Range estimate is useless, I NEED to know how many KWh left.

[ezjam]

I live 1300 ft above sea level. I work at about 80 ft. above sea level. Since the range estimate is based on driving condition from immediate past, it is completely useless to me. Thirty miles-to-empty mean two completely different things if I just drove downhill or uphill.

In addition all else being equal, thirty-miles-to-empty is plenty on my way to work... but 30-miles-to-empty on my way home will not get me there. What I do know is that I need 4KWh to get to work and 8KWh to get home. I don't want a percentage number rounded to the nearest (or lowest) 10 mile range, I need to know KWh in battery banks. Now I KNOW this information is stored somewhere in myLink. However, I have looked high and low but cannot get to this information.

Has anyone been able to hack myLink to get to this information?

ezjam

ps. This is one of many features I totally dislike about my EV.
The others are:
- FOB is too big,
- head-rest is too far forward,
- touch screen controls are just an accident waiting to happen, (who thought this was a good idea?)
- window lock only locks the front windows
- 3.3KW charger (I do wish they had the 6.6KW charger option)
- Heating elements suck up too much juice. Cannot wait for model with heat pump.
- Blind spots
- Speakers are a joke
- Lack of cargo space (I knew this going into the deal though.)

 

[Pegasus]

What I do is change the display to the Energy Details screen. This shows how many kWh have been removed from the battery since the last full charge (and it takes into account energy regained from regen.) The 2014 has a total capacity of 21kWh, of which I believe 19 is useable (at 70 deg. F, less when colder.) At that point, just subtract.

As for your other concerns, there are threads that address most of them here, especially the speakers one. (Adding a small powered sub rounds it out nicely.)
I planned to hack my car to do at least 6.6kW charging, 9.9 if there's room, (using booster chargers from Metric Mind and tying them into the charger coolant loop,) but there have only been a handful of times in the past year where it would have been really useful (and I've been very busy with work) so I haven't researched that option much yet.

 

[TonyWilliams]

Use the center screen to display kWh consumed.

2014 - 19kWh usable
2015 - 18kWh usable

 

[xylhim]

Using the energy details screen is a little misleading in terms of how many kWh remain. This screen just tells you how much energy you have used, but it doesn't tell you how much energy is left in the battery pack, these two are similar but different. Different conditions will result in you being able to extract more or less energy out of your battery pack. I've seen over 19 kWh of energy used since last full charge (and no charging in-between) and still have had a little bit left in the pack. You're going to have to just make a determination of how much energy it typically takes to get from point A to point B, but of course, this will change depending on the weather and how you drive. 60 Miles round trip with ~1200ft elevation change doesn't sound bad though, you won't even need to charge at work as long as you start with a full charge at home.

 

[Nashco]

The Volt/Ampera guys already figured out the CAN message for kWh and can view it on their cars. Want to join in on the party and see if you can replicate it on the Spark EV?

http://www.mychevysparkev.com/forum/viewtopic.php?f=10&t=3926&start=0

Heck, I'd be satisfied with a percentage of SOC instead of the guessometer, even such a simple thing is currently buried on a phone app instead of the cluster! When I'm racing, the green goop and guessometer are both nearly useless.

Bryce

 

[aviatorjames]

I'm confused about why you think the range estimates are "useless"
I have found them to be quite accurate and VERY useful.

You're imagining some different (simplistic, single number) way to estimate range...
But the engineers have actually come up with a very effective system!

The important data are...
mi / kWh (bottom center in Trip page)
Instantaneous Power & Regeneration (right side DIC)
Range estimates & Trend in formation (left side DIC)
Historical data in 5 miles increments (History page on Electrical)

I'm not concerned with how many kWh are in the battery...
I only want to know how far it will take me and how far I have to go!
...pretty simple, and the data above is a great way to provide that info.

IMHO... it's more complicated that simply knowing how many kWh are in the battery.

 

[Nashco]

I'm confused about why you think the range estimates are "useless"
I have found them to be quite accurate and VERY useful.

TLDR: They're only accurate if your recent past usage matches your immediate future usage.

I agree that the range estimates are pretty darn good and useful in the Spark EV when you use the "confidence band" display that shows the estimated range remaining as well as a maximum and minimum confidence band. This is based on approximately 100 miles of historical efficiency data, and seemingly weighted relative to how recent the data is. In practicality, that filter is very useful for most normal driving scenarios.

