BCU Software Version 30.00

18 posts / 0 new
Last post
PJD
PJD's picture
Offline
Last seen: 1 week 3 days ago
Joined: Wednesday, November 22, 2006 - 05:44
Points: 1416
BCU Software Version 30.00

I installed the version 30.00 of the BCU software on my C1214 and have now had a week to test it out. Here is what I noticed...

1. I loaded the software onto the BCU per the instructions in the BCU Software Tools document without problems. The first charging after installation showed that the state of charge indication problems were fixed - the gauge indicating 100% at end of charge. However, after a few rides it was apparent the fuel gauge was not reading correctly. A call to Current indicated that the gauge calibration routine in terminal mode needs to be run after the v30 software installation. I did this, and the gauge now works perfectly.

2. I'm not sure if it is just the state of charge working correctly, but the range seems to be improved. After 39 miles on riding, the SOC of still 25 percent, so a range of 50 miles should be comfortably achievable in the lower-speed 25 to occasionally 45 mph urban riding with some moderately long hills.

3. The function of the low/high switch has changed. Before, the low setting was, well, too low to be usable. Acceleration was too slow and I wouldn't even go 25 mph up many hills. But having a lower-current-draw "economy" setting for city riding is still a good idea; it just needed to be a bit higher than the current setting. Well, with v30.00, the "low" setting is higher, but so high it seems indistinguishable in acceleration from the "high" setting. (at least up to 45-50 mph, which is as fast as any traffic ever goes in this neck of the woods) Perhaps a little tweak on this would be nice for Version 30.01.

4. The regen braking now shuts off at about 5-7 mph, instead of continuing all the way down to virtually stopped. I assume that the main purpose of this modification was to eliminate the harmless but unappealing "thunking" of the Kelly seeming to not quite "know what to do" with the small regen current from the motor at very low speed. I developed a habit of simply letting off of the levers momentarily below 5mph when braking to a stop. This modification eliminates this, but unfortunately it also eliminates being able to use regen while slowly descending a hill due to backup at a traffic light, rough road, or other conditions. So maybe another minor tweak for v30.01.

I'm sure there are other changes too. I look forward to John commenting here about some of them.

jdh2550_1
jdh2550_1's picture
Offline
Last seen: 10 years 11 months ago
Joined: Tuesday, July 17, 2007 - 09:35
Points: 2335
Re: BCU Software Version 30.00

Hi Paul,

I installed the version 30.00 of the BCU software on my C1214 and have now had a week to test it out. Here is what I noticed...

Actually it's version V0.0.30 ;-). It will become V1.0 sometime in the next few months. This version is a strong candidate for V1.0 but there are a few updates still required (which will be shorthand V31 and in longhand V0.0.31). Want to really get into minutiae? I use a three part numbering scheme:

VA.B.C

A = major product release. Will be set to 1 when we have all the features in that we want for our mass produced bikes (did you see the news about teaming with a Tier 1 auto supplier for manufacturing? Very, very good news!)

B = minor external release. Will be bumped up if we add minor updates or bug fixes to mass produced bikes.

C = iterative development releases. This is the only number changing at present. Remember that you're a test pilot. You're part of development. That's a good thing! ;-) Now, because of the fact that there are currently around 30 test pilots on the road I treat this last component with a bit more formality - so you may see that you have V0.0.30RC5. RC5 means "release candidate 5" which means it's the fifth build addressing stuff we found before we wanted to put it in the field. (In truth it's probably the 100th build for "V30" that I've done as I bop back and forth between the code, the BCU test bench and a BCU on the bike - if there were two developers I might even add a D).

You're probably asleep by now - this is stuff only a software engineer could love!

But, hey, who's counting? ;-)

1. I loaded the software onto the BCU per the instructions in the BCU Software Tools document without problems. The first charging after installation showed that the state of charge indication problems were fixed - the gauge indicating 100% at end of charge. However, after a few rides it was apparent the fuel gauge was not reading correctly. A call to Current indicated that the gauge calibration routine in terminal mode needs to be run after the v30 software installation. I did this, and the gauge now works perfectly.

9 out of 10 folks we upgrade for them. V30 is a big change over V29A and everything else before it. Hence I'm not able to easily carry forward the configuration settings in the EEPROM (different EEPROM layout - getting rid of some cruft that's not needed and making space for new stuff). Hence we forgot to tell you that step - sorry about that. But Stephen was able to answer your questions.

2. I'm not sure if it is just the state of charge working correctly, but the range seems to be improved. After 39 miles on riding, the SOC of still 25 percent, so a range of 50 miles should be comfortably achievable in the lower-speed 25 to occasionally 45 mph urban riding with some moderately long hills.

