ElectricMotorcycleForum.com
Makes And Models => Zero Motorcycles Forum | 2013+ => Topic started by: Low On Cash on December 18, 2016, 07:54:17 AM
-
Thought it would make a good discussion concerning the short falls of Zero's Regen system - As most of you already know - instead of our "Regen Braking" being proportional to the throttle, our regen is either on or off.
Regretfully, even though all the logic is there in the Zero's throttle and controller system, for reasons unknown Zero elected to go with a basic "On or Off" approach instead of a linear Regen system that is proportional to the throttle and the bikes speed.
On my custom smartphone program, I dialed up 100% regen for the twisties and while it works just OK, its very abrupt when it engages and and disengages.
Thanks - Mike
-
Oh, do you think this hasn't been discussed to death for a few years now? It has. :) Not trying to be snarky, just saying this is well-trodden ground.
Just search for "variable regen" on the forum here for a taste.
-
The amount of regen and how it comes on is very much personal taste. Last time I went out to western NC/northern GA to play in the twisties I found having 100% regen set up in Custom mode was just too much. I switched over to Sport mode and stayed there the rest of the time. 100% Regen is nice on a few of the long hills out there as you can go for several miles without SOC dropping at all. One hill you actually recharged the battery a few percent as you had a higher SOC at the bottom than you had at the top. But for pushing hard in the twisties the amount of regen in Sport mode was fine. Maybe someday I will try riding from Maggie Valley to Cherokee in Sport mode and then do it again in Custom with 100% regen and document the difference in SOC.
-
I should say that I now ride constantly with 0% coasting regen and 100% braking regen and I find it the best way to use a Zero after you've spent a lot of time on it.
Knowing that the transition is there (a matter of having suitable expectations) and tuning the brake lever to soften it by having it just start to touch the pads as you engage the brake regen switch might be key to it.
-
I should say that I now ride constantly with 0% coasting regen and 100% braking regen and I find it the best way to use a Zero after you've spent a lot of time on it.
Knowing that the transition is there (a matter of having suitable expectations) and tuning the brake lever to soften it by having it just start to touch the pads as you engage the brake regen switch might be key to it.
There is another setting that works just as well as this if not slightly better as it gives you another option. Set coasting regen to 50-60% and braking to 100%. If you want to completely coast, just practice learning where the region is between throttle being off and regen being on. it can take some time to learn, and sometimes I have had the smallest amount of throttle when I thought I was coasting but that's ok, you can see it on the dash as it gives you bar graphs on whether power is going in or out.
-
There is another setting that works just as well as this if not slightly better as it gives you another option. Set coasting regen to 50-60% and braking to 100%.
I use a similar setting for off-throttle regen, 45%, which is perfect for me to maneuver through stop and go traffic without running into the backs of cars.
-
Thanks for all the reply's guys - sorry I was not trying to dig up old regen bones - I was more pointing to the question to why Zero would not refine their present system now they have a matured system.
If appears that all the logic is already there in their processors to enable a linear regen curve, which would just take some programming to make it work. Seems to me this would be another feather in their cap to have this advanced programmable braking system in their line of bikes.
Maybe I'm wrong but I feel it can all be done right in Zero App if we can get some hacker to look at our app. (When the Zero App works) it has real time speed and motor and battery data, how hard would it be to have your custom program increment the regen according to throttle position.
When I bought the very first BMW I3 years ago - the morons at BMW would only allow the generator to come on when the battery was almost dead (13%). Within a month the hackers got into the system and for $100 they provided a hack which allowed you to start the genny any time the battery was below 70% plus a number of other features like for US emissions cars they could not have a fuel tank over 1.9 gallons so the pencil heads at BMW programmed the car so it would turn off the fuel pump at 1.9 gallons so when you were stranded and out of fuel there was still a gallon of fuel in your tank. This same hack defeated that problem as well.
Regards
-
Having been a customer of BMW motorcycles for the past 30 years (strictly due to my customer-friendly dealership), I am not surprised that you ran afoul of BMW engineering, Low On Cash. A friend of mine, who used to be a freelance motorcycle reviewer (and was blackballed throughout the industry when he criticized their R1200S for a couple of goofy "features" in a new model review) use to say that BMW's motto was: "Ve are making ze perfect motorcycle, ze customer knows nothing!" ;)
-
BMW won't see a dime of my money - they are using their "once" fine name to lure customers into buying inferior products.
The BMW I3 electric is a train wreck with numerous problems. I dumped mine at a great loss and bought a Chevy Volt!