Unfortunately, if you have a driving pattern with erratic behavior, this filter and historical reference actually hurts your ability to know what the battery has to offer. As an example, when I drove 100 miles of highway driving to get to a race and was mostly charged up, my range estimate said I had ~75 miles of estimated range, with 93 minimum and 62. After four hillclimbs (~18 miles including transits), my estimated range was 16, with a maximum of 19 and minimum of 14. So in 18 miles of driving I lost 59 miles of estimated range (and about 50% of the green battery goop)! Repeat again a second day with some transit, charging, etc. My very last run up the hill (8th time now) showed a minimum of 5 miles of estimated range, but I ran out of battery in about two miles. This is an extreme example of what the original poster is mentioning regarding their usage with hills sometimes and not others. When I autocross, I can go through almost a third of the battery before even getting to 5 miles showing up on the history page! I can (and do) reset most of these trackers for average mi/kWh before racing, but unfortunately we can't reset the guessometer...so it references a the immediate past which is known to be unique from the immediate future.

I had to do my own calculations using the "kWh since last charge" and estimating my rate of usage when racing. If the car displayed % SOC remaining (like it does on the phone app) instead of showing 3D green goop, I could do this a hell of a lot easier. Unfortunately, at this particular hillclimb I couldn't even get phone service to get % SOC the slow way (OnStar app).

The moral of my story is that remaining kWh and/or % SOC would be VERY useful if you have erratic usage profiles. Another useful one would be an instantaneous mi/kWh, so the user could compare instantaneous usage to average mi/kWh (which is what is being used for the range estimator). For comparison, an instantaneous MPG display is extremely common in cars...so why not have mi/kWh in an EV? Some EVs do. Also, the battery management system is constantly tracking remaining kWh and % SOC (and adjusting depending on conditions)...so why not make that information available for those of us abnormal users who need more information? Other EVs have made this available because it's critical information, and GM is not the only one to launch an EV without discrete % SOC displayed on the instrument cluster...many OEMs are learning the hard way that this is expected information for some users.

I'm trying to work on a solution for this, but honestly this is low priority for me compared to other stuff I'm working with on my CAN project (trying to figure out how to disable the stability control!). I'd be glad to contribute if somebody else wants to develop a hack to workaround this problem. I could probably get somebody about halfway there (identify hardware, software, and CAN ids), but the information still needs to be displayed somehow (other than using a laptop, of course).

Bryce

 

[aviatorjames]

Fair enough...
Obviously I didn't buy a Spark to do hill climbs and autocross!!
...sort of think all bets are off with the "erratic" driving!

It seems like you answered your own question with the numbers you provided.
Making a couple of assumptions about where you were fully charged...

You got about 3.75 mi/kWh on the highway portion.
Then you burned through 59 miles of range in 18 miles of racing.
...that 59 miles range estimate was calc'd on 3.75 mi/kWh history
So 18/59 is about 30% which means you should've expected 30% of the estimated range (at best!)
...that equals about 1.1 mi/kWh while you're racing (at best!)
(curious what you're showing after the race in mi/kWh)

Your driving is obviously out of the ordinary and you want something that the engineers probably didn't anticipate!!
But I stand by my assertion that the values generated and displayed are totally adequate & effective for mere mortals!

Reference "recent past" vs "immediate future" ...I respectfully disagree again...
My single-charge commute is 53 miles round trip with about a 900' climb over a ridge each direction.
The five mile historical blocks are averaged for the purpose of providing updated range estimates based on actual mi/kWh consumption.
...that's actually a very smart way to keep the range information relevant!
i.e. If I leave showing 80 miles range, I get home with 27miles left +/- a mile or two ...can't ask for anything more accurate than that!

Again, once you start the racing, all bets are off ...but probably really fun!!!
...post some pictures of what a Spark EV looks like doing a hill climb!

 

[Nashco]

Nissan launched the Leaf without % SOC, but it is now a standard feature. Nissan has had instantaneous energy feedback since day one. BMW launched without % SOC, but will be releasing new software soon with it. This is not unique to racers, I just provided that as a very extreme example to show how this logic fails in cases of inconsistent driving.

I can think of about five ways GM could give the driver a better option for those with inconsistent usage profiles. Unfortunately, they chose to KISS. Fine for most, but I want more. I really like the idea of the confidence confidence bands, just not the delayed response. The green goop is worthless and some more precise information would be much more useful.

Bryce

 

[genosensor]

The Spark's range estimator is uncannily accurate for me.
However, I completely agree that it fails miserably if your destination is at a very different altitude.

The "green goop" gauge is just an indication of the number of kWh left in the pack.
We could wish it had more resolution, but its definitely usable as a rough guide.
As others have pointed out, there are auxiliary screens that will show total kWh quite accurately,
but they are not as conveniently accessible.