Yes, a lot of that is do with the new SoC software. The earlier versions weren't as sophisticated and I purposefully erred on the side of caution. The new version is more sophisticated, and more accurate. With the earlier pessimistic software SoC at 0% occurred earlier just because it counted too aggressively (remember SoC at 0% really means 20% remaining in pack - SoC is a percentage of "available energy" and "available energy" is to 80% DoD). The error in the old software was also cumulative meaning that folks were complaining of falling range.

3. The function of the low/high switch has changed. Before, the low setting was, well, too low to be usable. Acceleration was too slow and I wouldn't even go 25 mph up many hills. But having a lower-current-draw "economy" setting for city riding is still a good idea; it just needed to be a bit higher than the current setting. Well, with v30.00, the "low" setting is higher, but so high it seems indistinguishable in acceleration from the "high" setting. (at least up to 45-50 mph, which is as fast as any traffic ever goes in this neck of the woods) Perhaps a little tweak on this would be nice for Version 30.01.

We're experimenting with changing the purpose of the hi-lo switch. In "hi" we put the onus on the rider to decide how to manage his/her energy usage. We only put cutbacks when things get too hot. In "lo" we take the view of trying to give the rider the a "pre-programmed" balance between performance and range - that is to say we back off accel slightly and we back off at higher speeds. The reduced accel will hardly be noticed - it only applies between about 20mph and 40mph - however, if you were to look at a graph of the current draw you'd see that we've "lopped off" a significant current peak for very little impact in actual acceleration and even less impact on perceived acceleration. At the top end - you should find that in LO you top out around 55mph whereas in HI you will top out around 60 to 65 (on a C124). It can take a long time to get to those top speeds.

The calibration was developed using a C124. Unfortunately due to other commitments I wasn't able to finish a C130 calibration - so it shares the same calibration as the C124. As a result of the higher power of the C130 the differences between the HI & LO are even less apparent on the C130.

I'd like to hear what you and others think about this re-purposing of the switch (HI - you do the management; LO - we do the management). Is this a good idea? Or not? Is it a good idea but you also need a lower LO? If you can't have both which is preferable?

4. The regen braking now shuts off at about 5-7 mph, instead of continuing all the way down to virtually stopped. I assume that the main purpose of this modification was to eliminate the harmless but unappealing "thunking" of the Kelly seeming to not quite "know what to do" with the small regen current from the motor at very low speed. I developed a habit of simply letting off of the levers momentarily below 5mph when braking to a stop. This modification eliminates this, but unfortunately it also eliminates being able to use regen while slowly descending a hill due to backup at a traffic light, rough road, or other conditions. So maybe another minor tweak for v30.01.

Yup that's exactly why we did this. Ultimately I hope to provide the rider the ability to chose regen strategy themselves. How much on front brake? How much on rear brake? Cut off at low speed? etc. I think this is an area of great subjectivity which makes it difficult for Current Motor to "get right" for everyone.

I'm sure there are other changes too. I look forward to John commenting here about some of them.

You hit all the big ones. Another one is that the end-of-charging algorithms have been much improved. There should be less confusion related to the LED.

The good news is that this release seems to have stabilized things quite well - which means I'm working on the LCD screen that will replace the ammeter to the right of the speedo. When we have that we will still have the red LED but we will only use it less frequently.

And after that we've got a whole boatload of stuff to keep making our bike even better. Never a dull moment!

All the best. John H.

John H. Founder of Current Motor Company - opinions on this site belong to me; not to my employer
Remember: " 'lectric for local. diesel for distance" - JTH, Amp Bros || "No Gas.

PJD
PJD's picture
Offline
Last seen: 1 week 3 days ago
Joined: Wednesday, November 22, 2006 - 05:44
Points: 1416
Re: BCU Software Version 30.00

John, this is only partly related to the BCU software and partly related to hardware, but it seem to me that the most urgent improvement you need to address is the charging system.

What I am finding is happening, particularly in cooler weather, is that the bulk charging is ending, then the balance mode is taking a very long time, 2 hours or more, to get those last couple AH in the pack. This has nothing to do with balancing, but rather has to do the bulk mode cutting off then the cell voltages dropping way back down to 3.35-3.38 or so, resulting in a long period of charging at the 0.5 amp balance charge rate to get back to the full cell voltage at about 3.65 volts per cell or a pack voltage cutoff at about 88 volts. I assume the cells are well matched, so little or no balancing is needed or occurring. But, the balance mode takes excessively long time because of the long recovery in voltage after bulk charging ends. And this effect is going to get worse as the temperature drops and cells internal-resistance rises.

The problem is that your current charging protocol abruptly switches to a low-current mode that doesn't swamp the balancer, so it does not follow the ideal CC-CV protocol that charges LiFePO4 type cells most quickly and efficiently.