-
If appears that all the logic is already there in their processors to enable a linear regen curve, which would just take some programming to make it work. Seems to me this would be another feather in their cap to have this advanced programmable braking system in their line of bikes.
Maybe I'm wrong but I feel it can all be done right in Zero App if we can get some hacker to look at our app. (When the Zero App works) it has real time speed and motor and battery data, how hard would it be to have your custom program increment the regen according to throttle position.
It's fair to ask. The throttle input goes directly to the Sevcon controller, and it's the Sevcon settings and inputs that determine how the power train works. The phone app directs the MBB (main bike board) which can direct the Sevcon controller to change some settings on the fly.
It takes Sevcon software and a hardware connection to change that setting, and wiring up and assigning a new analog signal to yield a linear (say) regen experience is another bit of work.
-
I code for a living and do embedded programming as a hobby. It's amazing how people say "it's just a little bit of software" despite not having any idea what effort that requires or how it's done. If it's easy - give it a try yourself.
The main thing with all these possible software features is cost/benefit. It costs to redesign and manufacture PCBs. It costs to design, write and test firmware (especially if a bug could kill a customer). How does that cost compare to the net profit on any extra bikes you might sell?
-
...Set coasting regen to 50-60% and braking to 100%. ...
+1
I like this setting.
-
I like this setting.
This would make a great poll topic, if someone knows how to set that up... ("What are your preferred regen settings for coast and brake?")
-
I code for a living and do embedded programming as a hobby. It's amazing how people say "it's just a little bit of software" despite not having any idea what effort that requires or how it's done. If it's easy - give it a try yourself.
The main thing with all these possible software features is cost/benefit. It costs to redesign and manufacture PCBs. It costs to design, write and test firmware (especially if a bug could kill a customer). How does that cost compare to the net profit on any extra bikes you might sell?
You should do some code - the guy who did the I3 charged $100 and sold thousands of the fix's
-
I code for a living and do embedded programming as a hobby. It's amazing how people say "it's just a little bit of software" despite not having any idea what effort that requires or how it's done. If it's easy - give it a try yourself.
The main thing with all these possible software features is cost/benefit. It costs to redesign and manufacture PCBs. It costs to design, write and test firmware (especially if a bug could kill a customer). How does that cost compare to the net profit on any extra bikes you might sell?
You should do some code - the guy who did the I3 charged $100 and sold thousands of the fix's
That's an absurd suggestion, and shows that you have completely failed to grasp what he's trying to say.
-
Not following your reply Brian . . .
We're only talking about a linear regen feature here!
-
We're only talking about a linear regen feature here!
What's some code got to do, got to do with it?
-
Not following your reply Brian . . .
We're only talking about a linear regen feature here!
- You can't provide linear regen on a Zero without remapping how the throttle signal inputs to the Sevcon. You might want or need to provide an additional signal input, which requires skills in crimping and inserting wires into the Sevcon controller, which is a little more specialized than you'd suppose.
- Zero makes their own version of the Sevcon firmware; you'd have to binary patch the code in addition to patching the configuration settings. Developing a binary patch without source access is not going to be easy/cheap or reliable. The binary patch would also vary a bit per model/year.
- Applying the customization requires Sevcon cables and Zero's cables at the customer's site, or getting the dealer to do it (very unlikely outside of maybe two or three dealers in the world). In the 2017 world, there's a slight possibility of piggybacking on Zero's mobile app update arrangement, but this is a reverse engineering effort on top of the original effort.
- Zero firmware updates would often erase the customization and invalidate the way the customization works, requiring it to be re-developed after every Zero firmware update.
- These patches would invalidate your warrantee for anything related to the powertrain.
- The market for Zero owners willing to try this is maybe a few hundred, each willing to pay, sure $100-200, but that's not a lot of money to fund the development and maintenance costs (the author has to spend more hours upgrading the software per Zero firmware update, and will all of those customers want to pay again? Unlikely).
If it sounds like I'm making stuff up, I'll reveal that I had a reverse mode installed on my Zero, but it doesn't quite work right and someone's looking into it. This stuff does not just happen repeatably or affordably because one wishes it.
-
Exactly, Brian. Just a little bit of software. The ones and zeroes are even right there already, just not in the right order. ;-)
-
Not following your reply Brian . . .
We're only talking about a linear regen feature here!
- You can't provide linear regen on a Zero without remapping how the throttle signal inputs to the Sevcon. You might want or need to provide an additional signal input, which requires skills in crimping and inserting wires into the Sevcon controller, which is a little more specialized than you'd suppose.
