2015 Spark EV Battery Capacity from GDS2

Chevy Spark EV Forum

Help Support Chevy Spark EV Forum:

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

CCIE

Well-known member
Joined
Nov 22, 2016
Messages
174
Location
CT
I recently purchased a VCX Nano MDI (OBD) Interface that included GDS2 software. I got it to access GM TIS2WEB SPS so I could fix an issue with my Volt. But, the GDS2 was a nice bonus.

Anyway, I connected it to my 2015 Spark EV with 11,000 miles to see what I could see. One of the HPCM2 data readouts is actually labeled "Hybrid/EV Battery Pack Capacity." I looked up that label and it was referenced on gm-volt.com as a hex value representing the Amp-hour capacity of the battery with precision to 1/10th. In my case the hex value was 0209. That translates to 521, or 52.1Ah. The pack voltage on the same screen showed 395.7. Multiplying those two numbers and dividing by 1000 equals 20.61KWh.

I'm not sure if that number is total pack capacity, or usable. But either way, it's above what most people believe the total capacity of the 2015/16 LG battery is.

I took a screen cap, but don't see a way to insert it here without hosting it somewhere.

EDIT: As is stated below, nominal voltage (355V for our cars) must be used to calculate capacity. So, this should be 52.1 x 355 = 18.5KWh
 
I cross posted this at gm-Volt.com and was able to insert the pic there.

http://gm-volt.com/forum/showthread.php?294313-2015-Spark-EV-Battery-Capacity-from-GDS2&p=4154193#post4154193
 
CCIE said:
not sure if that number is total pack capacity, or usable. But either way, it's above what most people believe the total capacity of the 2015/16 LG battery is.
Click to expand...

Given the "usable" capacity is a Software limitation, I would think this measured the Total battery capacity, and if so, it is slightly above the theoretical total capacity (according to the article below)
https://cleantechnica.com/2016/09/03/chevy-bolt-chevy-spark-ev-battery-packs-compare-one-another/

If so, it also does not contradict that the usable capacity is estimated to be around 18.4 / 19 kWh, which is unrelated to the total capacity.
 
scrambler said:
CCIE said:
not sure if that number is total pack capacity, or usable. But either way, it's above what most people believe the total capacity of the 2015/16 LG battery is.
Click to expand...

Given the "usable" capacity is a Software limitation, I would think this measured the Total battery capacity, and if so, it is slightly above the theoretical total capacity (according to the article below)
https://cleantechnica.com/2016/09/03/chevy-bolt-chevy-spark-ev-battery-packs-compare-one-another/

If so, it also does not contradict that the usable capacity is estimated to be around 18.4 / 19 kWh, which is unrelated to the total capacity.
Click to expand...

Assuming I lost a bit of capacity in 11k miles, it seems like that number would have started around 21kwh. That's the same size as the 2014. I've never heard anyone claim the 2015/16 has a battery that big. Most people seem to think it's around 19 and has very little buffer.

Maybe they just lowered the usable capacity in the 2015/16 battery? The 2014s seem to start with more range and quickly lose it. So maybe GM saw that and decided to be a bit more conservative with the LG battery. That would explain why they start with less range but don't seem to lose it as fast.
 
Capacity is specified at nominal voltage, not peak voltage. SparkEV cells are 3.7V nominal and 96 in series, which gives 355V. 52 Ah is 18.47 kWh.

When charged to full, cell voltage is about 4.2 volts, 96 is 403V. 52 Ah is 21 kWh. But cell voltage can drop to as low as 3.2V (not sure exact value), 96 is 307V. 52 Ah is 16 kWh. That's why nominal value is more meaningful.

One thing to keep in mind is that load conditions will change Ah rating. If pushed too hard to too light, Ah will be less, which means kWh will be less, too.
 
SparkevBlogspot said:
Capacity is specified at nominal voltage, not peak voltage. SparkEV cells are 3.7V nominal and 96 in series, which gives 355V. 52 Ah is 18.47 kWh.

When charged to full, cell voltage is about 4.2 volts, 96 is 403V. 52 Ah is 21 kWh. But cell voltage can drop to as low as 3.2V (not sure exact value), 96 is 307V. 52 Ah is 16 kWh. That's why nominal value is more meaningful.

One thing to keep in mind is that load conditions will change Ah rating. If pushed too hard to too light, Ah will be less, which means kWh will be less, too.
Click to expand...

Ah, that makes sense.
 
I've been meaning to get one of those knock off GDS2s...

Where did you buy yours? Did you use the virtual machine some use? What sort of other things can you view?
 
JeremyW said:
I've been meaning to get one of those knock off GDS2s...

Where did you buy yours? Did you use the virtual machine some use? What sort of other things can you view?
Click to expand...