One solution would be to go to a three or four stage charging that would mimic CC-CV a little more closely - say 8-10 amp bulk mode, 4 amp, then 2 amp "top-off" modes, then 0.5 amp balance mode. I don't know how easy this would be to implement.

I know that this issue only involves the last 2-3 AH of charging, and one could just stop the charging and go riding with only a little effect on range, but most riders don't like taking off with charging still going on - they want a nice full pack before heading out.

The other change is finding a way to shut off the charger fan when charging is complete.

By the way, with charging stopped, but still plugged in and the charger door left open, how much current is being drawn from the pack?

MEroller
MEroller's picture
Offline
Last seen: 11 months 1 week ago
Joined: Monday, September 26, 2011 - 09:24
Points: 847
Re: BCU Software Version 30.00


I'd like to hear what you and others think about this re-purposing of the switch (HI - you do the management; LO - we do the management). Is this a good idea? Or not? Is it a good idea but you also need a lower LO? If you can't have both which is preferable?

Am none of your test riders, but have a pretty strong opinion on such limiting means. Of utmost importance to me is to have a very smooth and linear throttle control, plus of course a controller that will pump current into the motor accordingly. That way there is no real need for such a switch-induced self-restraint. I can use a smooth throttle from walking pace up to full blast and keep basically every conceivable constant speed in between, all within the travel of my right wrist. If you think a HI/LO setup is necessary I would stick to your formula "HI - you do the management; LO - we do the management", that latter being limited to typical city riding speeds around 30mph/50km/h. But ZEV's 3-step "gear shift" regime is in my opinion either proof of very poor controller "controllability" / jumpy throttle or some kind of bow to the gear-shifting habits of ICE-motorcyclists and thus tarnishing the beauty of the excellent torque / speed characteristics of PM-synchronous electric motors - a "back to the future" kind of gadget. Not necessary if throttle and controller work together smoothly.

4. The regen braking now shuts off at about 5-7 mph, instead of continuing all the way down to virtually stopped. I assume that the main purpose of this modification was to eliminate the harmless but unappealing "thunking" of the Kelly seeming to not quite "know what to do" with the small regen current from the motor at very low speed. I developed a habit of simply letting off of the levers momentarily below 5mph when braking to a stop. This modification eliminates this, but unfortunately it also eliminates being able to use regen while slowly descending a hill due to backup at a traffic light, rough road, or other conditions. So maybe another minor tweak for v30.01.

Yup that's exactly why we did this. Ultimately I hope to provide the rider the ability to chose regen strategy themselves. How much on front brake? How much on rear brake? Cut off at low speed? etc. I think this is an area of great subjectivity which makes it difficult for Current Motor to "get right" for everyone.

I am not sure why you see such a big problem with this point: an experienced motorcyclist knows instinctively if and how much he can use the rear brakes, particularly in hard braking. Thus only one brake lever should activate regen: the one for the rear brake. Because regen only brakes the rear wheel, right? I know I may be the odd one out, but I ride my bike in winter too. That is why I strongly prefer a clear distinction between front and rear brakes. I am not sure yet how comfortable I will be with the instant regen on both my brake switches when the roads get slippery. My ideal would be the following: Regen only on the brake lever for the rear, some hydraulic slack in the lever and over that little bit of travel continuoulsy increasing regen depending on how much I pull on that lever. Not easy to accomplish, but the most intuitive way of doing it, in my opinion. I know the grinding sound of the Kellys with regen to standstill, but I prefer this over a sudden drop due to regen being turned off below a certain speed. The compromise would be to keep regen to standstill if braking was initiated above say 10km/h, and not to activate regen if braking is initiated below this speed.

My rides:
2017 Zero S ZF6.5 11kW, erider Thunder 5kW

Derby
Derby's picture
Offline
Last seen: 6 years 11 months ago
Joined: Friday, October 7, 2011 - 06:56
Points: 74
Re: BCU Software Version 30.00

As a newcomer to the EV field, (one who is waiting on a status report regarding his on-order Red C124) I can only guess as to what would make the most sense to me. Once I get my Maxi-Scooter and get to ride it, I could change my mind. In the meantime:

In "hi" we put the onus on the rider to decide how to manage his/her energy usage. We only put cutbacks when things get too hot. In "lo" we take the view of trying to give the rider the a "pre-programmed" balance between performance and range - that is to say we back off accel slightly and we back off at higher speeds. The reduced accel will hardly be noticed - it only applies between about 20mph and 40mph - however, if you were to look at a graph of the current draw you'd see that we've "lopped off" a significant current peak for very little impact in actual acceleration and even less impact on perceived acceleration. At the top end - you should find that in LO you top out around 55mph whereas in HI you will top out around 60 to 65 (on a C124). It can take a long time to get to those top speeds.