- Zero makes their own version of the Sevcon firmware; you'd have to binary patch the code in addition to patching the configuration settings. Developing a binary patch without source access is not going to be easy/cheap or reliable. The binary patch would also vary a bit per model/year.
- Applying the customization requires Sevcon cables and Zero's cables at the customer's site, or getting the dealer to do it (very unlikely outside of maybe two or three dealers in the world). In the 2017 world, there's a slight possibility of piggybacking on Zero's mobile app update arrangement, but this is a reverse engineering effort on top of the original effort.
- Zero firmware updates would often erase the customization and invalidate the way the customization works, requiring it to be re-developed after every Zero firmware update.
- These patches would invalidate your warrantee for anything related to the powertrain.
- The market for Zero owners willing to try this is maybe a few hundred, each willing to pay, sure $100-200, but that's not a lot of money to fund the development and maintenance costs (the author has to spend more hours upgrading the software per Zero firmware update, and will all of those customers want to pay again? Unlikely).
If it sounds like I'm making stuff up, I'll reveal that I had a reverse mode installed on my Zero, but it doesn't quite work right and someone's looking into it. This stuff does not just happen repeatably or affordably because one wishes it.
Brian,
Regretfully - You are 100% "incorrect" again!
There is no re-wiring and crimping wires, or changing any mechanical aspects of the bike what so ever. It's all accomplished with lines of code. The regen is already there and fully adjustable which is proven by the fact you can control regen it to a degree with your smartphone. Making the Regen linear would be a simple as programing the degree (%) of Regen to a proportional position of the throttle, so instead of dropping to full regen at an instant, it would look at the throttle position and increment it in a proportional manner.
As I mentioned in my previous post, all the modifications of the BMW I3 electric was strictly software - a simple plug and play procedure, which also had the capability so when you re-installed the software it would put the system back to its original status for warranty reasons.
Regards!
-
Well, I'm glad we have someone who knows everything to tell us how things work. Good job! Have a great holiday.
-
Well, I'm glad we have someone who knows everything to tell us how things work. Good job! Have a great holiday.
Merry Christmas to you as well Brian!
-
Making the Regen linear would be a simple as programing the degree (%) of Regen to a proportional position of the throttle, so instead of dropping to full regen at an instant, it would look at the throttle position and increment it in a proportional manner.
Awesome news! I'd be happy to beta test that for you. When do you think you'll be finished?
-
I code for a living and do embedded programming as a hobby. It's amazing how people say "it's just a little bit of software" despite not having any idea what effort that requires or how it's done. If it's easy - give it a try yourself.
The main thing with all these possible software features is cost/benefit. It costs to redesign and manufacture PCBs. It costs to design, write and test firmware (especially if a bug could kill a customer). How does that cost compare to the net profit on any extra bikes you might sell?
You should do some code - the guy who did the I3 charged $100 and sold thousands of the fix's
That's an absurd suggestion, and shows that you have completely failed to grasp what he's trying to say.
Thanks for the reply - You mean you failed to grasp it right ?
-
So... are there examples of open source EV control to build this idea on?
-
So... are there examples of open source EV control to build this idea on?
There are open-source EV DIY platforms: http://wiki.p2pfoundation.net/Product_Hacking#Cars (http://wiki.p2pfoundation.net/Product_Hacking#Cars)
But I haven't run across anything like that specifically. Basically, it rounds up to generic firmware hacking, and there are open source toolkits for debugging and patching embedded binaries, but they're not usually made for customizing vehicles so they may not have the kind of stability desired.
There's literally a how-to for a linear regen lever, by the way:
http://electricmotorcycleforum.com/boards/index.php?topic=4357.0 (http://electricmotorcycleforum.com/boards/index.php?topic=4357.0)
Indexed on the wiki here:
http://zeromanual.com/index.php/Advanced_Modifications#Left_Hand_Variable_Regen_Lever (http://zeromanual.com/index.php/Advanced_Modifications#Left_Hand_Variable_Regen_Lever)
It's not something that's been replicated publicly. Honestly, the main problem is that everyone wants something different out of this kind of feature; the OP of that procedure is a motorcycle racer so they wanted a lot of performance and control. Others want the Vectrix style of regen-using throttle which is convenient but requires more extensive rework to implement.
All of this is clouded by the misconception (repeated here) that what the app does is equivalent to variable regen, but it is not - it's changing a static numeric setting and getting the throttle or an analog lever to work dynamically is an entirely different task that requires bypassing a lot of Zero's arrangements.
