Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[gt13013]

My FXS 2016 was using his original firmware until July 2021 (Firmware rev. 47).
In July 2021 it has been repaired (controller replaced), and the firmware became rev. 56.

Before, the SOC was a pretty good measure of the remaining capacity of the battery, and a very good indicator of the remaining range.
But since the repair, I get strange behaviour of the SOC: decreasing and rising inconsistently, decreasing much faster than before, and consequently the range has decreased a lot.

I rode approximately the same trip several times.
Since I keep the logs for the entire life of my bike, I was able to extract the data from 4 of these trips. And that shows a clear difference between the firmwares.

These data confirm the overall impression:
- the displayed SOC decreases much quicker with the new firmware, and the range has decreased since the bike seems to not accept the lowest pack voltages that it accepted with the old firmware. I have a log of a full battery depletion with the old firmware. I have to find some time in order to do a full depletion with the new firmware and compare the data.
- another impression is that the new firmware SOC is mainly based on the battery pack voltage, in such a way that it decreases a lot if you ride the bike intensively, but then increases as soon as you ride more quietly. With the new firmware, when I start the bike, I can get 2% of SOC decrease in 2 minutes, without riding the bike, just due to the headlight decreasing the battery voltage! The old firmware was smarter, and probably taking into account the amount of energy used. Its behaviour was really good and it gave a pretty accurate SOC.

The most interesting part is in the attached spreadsheet. Be sure to explore the 3 tabs in this spreadsheet.

[sharagan]

either they want to stretch the lifetime of the battery so it survives until the warranty ends or to push the 80% mark a little lower to get rid off some warranty claims or they will slowly push the range lower even further so that they can offer us a "range extension" via the cypher store and implement the original values.

[gt13013]

I take great care of my battery pack. It has now 6 years and I think it has kept almost the same capacity as at the beginning. Zero would better allow us to use the full capacity of the battery, and tell us clearly what is good or bad to keep it in good condition. When I got my bike, the manual advised to let it charge all the time (even for long periods of storage!). But at that time, it was already knew that it is bad!

A least, it would be fine to get the entire range (without paying!), as soon as the battery pack is no longer under warranty...

[sharagan]

maybe it would be worth to send out a message to Zero and ask them explicitly why the new firmware decreases the usable voltage range of the battery. I am pretty sure it is intentional and not an error.

[gt13013]

I must first do a full discharge test with the new firmware, and compare it to the one I had done with the old firmware.

By comparing the operating voltage limits of the batteries, we will be able to see if Zero has actually narrowed the range of usable voltages or not.

Until I have done this test, I will not be sure that the autonomy has really decreased. It remains possible that only the SOC displays are miscalculated...

[gt13013]

I made a total discharge of my battery (until the bike stops), and it seems that the cutoff voltage has not changed (compared to the old firmware). It is just the SOC that decreases very quickly. With a SOC at 0%, I was able to ride about 15 km.
The bike stopped when
- the lower cell went to 3100 mV
- the pack went to 87 V
I do not know which condition between these two made the bike stop.
After 4 minutes with the system off, the voltages probably increased a little bit, and the bike accepted to ride me to my garage (about 1 km)

[gt13013]

I have tested twice a complete discharge of my bike.
- March 17th, 2019: with the original firmware 47. I stopped the test after 5 kms of riding with 0% SOC, but the bike was able to ride more.
- Feb. 13th, 2022: with the updated firmware 56. I rode the bike until it stops.
I have extracted the data from the logs and compared them in the attached spreadsheet.
The capacity of the batteries seems to have stayed almost unchanged, despite the bike enters in its 7th year.
It confirms that the new SOC is very pessimistic compared to the old one, and also it is fluctuating a lot.

[Curt]

That's some nice data analysis.

The older firmware behavior seems superior. The new one behaves more like what we Chevy Bolt EV owners call the "Guess-o-Meter" (GOM), which is a range readout that uses advanced estimation techniques to confuse the hell out of everyone.

Glad to hear that 2016 batteries are holding in there. I haven't noticed any degradation myself, but also haven't been taxing it as much as I used to.

[gt13013]

Concerning the problem about wrong estimations of the SOC, here is some recent information that I have got from studying the logs.

Clearly, the way the SOC is estimated has changed due to a firmware update, and perhaps some reprogramming from the Zero dealer.

Original firmware:
Serial number      2016_mbb_ca1a29_02158
VIN                538XXAZA1GCJ06548
Firmware rev.      47
Board rev.         3
Model              FXS

After update:
Serial number      2016_mbb_ca1a29_02158
VIN                538XXAZA1GCJ06548
Firmware rev.      56
Board rev.         3
Model              FXS

Starting from the firmware update, new lines appear in the BMS logs.
Here are 2 lines from the new BMS logs. Lines such as line 01470 did not appear with the original firmware.
These new lines give 8 different values related to the capacity and the SOC.