This sounds like what I would want. My type of driving (mostly short trips in a city) would most likely be done almost always in LO but it would be nice to have HI when needed.

The compromise would be to keep regen to standstill if braking was initiated above say 10km/h, and not to activate regen if braking is initiated below this speed.

I like the compromise suggested by MEroller.

PJD
PJD's picture
Offline
Last seen: 1 week 3 days ago
Joined: Wednesday, November 22, 2006 - 05:44
Points: 1416
Re: BCU Software Version 30.00

Just so you know, Current scooters have three regenerative braking levels, a low level - low enough to use in a lot of downhill coasting situations - i.e. barely perceptible - activated with the right lever; a medium level, suitable for a gradual slowing stop at a traffic light or steep hills, actuated by the left lever, and a high level, suitable for moderately abrupt traffic light stops, actuated by squeezing both levers. Regen braking causing a rear wheel skid is unlikely, because the regen braking force is a function of wheel rpm, an abrupt reduction in wheel rpm leads to an abrupt reduction in braking force - the opposate of friction brakes; at zero wheel rpm, there is no regen force. So, regenerative braking is inherently "anti-lock".

Having a variable regen with a brake lever would be difficult to implement. It would require a rather specialized sensor at the brake lever.

dgalb100
Offline
Last seen: 8 years 8 months ago
Joined: Sunday, January 30, 2011 - 18:50
Points: 20
Re: BCU Software Version 30.00

First on the hi-low switch, I don't see a need for it at all once you have a LCD screen. Once I was able to accurately track the temps, the SoC and other numbers, I don't want the bike trying to out guess me... Pre-LCD, I can see that might be useful because you don't accurately know those numbers so you want the bike to error on the side of caution and get you the max range but not run out, you don't know how hot the motor is, so you want the bike to protect you so you don't overheat in the middle of the big climb you will have coming up in 3 miles. Once you can see the temps are climbing high, or the SoC is getting lower than you are expecting, its easy to adjust your riding style and baby the bike until you get over the climb and things get back to where you want them. Thats my experience having a LCD screen where I can see the status of everything.

Now, as to Re-Gen braking.... If I could wave my wand, and just get what I want... (I suspect this method is patented, but...) The first 20% (maybe 10%, 5% have to play with it) would act like a % regen selection... this would mimic engine braking on a ice bike... At greater than 20% throttle, you should get no regen, and if you are going faster than 20% throttle then you get a coast much like 0% throttle is now. As you bring the throttle down lower the regen braking should increase until when you are at 0% throttle, regen is at X% (would need to play to see what % that is) Basically, I'd like it to mimic engine braking at 20% throttle and go to downshift/engine brake at 0% throttle. So if you want to coast like an electric, you'd need to hold the throttle slightly on. This regen/engine brake is followed even if the brakes are on.. So brakes on, but throttle slightly engaged with back of the regen from 100% to 80 to 70% and so on... so to get 100% regen you'd let go of the throttle and touch both brakes...

If you pull the brakes, it does what it does now. Currently pulling the front brake in just until regen engages is like downshift engine braking on a normal motorcycle (with automatic anti-lock), so I don't mind the pull the front brake currently gets 30%(?) regen on the rear wheel. In the above method, because the throttle is off you are already at 30-35% regen, front brake adds another 0-10% to 30-45% total regen, back brake adds 55-70% and you're at 100% regen if throttle is at 0%, and front/rear brakes are engaged. You could even make front brake do no engine braking depends on how it feels... What would be cool is all of these numbers are selectable... so I can set throttle 20,10,0% regen numbers and front/back brake regen numbers... That way I can tune the bike so it works the exact way I want it to.... :)

At slow speed... As you get slower and slower, regen drops down. (So at say 15mph, regen is down to 60%, 10mph 30% and so on... just making numbers up, but you get the idea), the regen should drop off little by little until where it kicks off (5mph) you won't even feel it let go because you are already using the 90% disk brakes to slow the bike. Smmmooooooth should be the idea, you shouldn't ever feel regen "TURN ON" or "TURN OFF"... it should just smoothly engage and disengage... Here I can set the speed regen parameter numbers, so I can say 20mph regen=-20%, 15mph regen=-40%, 10mph regen=-60%, 7mph regen=-80%, 5mph regen=-100%. Then current could set what the defaults are, and a user can program whatever values they want and upload them into the BCU.

My goal of these ideas is that regen shouldn't be a noticeable force when it initially comes on, or when it turns off. Right now when I touch the back brake I can really feel the regen come on. That in my opinion is a problem (minor, but maybe solvable), the braking comes on to strong and you can actually feel the bike "pulling" slower, more like you've down shifted two gears and the engine is revving hard sort of braking, not something I want to be an all/nothing type switch...... a light touch on the brakes should barely be noticeable... like the way it currently does with the front brake... you feel a drag, you know its there, but its like you just let the throttle off...

