ElectricMotorcycleForum.com

Makes And Models => Zero Motorcycles Forum | 2013+ => Topic started by: dhln on May 22, 2018, 01:25:56 PM

Title: SR 2016 acts like an "S"-model after firmware update
Post by: dhln on May 22, 2018, 01:25:56 PM
Hello folks,

I am fairly new here and read a lot about my new Zero SR 2016 here.
I recently visited a Zero dealer to have the Touring windshield installed and to have my firmware updated to the latest (I think it's v53, coming from v51).
After the update, my SR is no longer recognized in the app (see attachment) and the top speed (in the app settings and in real life) is limited to 151 km/h (93mph) instead of 164 km/h (102mph) before the update.
The counters for charging cycles and range/consumption were also cleared :-/
Is it possible to make an "SR"-model to an "S"-model by mistake by using a different firmware?
Can I safely drive like this?

I am already in contact with the dealer but he is waiting for information from Zero.
I live in Germany if that's somehow relevant.

Thanks in advance!
Dennis
Title: Re: SR 2016 acts like an "S"-model after firmware update
Post by: kashography on May 22, 2018, 05:55:13 PM
Newest should be 53.
I also noticed that in my logfile the "Model" is marked with "NON" instead of "SR", so i think they failed to privide the Model totally and now takes some fallback/standard values in the app
Title: Re: SR 2016 acts like an "S"-model after firmware update
Post by: dhln on May 22, 2018, 07:02:14 PM
Do you also have limited top speed? Which are the max-values of the top speed slider in custom-mode?
Title: Re: SR 2016 acts like an "S"-model after firmware update
Post by: Richard230 on May 22, 2018, 07:40:24 PM
When I encountered the latest BMS and MMB versions 31 (Zero apparently switched firmware revision numbers on the new models), used on my 2018 S, the information supplied by my Zero app was neutered to look like yours and I discovered that the maximum speed on the "custom mode" slider had been dropped to something like 93 mph.  I can't quite recall exactly what the top speed had been reduced to because I immediately slid it back up to 102 mph and so far it seems to be staying there.  :)

Zero seems to be on a march to reduce the amount of information regarding the condition of the battery pack displayed on their app from what they originally provided when I first used the app on my 2014 S, to the latest information provided with firmware version 31, which apparently updated via the firmware instead of the usual cell phone update method.  Apparently Zero would prefer that their customers not know the status of their battery packs as the batteries age over time.  :(  I can only surmise as to what that reason might be.  ::)
Title: Re: SR 2016 acts like an "S"-model after firmware update
Post by: dhln on May 23, 2018, 12:11:12 AM
Thanks for the replies. I can not slide the values up as they are already maxed out (the limits are lower now).
I have to wait for the dealer, he says we will be able to provide the information during the week.
Title: Re: SR 2016 acts like an "S"-model after firmware update
Post by: BrianTRice@gmail.com on May 23, 2018, 06:11:14 AM
Richard's observation is interesting, and I feel like I have something to say in this area:

Based on what I've observed, Zero is not limiting access to battery state because of aging, but seemingly so they can try more dynamic approaches to range and charge state indications.

Having observed some tests with recent (but not latest) firmware, I think these approaches should be reconsidered, based on my experience in the software industry, engineering mathematics education, and having spent some time calibrating electrical instrumentation and performing equipment analysis in the US nuclear navy.

I have not seen the latest dealership firmware, but the general firmware management process has seemed too raw and error-prone. Zero reacts to such tension by clamping down, and I would strongly encourage any engineer or manager there to resist this trend and consider that operational ergonomics is an actual studied field that they can approach without risking project deadlines or increasing risk for dealers or customers.

Customers need a simple but effective model of how their vehicle operates in order to understand its state in both normal and abnormal conditions. Dealers also need a less simple but effective model of the vehicle to troubleshoot and help keep the vehicles running affordably and provide feedback when failures are triaged. Zero has succeeded in simplicity for the owner but erred in effectiveness for both owners and dealers as a class.

I'm using the term "effectiveness" in a narrow technical sense of whether an owner or technician can read an indication and act on it safely and correctly.

The user is in almost all cases not informed how to act on a Zero motorcycle's indications in a meaningful way, and my self-assigned role in charting the paths through working with Zero's systems in the unofficial manual is unpaid work for Zero to compensate for this. I can only hope to help dealerships out, but I have no visibility there and just vaguely hope that I'm not causing them grief with customers harassing them or ignoring them.

Obviously, I'm going to proceed with or without Zero's assistance, but I don't agree with the manifestation of their strategy both as observed and as declared. I think the bikes can be more serviceable, more reliable, and more understandable than they are; it's debatable how Zero can get there, but the ingredients are available as long as the process changes to improve them.
Title: Re: SR 2016 acts like an "S"-model after firmware update
Post by: Doug S on May 23, 2018, 08:29:58 PM
...I think these approaches should be reconsidered, based on my experience in the software industry, engineering mathematics education, and having spent some time calibrating electrical instrumentation and performing equipment analysis in the US nuclear navy.

I have a lot experience designing products with batteries too, and I agree. Cell voltage, with small corrections based on temperature and current draw/charging current, is really the best way to evaluate SoC on a well-behaved battery pack. Coulomb counters are a red herring; charging/discharging efficiency can vary pretty substantially depending on several things. Similarly, distance traveled means very little since the conditions can vary so much. Tiny changes to the "full battery" capacity estimate can be made over time to correct for battery aging, but that's about all that's really needed to get as close as you can to an accurate estimate of battery SoC.
Title: SR 2016 acts like an "S"-model after firmware update
Post by: BrianTRice@gmail.com on May 23, 2018, 09:19:58 PM
...I think these approaches should be reconsidered, based on my experience in the software industry, engineering mathematics education, and having spent some time calibrating electrical instrumentation and performing equipment analysis in the US nuclear navy.

I have a lot experience designing products with batteries too, and I agree. Cell voltage, with small corrections based on temperature and current draw/charging current, is really the best way to evaluate SoC on a well-behaved battery pack. Coulomb counters are a red herring; charging/discharging efficiency can vary pretty substantially depending on several things. Similarly, distance traveled means very little since the conditions can vary so much. Tiny changes to the "full battery" capacity estimate can be made over time to correct for battery aging, but that's about all that's really needed to get as close as you can to an accurate estimate of battery SoC.

Current helps estimate and mitigate the effects of lift and sag in battery voltage at the terminals, and then there are temperature effects.

So, I can understand the attraction to coulomb counting, but here’s what annoys me:

As an owner and operator of the vehicle, I have no indication of which method the vehicle is estimating its charge with, nor do I have a control or toggle to reset to voltage based mode, nor is there a described or understandable way to know when the bike will reset the method to voltage.

Even access to voltage based SoC calculation while the vehicle prefers coulomb counting would be a benefit.

PS of course the owner can press the soft reset on the BMS, but this is not explained anywhere and I don’t think removing the BMS cover is suitable for anything but an emergency situation.