01469    01/09/2023 13:08:58   Discharge level      002 AH, SOC: 70%, I:034A, L:3756, l:3971, H:3807, B:051, PT:022C, BT:026C, PV:106014, M:Bike On
01470    01/09/2023 13:08:58   SOC:2341,7632,2672,26000,90,85,70,53   

In the 8 numbers that follow SOC, it seems that the third one, 2672, is the amount of mAh that the pack has delivered. I guess that this measure is obtained by Coulomb counting (or integration of the intensity). This value is rounded to 2 Ah in the line 01469 just before.
26000 is the nominal capacity of the pack in mAh. It has been stored in the BMS by the Zero dealer.
It seems that 90,85,70 are 3 different estimations of the SOC calculated from different ways.
The one that is used is the third (70%). Probably mainly influenced by the pack voltage.

From the line 01470, I have a pack with 26000 mAh and it has delivered 2672 mAh, i.e. taking only these numbers into account, its SOC should be 89.7%. Note that it is equal to the first of the 3 SOC estimations (90,85,70).
So, it seems that the first SOC estimation (90%) is derived from the Coulomb counting.

In my bike, I have two battery packs, one BMS in each pack, one log for each BMS.
And it seems that the MBB makes a mix of the information retrieved from the 2 BMS (average of their SOC).
And the dash gives this mixed information.

Since information displayed uses the third estimation of  70%, it means that instead of telling me that I have used 10% of the capacity, the dash tells me that I have used 30% !!!
That is why I often get a big range panic !  This day (last January 9th), I was thinking that one of my battery packs was dead or disconnected ! But finally both packs were working, and the range was not so bad.

To those who know how to make Android applications taking their data from Zero bikes: probably many of us would be happy with a range estimated from
- Coulomb counting
- entering manually the pack capacity (so we could take manually into account the real capacity of the pack, take into account smaller capacity in winter, or keep a reserve capacity for safety)
- giving an estimate of our Wh/km consumption

[Bodo]

Gerard, looking at Zero's fimware changelog, I suspect a mismatch.

The latest official firmware offered for Cypher II bikes is MBB v37, BMS v44 - yours was 47 and now is 56.

That said, mine is on firmware mbb_fw_ver: 36 and I don't see problems with my dashboard SoC.

[Richard230]

I really appreciated the first version of the Zero app that I used for my 2014 Zero S. It was able to display the remaining power in the battery pack by showing the kWh remaining in the pack. As near as I could tell it was pretty accurate, based upon the amount of power used to recharge my bike as measured by my Kill A Watt meter. At some point, about a year later I think, that information disappeared from the app and battery information seems to have become less informative with every new app upgrade. 

[DerKrawallkeks]

Makes me glad my bike has not been updated since 2016. It works perfectly with no weird behavior and almost no magic charging. (just sometimes if it was super cold and warms up, it might get one more bar, which I actually appreciate cause it represents the true available energy, unlike in a phone, where it is consistantly going down.)

Well I wouldn't want a percentage number to increase, that would feel weird, but since my Zero has bars, it don't mind one coming back if the bike feels that is good.

I also really like the old app, even though it doesn't work correctly with all values. I still like the idea that it displays some more information.

[gt13013]

Gerard, looking at Zero's fimware changelog, I suspect a mismatch.
The latest official firmware offered for Cypher II bikes is MBB v37, BMS v44 - yours was 47 and now is 56.

No, look at the "2016" section on page https://www.zeromotorcycles.com/fr-fr/firmware
MBB last version for these models should be 56

[gt13013]

I really appreciated the first version of the Zero app that I used for my 2014 Zero S. It was able to display the remaining power in the battery pack by showing the kWh remaining in the pack. As near as I could tell it was pretty accurate, based upon the amount of power used to recharge my bike as measured by my Kill A Watt meter. At some point, about a year later I think, that information disappeared from the app and battery information seems to have become less informative with every new app upgrade. 

If you are with Android, you can use old versions of the Zero App. For Apple system I don't know.
I personally still uses version 1.6.50 on my 2016 FXS. You must tell your phone that you do not want automatic upgrades of the App, and it will keep this version.
Here is a topic about this, and it ends with an Excel worksheet where I compare 3 versions:
https://electricmotorcycleforum.com/boards/index.php?topic=7720.15

[Bodo]

Gerard, looking at Zero's fimware changelog, I suspect a mismatch.
The latest official firmware offered for Cypher II bikes is MBB v37, BMS v44 - yours was 47 and now is 56.

No, look at the "2016" section on page https://www.zeromotorcycles.com/fr-fr/firmware
MBB last version for these models should be 56

Thanks, I see.

What a mess this page is: fw versions classed by model years & operating systems, fw versions classed by models and some model years where different model&model year combinations overlap, fw versions just by a single model year w/o model, but overlapping by other model year and model combinations, incoherent versioning for the same device name, overlapping device naming for different devices.

People need profound knowledge of Zero's model range, model history and operating system strategy to determine which version is correct for their bike.

Zero is clearly unable to cope with software. Gerard, you should contact them with your SoC problem so that they can get it fixed when the newer fw version caused the unexpected SoC display.