Using my described method, already having the regen slightly on 30-40% (because the throttle is at 0%) means that pulling the back brake goes from 30-70% regen which isn't near as much of a jump as 0-70% and as the bike slows, dropping the regen down means you compensate by pulling the brakes more (but its small enough you wouldn't even notice you're having to compensate) and when the regen finally drops off you are already 80% on the disk brakes anyway so you don't feel a surge forward when the regen turns off and you have to quickly grab more brake to keep the bike smoothly stopping...

Hopefully I've described the system I'd like to see. of course once I see it I'll probably change my mind and attempt to redesign the entire thing again... Hopefully if I'm given the parameters I can tweak it until it does what I want. :) If we can't get throttle activated regen I'd at least like to see the slow speed regen idea where the regen "ramps down" as you go slower until it hits 0... so Regen=100,80,60,40,20,0 sort of process as you go from 20mph to 0mph... instead of the 100,0 at 7mph (numbers made up but you get the idea)... Does this mean you lose out on some regen? yes... so you lose out on 1% of the 1% extra range regen added (numbers are made up). With the parameters you can tweak it till you get the balance you want of regen vs. bike feels the way you want it to feel...

Laters,
-d.

jdh2550_1
jdh2550_1's picture
Offline
Last seen: 10 years 11 months ago
Joined: Tuesday, July 17, 2007 - 09:35
Points: 2335
Re: BCU Software Version 30.00

John, this is only partly related to the BCU software and partly related to hardware, but it seem to me that the most urgent improvement you need to address is the charging system.

What I am finding is happening, particularly in cooler weather, is that the bulk charging is ending, then the balance mode is taking a very long time, 2 hours or more, to get those last couple AH in the pack. This has nothing to do with balancing, but rather has to do the bulk mode cutting off then the cell voltages dropping way back down to 3.35-3.38 or so, resulting in a long period of charging at the 0.5 amp balance charge rate to get back to the full cell voltage at about 3.65 volts per cell or a pack voltage cutoff at about 88 volts. I assume the cells are well matched, so little or no balancing is needed or occurring. But, the balance mode takes excessively long time because of the long recovery in voltage after bulk charging ends. And this effect is going to get worse as the temperature drops and cells internal-resistance rises.

The problem is that your current charging protocol abruptly switches to a low-current mode that doesn't swamp the balancer, so it does not follow the ideal CC-CV protocol that charges LiFePO4 type cells most quickly and efficiently.

One solution would be to go to a three or four stage charging that would mimic CC-CV a little more closely - say 8-10 amp bulk mode, 4 amp, then 2 amp "top-off" modes, then 0.5 amp balance mode. I don't know how easy this would be to implement.

Paul - I'll make sure Erik looks at this. Now might be a good time to consider this update.

The other change is finding a way to shut off the charger fan when charging is complete.

Fan on all the time was a "temporary hack" to ensure the longevity of the chargers. The latest batch from the factory now switch off. Also, we're contemplating different chargers (hence your timely input).

By the way, with charging stopped, but still plugged in and the charger door left open, how much current is being drawn from the pack?

I'll get you a figure later today.

John H. Founder of Current Motor Company - opinions on this site belong to me; not to my employer
Remember: " 'lectric for local. diesel for distance" - JTH, Amp Bros || "No Gas.

jdh2550_1
jdh2550_1's picture
Offline
Last seen: 10 years 11 months ago
Joined: Tuesday, July 17, 2007 - 09:35
Points: 2335
Re: BCU Software Version 30.00

Am none of your test riders, but have a pretty strong opinion on such limiting means. Of utmost importance to me is to have a very smooth and linear throttle control, plus of course a controller that will pump current into the motor accordingly. That way there is no real need for such a switch-induced self-restraint. I can use a smooth throttle from walking pace up to full blast and keep basically every conceivable constant speed in between, all within the travel of my right wrist. If you think a HI/LO setup is necessary I would stick to your formula "HI - you do the management; LO - we do the management", that latter being limited to typical city riding speeds around 30mph/50km/h. But ZEV's 3-step "gear shift" regime is in my opinion either proof of very poor controller "controllability" / jumpy throttle or some kind of bow to the gear-shifting habits of ICE-motorcyclists and thus tarnishing the beauty of the excellent torque / speed characteristics of PM-synchronous electric motors - a "back to the future" kind of gadget. Not necessary if throttle and controller work together smoothly.

Hi MEroller - thank you for taking the time to comment. You're input (and everyone else's) is definitely most welcome.

