you still have ~2% steps at first at least. Although I guess it's when you're running close to empty that you really want the finer granularity...

Duh! :-0 Of course! If it's accurate at all this makes it an even better find.

arnold
 
Caracalover said:
I have tried this, but it is not all that easy or it seemed to me accurate. I use L1 to charge most of the time, and the reading seems padded so that the car is charged before the time estimated. Was this someone that knew how the time was determined, or was this just another driver?

The padding won't matter as long as it is consistent. What matters most is a consistent indicator of SOC. I obviously haven't used technique before but I'm going to start trying it. I don't know the qualifications of the person that suggested it although I get the impression he has used it.

arnold
 
arnolddeleon said:
I have not heard about this tip in any of the discussion on the SOC meter (any my attempts to search for it didn't turn up any matches). I didn't get the person's name to give me proper credit.
Arnold, thank you for starting the thread. Glad to see that the word got out and there is some interest, and I would hazard a guess that you spoke to Tony. I've been using this method on a daily basis for about a month now and I think that it has some potential. I reached out to TonyWilliams, Stoaty and gascant early last month to inquire if this has been attempted before. I remember seeing dashboard pictures with the charging display being prominently shown, and it looked like an intriguing possibility. Even or perhaps because it was not specifically discussed on the forum. I was hoping that we could gather more data and validate this approach using Gary's SOC meter.

I could be wrong, but I believe that 1 hour on the 120V charging time display corresponds to about 0.8 kWh. This is likely due to the fact that Nissan did not recalibrate its algorithm from Japan (100V) to the US (120V). I reported this on Phil's spreadsheet of issues for the Nissan engineering team as well.

In any case, if you turn off your charge timer, the low battery warning will always come on at the 21-hour mark. If you are using an 80% timer, it will show up at the 17-hour mark. The difference between 80 and 100% charge seems to be exactly 4 hours, which is a bit less than expected. Turtle should come on around the 24-hour mark, although I have seen at 22 hours once. This was likely circumstantial, since it happened on a steep incline.

The charging time display has much better granularity than battery gauge bars. In the upper range, it shows not only full hours, but also half-hours. This means that it's about three times better than battery bars, and we can estimate the SOC in 400 Watt increments. That's the extent of what I know, and I believe this could be a very useful SOC meter proxy once we have collectively gathered more data.

George
 
surfingslovak said:
I could be wrong, but I believe that 1 hour on the 120V charging time display corresponds to about 0.8 kWh. This is likely due to the fact that Nissan did not recalibrate its algorithm from Japan (100V) to the US (120V). I reported this on Phil's spreadsheet of issues for the Nissan engineering team as well.

In any case, if you turn off your charge timer, the low battery warning will always come on at the 21-hour mark. If you are using an 80% timer, it will show up at the 17-hour mark. The difference between 80 and 100% charge seems to be exactly 4 hours, which is a bit less than expected. Turtle should come on around the 24-hour mark, although I have seen at 22 hours once. This was likely circumstantial, since it happened on a steep incline.

The charging time display has much better granularity than battery gauge bars. In the upper range, it shows not only full hours, but also half-hours. This means that it's about three times better than battery bars, and we can estimate the SOC in 400 Watt increments. That's the extent of what I know, and I believe this could be a very useful SOC meter proxy once we have collectively gathered more data.

George

0.8kWh sounds about right for an hour of charging at 100V. There seems to about 200 watts of car overhead/inefficiency when the car is charging leaving 800 watts for the car. If they had reclalibrated for 120V we would get a nice round kWh. Does the car switch to only 1 hour increments at lower SOC? That would be a bummer because that's where you need more information.

From what I gathered at the meeting the SOC is at best a guess when the car is running and under load so getting and displaying very precise numbers is going to near impossible but I'm ok with that. I would be quite happy to have any gauge that is repeatable and accurate to with a 2 to 3% (N.B. I'm trying to use precise and accurate carefully).