I bought it from Amazon. VCX Nano GM. It was even prime. Yes, it uses a VM for GDS2. You can see the status of pretty much everything and have some limited control for testing purposes.

Access to SPS is required for programming and some security functions. I had bought a 2 day SPS subscription to fix an issue with my Volt. After I did that I tried using it to check the module code versions on my Spark EV. But, neither the Spark nor Spark EV show up in SPS online. I checked in the regular Chevolet Section, as well as the Chevrolet (GM Korea) Section. Very odd.
 
Oh, and can you run the coolant and ATF pumps and the anti-lock brake pump on demand to allow flushing the fluids like a pro?

Where and how much!?
 
NORTON said:
Oh, and can you run the coolant and ATF pumps and the anti-lock brake pump on demand to allow flushing the fluids like a pro?

Where and how much!?
Click to expand...

I didn't look for those, but presumably GDS2 can do it. Link for the one I bought is below. $130 and you need a laptop. They recommended Windows 7 64bit, which I have. Not sure if Windows 10 will work. The VM is actually Windows XP and uses VMWare Player.

http://a.co/10s9QKk
 
By the way, something of interest might be voltage over % used. At full charge, it should be close to 400V, but what is it when it's almost empty? If you can find the voltage at each % and plot it, it would be interesting. Or if you have the raw data and not deal with plotting and hosting, I can make plots and in my blog post (giving you credit, of course).

Another is voltage data from each cell over discharge for many cycles without charging to full. I'm curious if cell balancing only take place on full charge on L1/L2. I don't normally charge to full, but few times charging to full with L2 seem to "bring back" the lost capacity. If I could measure cell voltage differences increasing over many not-to-full charge cycles, but they equalize on full charge, that will mean full charge is needed to balance the cells. Again, I can work with data and plot in a blog post if you rather not plot them.

Not-so-full charge test can also be done using only DCFC. But only doing with DCFC might be an issue since many Bolt get free charging and you have to wait for them.

Yet another is to test above on 2014 SparkEV's LFP battery. LFP typically has flattest voltage characteristics among all Lithium batteries, and I'm curious if that's true compared to NMC on 2015+ SparkEV. Flat voltage characteristics make it very difficult to balance the cells unless close to fully charged.

Finally, when you figure out a way to log the voltage (pack and cells), current Ah rating, state of charge, discharge current, charge current, all of them to a file, please post.
 
CCIE said:
NORTON said:
Oh, and can you run the coolant and ATF pumps and the anti-lock brake pump on demand to allow flushing the fluids like a pro?

Where and how much!?
Click to expand...

I didn't look for those, but presumably GDS2 can do it. Link for the one I bought is below. $130 and you need a laptop. They recommended Windows 7 64bit, which I have. Not sure if Windows 10 will work. The VM is actually Windows XP and uses VMWare Player.

http://a.co/10s9QKk
Click to expand...

I took another quick look and GDS2 does allow running an automatic break bleed procedure. I didn't see anything to run the ATF pump. There are also options to control the battery coolant pumps, active shutters, HV contactors, and a bunch of other stuff.

Also, I found the Spark EV in SPS. It is under GM Korea and is called Spark M-350. I'm not sure how I missed it yesterday. I took a look at the OS and Calibration levels for a few modules. My HPCM2 and Radio were out of date, but I didn't want to risk bricking them by updating. Car has been great, so no reason to chance it.
 
SparkevBlogspot said:
By the way, something of interest might be voltage over % used. At full charge, it should be close to 400V, but what is it when it's almost empty? If you can find the voltage at each % and plot it, it would be interesting. Or if you have the raw data and not deal with plotting and hosting, I can make plots and in my blog post (giving you credit, of course).

Another is voltage data from each cell over discharge for many cycles without charging to full. I'm curious if cell balancing only take place on full charge on L1/L2. I don't normally charge to full, but few times charging to full with L2 seem to "bring back" the lost capacity. If I could measure cell voltage differences increasing over many not-to-full charge cycles, but they equalize on full charge, that will mean full charge is needed to balance the cells. Again, I can work with data and plot in a blog post if you rather not plot them.

Not-so-full charge test can also be done using only DCFC. But only doing with DCFC might be an issue since many Bolt get free charging and you have to wait for them.

Yet another is to test above on 2014 SparkEV's LFP battery. LFP typically has flattest voltage characteristics among all Lithium batteries, and I'm curious if that's true compared to NMC on 2015+ SparkEV. Flat voltage characteristics make it very difficult to balance the cells unless close to fully charged.

Finally, when you figure out a way to log the voltage (pack and cells), current Ah rating, state of charge, discharge current, charge current, all of them to a file, please post.
Click to expand...

I don't usually run the car down below two bars, but next time I do I'll take a look at the pack voltage.