Oh... The front headrests are very uncomfortable. Most probably, they'll save your life in an accident, but you'll be so distracted that they may make having that accident more likely

I was so desperate, that I replaced them with the rear headrests from our VW Jetta Mark IV.
Guess what... they fit well and are much more comfortable.
They don't go all the way down, however, unless you cut the supports. I'm 5' 9" and didn't need to bother with that.
Later, I found some nice black leather ones for about $50 on Ebay.
PM me if you want more info...

 

[solder]

I'm trying to work on a solution for this, but honestly this is low priority for me compared to other stuff I'm working with on my CAN project (trying to figure out how to disable the stability control!). I'd be glad to contribute if somebody else wants to develop a hack to workaround this problem. I could probably get somebody about halfway there (identify hardware, software, and CAN ids), but the information still needs to be displayed somehow (other than using a laptop, of course).

(Do you actually need to disable the stability control over CAN? The service manual says you should be able to press the traction control button once to turn off traction control, or hold down for five seconds to disable StabiliTrak (p5-8). Seems to work on my car.)

We have all of the information we need to get all this already with mode $22 diagnostic commands -- take for example

http://www.mynissanleaf.com/viewtopic.php?f=44&t=12561

The Dash Display is a small 3.9" x 1.9" x 0.6" blue OLED display (128x32) that connects to the OBD2 port on the Leaf under the dashboard.

The primary page shows Gids (of course) or a precise SOC (xx.xx%) or Gid% (xx.xx% Gid*0.356) or Ah (xx.xx Ah_CAP * SOC).
The primary page also displays real time current in or out of the pack graphically at the top of the display and battery pack voltage as a horizontal marker indicating 332-394V.
The primary page also displays the warmest of the 4 battery pack temperatures in tenths of a degree C or F. A small graph to the right of the battery temperature shows the relative temperatures of all 4 sensors just to give a feel for the relative distribution.

So we don't have Gids or Ah...-- the one thing we do have is "precise SOC" (7E4#03.22.43.AF -> high res state of charge, (1/25600 % per unit)).
Battery pack current is 7E4#03.22.43.56 (16 bit signed int, 1/9832 A per unit).
Overall voltage is 7E4#03.22.43.2d (unit is 0.52 V)
We have 6 battery pack temperature sensors, at least in my 2015 -- 7E7#03.22.40.D7/D9/DB/DD/DF/E1 (unfortunately, only accurate to 1 deg C)
We have 96(2015) or 112(2014) cell set voltages -- 7E7#03.22.41.81 - 03.22.42.50 (5/65536 V per unit)

Do we need anything else? Other than a display, a microcontroller (Teensy 3.1?), CAN PHY, 12v->5V DC-DC converter, OBD cable, fancy enclosure, a good place to mount it...oh and some firmware, but that's just a little bit of typing.

 

[Nashco]

(Do you actually need to disable the stability control over CAN? The service manual says you should be able to press the traction control button once to turn off traction control, or hold down for five seconds to disable StabiliTrak (p5-8). Seems to work on my car.)

Go 31 mph and you'll learn that it's a short lived disable. I've got some ideas, both analog and CAN manipulation possible, but no time to tinker lately.

Bryce

 

[solder]

Go 31 mph and you'll learn that it's a short lived disable. I've got some ideas, both analog and CAN manipulation possible, but no time to tinker lately.

You've probably already thought of this, but if you're willing to throw a code in the process, shorting the Traction Control switch signal to ground will give you DTC B2745 02 (p5-8), which will disable the stuff you don't want. Or so the manual claims. (Maybe it's sufficient to just hold the button down continuously? But there's text in there that implies it's not a simple switch, but really uses "serial data". LIN bus? Dunno.)

Anyway, back on this topic, I ordered a SPI 128x64 OLED display for $8 on eBay, I'll try wiring it up to a Teensy 3.1 and a CAN transceiver and see if I can make something pretty show up on the screen that vaguely reflects the battery state.

 

[JackHickey]

"... 60 Miles round trip with ~1200ft elevation change doesn't sound bad though, you won't even need to charge at work as long as you start with a full charge at home."

Bad advice. If you start with a full charge at home, you can not recover the stored energy with regenerative braking going downhill without overcharging the battery. I don't think the Spark will let you do that. Your disc brake pads will take the heat. A 3000 lb vehicle (Spark) has 1.36 KWH stored at an elevation of 1200 feet.

Tesla has the ability to set the maximum charge level. I have suggested to GM that they include this in a software upgrade for the Spark. It's a no-brainer. I invite others to let GM know you would like that feature.