The question that driver needs to answer is "will I make it". The current reality is the battery pack doesn't store enough energy to make this a moot question. This leads to the fundamental need for a gauge that behaves in a somewhat consistent manner so the driver can *learn* what *their* range is. I suspect most drivers don't think of trips in miles (until they start driving EVs). What I've learned in almost 10 years of driving a RAV4-EV is that I need a certain level of SOC from certain locations to get home. I could plan future excursions from those known points. BTW I think the DTE (Distance To Empty) indicator is a useful addition, it is just not sufficient. The unfortunate mistake was to make it so prominent and not give us a percent SOC meter that we can read as easily.

arnold
 
arnolddeleon said:
0.8kWh sounds about right for an hour of charging at 100V. There seems to about 200 watts of car overhead/inefficiency when the car is charging leaving 800 watts for the car. If they had reclalibrated for 120V we would get a nice round kWh. Does the car switch to only 1 hour increments at lower SOC? That would be a bummer because that's where you need more information.
I drove from full to turtle today to confirm couple of key metrics, but I'm too tired to post all the data I've gathered. What I can say with some certainty is that 1 hour on the 120V charging display corresponds to about 840 Wh. The display shows half hour increments in the upper part of the range, but they are not displayed in the bottom half (below 12:00).

arnolddeleon said:
From what I gathered at the meeting the SOC is at best a guess when the car is running and under load so getting and displaying very precise numbers is going to near impossible but I'm ok with that. I would be quite happy to have any gauge that is repeatable and accurate to with a 2 to 3% (N.B. I'm trying to use precise and accurate carefully).
Based on my turtle experience last month, I had a hunch that Nissan used pack voltage to estimate SOC values. If I recall Kadota-san's slide correctly, he mentioned that in his presentation as well. As GregH said couple of months ago, pack voltage can plummet rather easily when the battery is below the knee, which roughly corresponds to the very low battery warning. In my experience, sharp pack voltage drop can trigger premature onset of power limited mode. I would be careful with the accelerator pedal and either avoid going uphill or at least slow down considerably on inclines after the LBW and particularly after the VLBW.

arnolddeleon said:
The question that driver needs to answer is "will I make it". The current reality is the battery pack doesn't store enough energy to make this a moot question. This leads to the fundamental need for a gauge that behaves in a somewhat consistent manner so the driver can *learn* what *their* range is. I suspect most drivers don't think of trips in miles (until they start driving EVs). What I've learned in almost 10 years of driving a RAV4-EV is that I need a certain level of SOC from certain locations to get home. I could plan future excursions from those known points. BTW I think the DTE (Distance To Empty) indicator is a useful addition, it is just not sufficient. The unfortunate mistake was to make it so prominent and not give us a percent SOC meter that we can read as easily.
Yes, I agree completely. To Nissan's credit, they tried to make the DTE indicator easy to use. However, it would have been a prudent design decision to include a simple SOC meter display option as a fallback. Just in case their rather simplistic DTE algorithm should prove to be inadequate in the field.

Based on my trip today, I'm pretty confident that the rock bottom on the 120V indicator is 25 hours. Turtle mode is getting very close when the 240V indicator ticks over from 7:00 to 7:30 hours. The two indicators (120V and 240V) on the charging time display don't move in lockstep, they appear to be calculated separately.
 
From what I heard, pack voltage is not a good measure of "fullness" while driving because the load affects it.
If you stop the car, and hook up to charge it can get a "good reading" and re-orient. But if you are doing a long drive, it may have to do some extrapolation and 'guestimation' of 'fullness' since the voltage is being affected by the load on the pack. Maybe I don't have the details exactly right, but I think the point is that there is no "easy" way to just get one quick reading of accurate pack SOC. Just layers on top of layers of estimations.
 
TEG said:
From what I heard, pack voltage is not a good measure of "fullness" while driving because the load affects it.
If you stop the car, and hook up to charge it can get a "good reading" and re-orient. But if you are doing a long drive, it may have to do some extrapolation and 'guestimation' of 'fullness' since the voltage is being affected by the load on the pack. Maybe I don't have the details exactly right, but I think the point is that there is no "easy" way to just get one quick reading of accurate pack SOC. Just layers on top of layers of estimations.
What about (as I noted in another thread) shutting the car down? Is it necessary to charge to re-orient? Or is re-booting the car sufficient?

Should be an easy experiment, aside from needing to run it down close to empty with a varying discharge curve in an attempt to first DIS-orient it -- then shut down and reboot and see if the DTE changes from the number displayed before the reboot.
 