The wiki links to the Sevcon documentation which I've been through. This can be done, but it's a little complicated and is best done with Sevcon's software which has become locked down in the last few years, and the cables which are a little easier to acquire but are not cheap. Anyway, the point is that implementing this feature requires the dedication of a racing team or a dealership.
-
The mod does not have to be perfect - anything other than the "On & Off" switch it has now would be a vast improvement. I dialed up full regen and braking on my custom program but shortly after changed it because it was just too abrupt when letting off the throttle.
I sure not a programmer but I would "Guess?" it might be accomplished in just the phone app if a guy was sharp.
-
The app can't do it - bluetooth messages cannot be reliably sent and responded to quickly enough or reliably enough to get the desired effect. I'm not guessing and I don't have to guess (I've done a preliminary bluetooth sniffing session on the app and wish I had more time, but the basic gist is that the messages are not meant to be sent that fast and bluetooth isn't reliable anyway).
The MBB isn't involved with the throttle signal for exactly this reason, and the MBB is the only thing the app talks to. The MBB can change the Sevcon settings via CAN bus messages, but the refresh rate on that is about a few seconds at best for the full round trip, and it is not accessing Sevcon's variable regen feature.
-
The app can't do it - bluetooth messages cannot be reliably sent and responded to quickly enough or reliably enough to get the desired effect. I'm not guessing and I don't have to guess (I've done a preliminary bluetooth sniffing session on the app and wish I had more time, but the basic gist is that the messages are not meant to be sent that fast and bluetooth isn't reliable anyway).
The MBB isn't involved with the throttle signal for exactly this reason, and the MBB is the only thing the app talks to. The MBB can change the Sevcon settings via CAN bus messages, but the refresh rate on that is about a few seconds at best for the full round trip, and it is not accessing Sevcon's variable regen feature.
Brian,
I just wanted to make sure you understand how the custom program in our App works - The Zero does not depend on the App nor the Bluetooth link to operate!
Once the custom program (or perhaps linear brake program) is uploaded to the Zero's processor, the App has finished its job. When the bike is running, the operation of the phone App, its settings and Bluetooth link have absolutely nothing what so ever to do with the processors operation. The phone App and its Bluetooth link is only a source to alter the operation routine when stopped and not control it.
For this same reason if the phone app was modified for regen, it would then have the capability to alter the Zero's code to follow a protocol to apply regen according to the throttles position to make the regen linear and not abrupt as it presently is.
I also feel that in due time, Zero will incorporate this feature in all their models.
Regards - Mike
-
I lost patience reading the last few posts, so forgive me if it was already mentioned, but to have a variable regen we need a variable input to the regen: we need a way to sense how much brake is applied. That's hardware, which the bike does not have.
On the other hand, to have a gradual regen level increase, meaning it would kickoff smoothly and increase to whatever max level the user sets and stay there for as long as the brake is applied (or whatever the conditions to start regening), should have been implemented as a standard. Arguments that it is a hard thing to do for the software developers at Sevcon or Zero I think don't hold water - if the controller has the capability, the programming would be trivial. They could have added a second parameter: how quickly the regen ramps-up from 0 to max (the other parameter being the "max" itself).
Sure, custom-programming a closed, proprietary system is not easy. But the programming of said system to do gradual reven, when done by the creators of it, shouldn't be rocket science, provided the hardware is there to support it.
-
I think that what Zero does in the way of programming changes has more to do with what they think will sell more motorcycles to their future customers and less about what can be done to keep some current owners happy. While I agree that a smoother transition between on and off regen when slowing down to a stop, or when the battery is at 100% and not accepting regen for more than a second or two before shutting it off would be nice, I am pretty happy with the way that the system currently performs on my 2014 S. I just shut it off when the throttle is closed and coast, while letting it supplement my braking when I touch my front brake lever lightly. That works for me. :)
-
When I said "everyone wants something different out of this feature", I meant that literally we're all talking about different goals.
I was responding only to the idea that some enterprising third party could easily patch the firmware and make a worthwhile amount of money.
Honestly, the word "trivial" is one of those words people use about software development that I always regard with skepticism, because I know it means only casual consideration for how it gets made. And nothing about hardware control by software is trivial.
The Sevcon does not support an implicit regen ramp up out of the box and isn't motivated to because their larger market is served fine with how they do it now.
For what it's worth, an implicit ramp creates some slight difficulties making regen work right if it's toggled too quickly.