I'll let other rider's comment but I believe we have a very good throttle response. Far more control than on the low-end bikes. Not quite as good as a Vectrix on the low end - but as good as it at the higher end.

Yes, I think I'm coming to the conclusion that a more restrictive LO might be more useful.

I am not sure why you see such a big problem with this point: an experienced motorcyclist knows instinctively if and how much he can use the rear brakes, particularly in hard braking.

I think you've answered your own question. It's a problem because not every motorcyclist is experienced - and we're aiming for the mass-market and lots of new riders.

Thus only one brake lever should activate regen: the one for the rear brake. Because regen only brakes the rear wheel, right? I know I may be the odd one out, but I ride my bike in winter too. That is why I strongly prefer a clear distinction between front and rear brakes. I am not sure yet how comfortable I will be with the instant regen on both my brake switches when the roads get slippery. My ideal would be the following: Regen only on the brake lever for the rear, some hydraulic slack in the lever and over that little bit of travel continuoulsy increasing regen depending on how much I pull on that lever. Not easy to accomplish, but the most intuitive way of doing it, in my opinion. I know the grinding sound of the Kellys with regen to standstill, but I prefer this over a sudden drop due to regen being turned off below a certain speed.

And that's the other part of your answer - you have a very strong idea of what you want. Which is different than some others. So, if I make it user selectable then both folks can be happy.

The compromise would be to keep regen to standstill if braking was initiated above say 10km/h, and not to activate regen if braking is initiated below this speed.

I'm not sure that's a very effective compromise.

John H. Founder of Current Motor Company - opinions on this site belong to me; not to my employer
Remember: " 'lectric for local. diesel for distance" - JTH, Amp Bros || "No Gas.

jdh2550_1
jdh2550_1's picture
Offline
Last seen: 10 years 11 months ago
Joined: Tuesday, July 17, 2007 - 09:35
Points: 2335
Re: BCU Software Version 30.00

First on the hi-low switch, I don't see a need for it at all once you have a LCD screen. Once I was able to accurately track the temps, the SoC and other numbers, I don't want the bike trying to out guess me... Pre-LCD, I can see that might be useful because you don't accurately know those numbers so you want the bike to error on the side of caution and get you the max range but not run out, you don't know how hot the motor is, so you want the bike to protect you so you don't overheat in the middle of the big climb you will have coming up in 3 miles. Once you can see the temps are climbing high, or the SoC is getting lower than you are expecting, its easy to adjust your riding style and baby the bike until you get over the climb and things get back to where you want them. Thats my experience having a LCD screen where I can see the status of everything.

However, I don't think most riders will monitor things nearly as closely. If they do, then they can just leave it in HI.

Now, as to Re-Gen braking.... If I could wave my wand, and just get what I want... (I suspect this method is patented, but...) The first 20% (maybe 10%, 5% have to play with it) would act like a % regen selection... this would mimic engine braking on a ice bike... At greater than 20% throttle, you should get no regen, and if you are going faster than 20% throttle then you get a coast much like 0% throttle is now. As you bring the throttle down lower the regen braking should increase until when you are at 0% throttle, regen is at X% (would need to play to see what % that is) Basically, I'd like it to mimic engine braking at 20% throttle and go to downshift/engine brake at 0% throttle. So if you want to coast like an electric, you'd need to hold the throttle slightly on. This regen/engine brake is followed even if the brakes are on.. So brakes on, but throttle slightly engaged with back of the regen from 100% to 80 to 70% and so on... so to get 100% regen you'd let go of the throttle and touch both brakes...

If you pull the brakes, it does what it does now. Currently pulling the front brake in just until regen engages is like downshift engine braking on a normal motorcycle (with automatic anti-lock), so I don't mind the pull the front brake currently gets 30%(?) regen on the rear wheel. In the above method, because the throttle is off you are already at 30-35% regen, front brake adds another 0-10% to 30-45% total regen, back brake adds 55-70% and you're at 100% regen if throttle is at 0%, and front/rear brakes are engaged. You could even make front brake do no engine braking depends on how it feels... What would be cool is all of these numbers are selectable... so I can set throttle 20,10,0% regen numbers and front/back brake regen numbers... That way I can tune the bike so it works the exact way I want it to.... :)

At slow speed... As you get slower and slower, regen drops down. (So at say 15mph, regen is down to 60%, 10mph 30% and so on... just making numbers up, but you get the idea), the regen should drop off little by little until where it kicks off (5mph) you won't even feel it let go because you are already using the 90% disk brakes to slow the bike. Smmmooooooth should be the idea, you shouldn't ever feel regen "TURN ON" or "TURN OFF"... it should just smoothly engage and disengage... Here I can set the speed regen parameter numbers, so I can say 20mph regen=-20%, 15mph regen=-40%, 10mph regen=-60%, 7mph regen=-80%, 5mph regen=-100%. Then current could set what the defaults are, and a user can program whatever values they want and upload them into the BCU.