From my trip today. Time to 100% Charge on 120V in relation to Gids:

2 hours- 253 Gids
2.5 hours- 248
3 hours- 240
3.5 hours- 234
4 hours- 228
4.5 hours- 221. Return trip- 224
5 hours- 216. Return trip- 219
5.5 hours- 210. Return trip- 211
6 hours- 203. Return trip-205
6.5 hours- 197. Return trip- 198
7 hours- 191
7.5 hours- 186. Return trip-185
8 hours- 179
8.5 hours- 172. Return trip- 175
9 hours- 168. Return trip- 169
9.5 hours- 160. Return trip- 161
10 hours- 155
Not sure if this was at 10.5, 11 or 11.5 hours- 150
Next was definitely:
12 hours- 140
13 hours- 122
14 hours- 109
15 hours- 99
16 hours- 87
17 hours- 79
18 hours- 65
 
surfingslovak said:
In any case, if you turn off your charge timer, the low battery warning will always come on at the 21-hour mark. If you are using an 80% timer, it will show up at the 17-hour mark. The difference between 80 and 100% charge seems to be exactly 4 hours, which is a bit less than expected. Turtle should come on around the 24-hour mark.....


This is about 10 Gid, 3.3 miles into VLB.


44d2de53.jpg
 
TonyWilliams said:
This is about 10 Gid, 3.3 miles into VLB.
<img>
Tony, why is your 240V time-to-charge reading so similar to your 120V reading? In my car the 240V reading is about 37.5% of the 120V reading - which is to be expected, because at 16A and 240V you get enough additional power to charge in 37.5% of the time that it would take at 12A and 120V. So the display in my car would show, to 80% charge, 240V 6 hrs, 120V 16 hrs. For that matter, a full charge at 240V only takes 7 hours. So why is your car saying 11 hrs to just 80%?
 
johnr said:
TonyWilliams said:
This is about 10 Gid, 3.3 miles into VLB.
<img>
.....a full charge at 240V only takes 7 hours. So why is your car saying 11 hrs to just 80%?

I don't have any idea. That's why I posted it. It doesn't meet the proposed expectations to determine SOC. This picture was taken early, early Sunday morning.
 
91040 said:
From my trip today. Time to 100% Charge on 120V in relation to Gids:

2 hours- 253 Gids
2.5 hours- 248
3 hours- 240
3.5 hours- 234
4 hours- 228
4.5 hours- 221. Return trip- 224
5 hours- 216. Return trip- 219
5.5 hours- 210. Return trip- 211
6 hours- 203. Return trip-205
6.5 hours- 197. Return trip- 198
7 hours- 191
7.5 hours- 186. Return trip-185
8 hours- 179
8.5 hours- 172. Return trip- 175
9 hours- 168. Return trip- 169
9.5 hours- 160. Return trip- 161
10 hours- 155
Not sure if this was at 10.5, 11 or 11.5 hours- 150
Next was definitely:
12 hours- 140
13 hours- 122
14 hours- 109
15 hours- 99
16 hours- 87
17 hours- 79
18 hours- 65

Those data are beautifully linear. In fact R = -0.997!

You get Gids using the following formula:

Gids = hours (to 100% charge at 120V) x -11.65 + 273.2

The linear fit thus estimates maximum Gids to be 273 and charge time from empty to be 23.4 hours.

P.S. I took all data points, except 10.5-11.5. I also assumed that on the return trip the 7 and 8 hour data points were identical to those during the first trip.
 
lukati said:
You get Gids using the following formula: Gids = hours (to 100% charge at 120V) x -11.65 + 273.2

If you force the fit through 0/281, assuming 281 is equivalent to 100% SOC, you get:

Gids = hours (to 100% charge at 120V) x -12.42 + 281

From this, SOC can be approximated simply by calculating

SOC = 100 - 4.42 x hours (to 100% charge at 120V)
 
fooljoe said:
11 hours for 80% on 240v? Tony, I think your car's broken.
The time to charge auto-adjusts depending on the available supply last time you plugged in. Using my Rev. 1 modified EVSE which supplies only 12A at 240V, the reported time to charge at 240V is significantly longer. Even with "standard" L2 charging, there is likely to be some variation.

Tony, what were you last plugged into before taking that photo?
 
Back
Top