GDS2 shows voltage values for each of the 96 cell groups. But you can only view one "section" at a time, so it would be hard to log them all while driving. You should look at getting the VCX Nano since it sounds like you'd love the data it shows.

My understanding has been that cell balance only happens after full charge. It's done by using internal resistors to drain overcharged cell groups.
 
I don't want you to do things dangerous. I thought it's like "set it and forget it" type of data collection for unlimited number of parameters, but it sounds like it's not so simple.

If it's using bleed resistors to equalize, that's pretty inefficient. With only few more components, they could've "recycled" the energy. But I suppose there's no need for max recycling when the battery pack is so small.

For end of charge balancing, 2014's LFP cells would almost require that since the differences in voltages aren't much until almost full. But for NMC, I think the difference would be far more noticeable and able to equalize at any time, especially if it's simple bleed resistor type of equalization. It's only software change, but then they probably didn't bother to go the extra step for 2015, especially if equalization is not recycling the energy.

But even for end of charge balancing, I'm not sure if that happens if DCFC is used to take it to 100%. "Leaf" screen does not reset when DCFC was used to charge to 100%; people at Panda Chinese food took 20 minute to fill my simple order, and it got charged to 100% using DCFC.
 
SparkevBlogspot said:
I don't want you to do things dangerous. I thought it's like "set it and forget it" type of data collection for unlimited number of parameters, but it sounds like it's not so simple.

If it's using bleed resistors to equalize, that's pretty inefficient. With only few more components, they could've "recycled" the energy. But I suppose there's no need for max recycling when the battery pack is so small.

For end of charge balancing, 2014's LFP cells would almost require that since the differences in voltages aren't much until almost full. But for NMC, I think the difference would be far more noticeable and able to equalize at any time, especially if it's simple bleed resistor type of equalization. It's only software change, but then they probably didn't bother to go the extra step for 2015, especially if equalization is not recycling the energy.

But even for end of charge balancing, I'm not sure if that happens if DCFC is used to take it to 100%. "Leaf" screen does not reset when DCFC was used to charge to 100%; people at Panda Chinese food took 20 minute to fill my simple order, and it got charged to 100% using DCFC.
Click to expand...

It's not dangerous and GDS2 does have some graphing capability. But, it seems like it's only for the module and parameter that's being actively looked at. It's really designed for break-fix troubleshooting, not long term data logging. It provides a lot of data, but the program itself looks and behaves like it was written by a 12 year old. Though I guess it's better than the old Tech2.

My understanding is that the Volt, Spark EV, and Bolt all use resistors to bleed overcharged cells and convert the energy to heat. I agree that it seems like a waste, but maybe cell imbalances are normally so small that it's not worth doing something more advanced.

I don't normally DCFC to 100%, though I have and the car definetly stops charging and the status light goes into slow blink green mode. I'll try to remember to bring the GDS2 next time I DCFC.
 
When I DCFC to 100% in my 2014 Spark, it will reset the energy screen and flash the green LED, if you let it. The indicator on the DIC and fast charger will show 100% charge, however it takes a few more minutes (5 or so) of charging at 1-3 KW to allow the batteries to equalize and reach true 100% to reset everything. At least that's what I have observed and think is happening. I am able to see the additional energy draw past the charge complete indication with my OBDII dongle and Torq Pro on my smart phone.

Keith
 
I bought a used 2014 Spark EV and the top end number always seemed to be 84/85 miles range or so when using an L1 charger for a while.

I recently acquired a GE WattStation and after 2 days charging at L2, the top end range went to 91 today. Which surprised me because I was used to seeing an eight something for so long.
 
I connected my 2015 Spark EV with 16,700 miles to GDS2 again today. 3 miles showing on the GOM. Actual pack charge is at 7%, while the GOM is being told to display 4%. Pack voltage is 315V. Individual battery modules are between 3.24V and 3.34V. Pack resistance is 163.5 Ohms.

Battery capacity now shows 01FF. That comes out to 18.14KWh. Down from my reading at 11,000 miles.
 

Similar threads

S
ABS / Brake Bleed issue with 2015 Spark EV
Replies
0
Views
2K
SmokeyPete
S
I
[Software Glitch] Switching from Accessory/Service Mode to "Start" Bricks 12V System Causing 100A+ AGM Battery Drain
Replies
4
Views
6K
BetteInSCW
B
T
Battery Estimation Incorrect Recall/Fix and Battery Capacity
Replies
13
Views
29K
TheLondonBroiler
T
M
Potential purchase of 2015 Spark EV 1LT DCFC 40K miles
Replies
8
Views
8K
SparkE
S
M
Battery Degradation - 2014 vs 2015 Spark EV
2
Replies
26
Views
38K
Planerench
P

Latest posts

Back
Top