My goal of these ideas is that regen shouldn't be a noticeable force when it initially comes on, or when it turns off. Right now when I touch the back brake I can really feel the regen come on. That in my opinion is a problem (minor, but maybe solvable), the braking comes on to strong and you can actually feel the bike "pulling" slower, more like you've down shifted two gears and the engine is revving hard sort of braking, not something I want to be an all/nothing type switch...... a light touch on the brakes should barely be noticeable... like the way it currently does with the front brake... you feel a drag, you know its there, but its like you just let the throttle off...

Using my described method, already having the regen slightly on 30-40% (because the throttle is at 0%) means that pulling the back brake goes from 30-70% regen which isn't near as much of a jump as 0-70% and as the bike slows, dropping the regen down means you compensate by pulling the brakes more (but its small enough you wouldn't even notice you're having to compensate) and when the regen finally drops off you are already 80% on the disk brakes anyway so you don't feel a surge forward when the regen turns off and you have to quickly grab more brake to keep the bike smoothly stopping...

Hopefully I've described the system I'd like to see. of course once I see it I'll probably change my mind and attempt to redesign the entire thing again... Hopefully if I'm given the parameters I can tweak it until it does what I want. :) If we can't get throttle activated regen I'd at least like to see the slow speed regen idea where the regen "ramps down" as you go slower until it hits 0... so Regen=100,80,60,40,20,0 sort of process as you go from 20mph to 0mph... instead of the 100,0 at 7mph (numbers made up but you get the idea)... Does this mean you lose out on some regen? yes... so you lose out on 1% of the 1% extra range regen added (numbers are made up). With the parameters you can tweak it till you get the balance you want of regen vs. bike feels the way you want it to feel...

Laters,
-d.

This is great stuff! I need to read it in more detail - but I think some of it might be doable (although not all). The "user programmable" part would fit in quite well because I feel pretty sure some riders won't like "crack open the throttle to coast" (so they could switch off that aspect).

Thanks everyone for the feedback.

John H. Founder of Current Motor Company - opinions on this site belong to me; not to my employer
Remember: " 'lectric for local. diesel for distance" - JTH, Amp Bros || "No Gas.

ErikK
Offline
Last seen: 11 years 1 month ago
Joined: Sunday, July 13, 2008 - 21:16
Points: 10
Re: BCU Software Version 30.00

Paul -
Your suggestion for tapering off charging amps, instead of switching directly from about 10 to 0.5 is a good one. We can't do it with our current charger, but it's something I will keep in mind as we work on new chargers.

You are correct that it only affects the last couple Ah, and doesn't affect the overall efficiency of the system much. It just takes longer.

With the charging stopped, but the charge door left open, draw from the pack is under 0.1 amps. I just measured it again to verify. And if the charger is plugged in it is drawing something like 11 watts from the wall. Not as good as zero, but not horrible either.

- Erik

-- Erik Kauppi, Chief Engineer, Current Motor Co --

MEroller
MEroller's picture
Offline
Last seen: 11 months 1 week ago
Joined: Monday, September 26, 2011 - 09:24
Points: 847
Re: BCU Software Version 30.00
I am not sure why you see such a big problem with this point: an experienced motorcyclist knows instinctively if and how much he can use the rear brakes, particularly in hard braking.

I think you've answered your own question. It's a problem because not every motorcyclist is experienced - and we're aiming for the mass-market and lots of new riders.

OK. In that case the best thing would be a true combination ABS braking system, including regen if you like. One brake lever activates both brakes with automatic ABS break force balancing between front and rear, and the other lever activates only the front brake, with ABS too. But that would blow the bike's cost way over the top.
Mass market and new riders will most likely be intimidated by such a powerful E-scooter, so there is some real danger in this big spread of your target customers and your bike's performance. I thus understand your concern for finding the best possible solutions to this dilemma...

My rides:
2017 Zero S ZF6.5 11kW, erider Thunder 5kW

jdh2550_1
jdh2550_1's picture
Offline
Last seen: 10 years 11 months ago
Joined: Tuesday, July 17, 2007 - 09:35
Points: 2335
Re: BCU Software Version 30.00

Mass market and new riders will most likely be intimidated by such a powerful E-scooter, so there is some real danger in this big spread of your target customers and your bike's performance.

That's what we have the "LO" switch for :-) (I'm coming to the conclusion that we want to make "LO" somewhere between where it was (too low) and where it is now (not enough differentiation with high))

Thanks for everyone's input - I really appreciate it.

John H. Founder of Current Motor Company - opinions on this site belong to me; not to my employer
Remember: " 'lectric for local. diesel for distance" - JTH, Amp Bros || "No Gas.

adstriegel
Offline
Last seen: 12 years 1 month ago
Joined: Saturday, October 15, 2011 - 19:39
Points: 10
Re: BCU Software Version 30.00

Cool, now I can put my new programming adapter that came last week to good use when I get back from traveling.

I would second the comment on the fan. I don't mind it terribly but my wife has commented that it is quite loud and runs all night when plugged in. Glad to hear the new ones resolve that.

Related to the LCD thread and I'll drop a comment there, when it is charging, it would be nice to show an estimated range / time left to charge. Some sort of an X:XX estimated time left until charge. Some sort of an animation / graphic while charging would be great.

PJD
PJD's picture
Offline
Last seen: 1 week 3 days ago
Joined: Wednesday, November 22, 2006 - 05:44
Points: 1416
Re: BCU Software Version 30.00

John,

I spent a few days riding to be sure this is the case, but on my C124 anyway, it appears that the V.0.0.30 regen cutoff did not entirely fix that bit of inelegant thunking/sputtering from the motor upon coming to a stop. The regen audibly and palpably cuts off at 5-7 mph, but a certain amount of that sputtering (but it seems less than before) still occurs with the last revolution or two of the wheel before stopping. It seems to be a bug in Kelly controller? This is purely an aesthetic, not a performance, reliably or safety issue, but I thought I'd bring it up.

BTW, with regard to the charging issue, I'd also love to have a charger that shuts off the fan upon completion. Are the newer chargers showing enough robustness to consider turning the charging current back up to 10 amps? Speed of charging can be considered as important an aspect of overall performance as how fast the scooter goes.

jdh2550_1
jdh2550_1's picture
Offline
Last seen: 10 years 11 months ago
Joined: Tuesday, July 17, 2007 - 09:35
Points: 2335
Re: BCU Software Version 30.00

John,

I spent a few days riding to be sure this is the case, but on my C124 anyway, it appears that the V.0.0.30 regen cutoff did not entirely fix that bit of inelegant thunking/sputtering from the motor upon coming to a stop. The regen audibly and palpably cuts off at 5-7 mph, but a certain amount of that sputtering (but it seems less than before) still occurs with the last revolution or two of the wheel before stopping. It seems to be a bug in Kelly controller? This is purely an aesthetic, not a performance, reliably or safety issue, but I thought I'd bring it up.

BTW, with regard to the charging issue, I'd also love to have a charger that shuts off the fan upon completion. Are the newer chargers showing enough robustness to consider turning the charging current back up to 10 amps? Speed of charging can be considered as important an aspect of overall performance as how fast the scooter goes.

Hi Paul,

Yes, I'm noticing the same regen behavior - but not on all setups. I will try tweaking some of the Kelly settings to see if I can get rid of this. Annoying - but not critical. I also intend to ramp out the regen at slow speed rather than the current step change and also add in some hysteresis. This should just make for a smoother implementation.

Charger work is coming along. We're actually investigating a fanless solution Yes, we will go back up to a full 10A with the replacement solution. The step down to 8A is a "quick fix" while we source, test and integrate a better charger. BTW, if we add RTC hardware (see the LCD page) we can also add the ability for timer control of the charger. There's lots and lots of cool stuff we can still think to implement! (After we get the important stuff done first)

John H. Founder of Current Motor Company - opinions on this site belong to me; not to my employer
Remember: " 'lectric for local. diesel for distance" - JTH, Amp Bros || "No Gas.

Iccarus
Iccarus's picture
Offline
Last seen: 6 years 9 months ago
Joined: Saturday, December 20, 2008 - 08:05
Points: 347
Re: BCU Software Version 30.00

[quote jdh]Yes, I'm noticing the same regen behavior - but not on all setups.

John, On my KBL 12251H (on my xm3500li)I had the big round multipin connector loosen up slowly over time. While this was happening I noticed that both the throttle was getting more touchy and the regen brakes were getting rough and seemed to be slipping and making the noise you that is being described. I finally noticed as things slowly got worse that it was the connector. I noticed while I was connecting my laptop to the controller to try playing with the settings :) That connector will loosen up if you don't make it real tight. This might save you time if a customer calls with this problem. With my fitting tight everything is so smooth, for me at least, the 2 speed switch is unnecessary. This happened to me more than once,( the loosening) so be advised.
John, I admire your patience, keep up the good work.
Bill

2008 XM3500li Mods/Kelly KBL12251/84v 28cell 40AH pack/ Variable regen brake trigger on left brake handle/Givi/Cycle Analyst/Homemade BMS

KMX Typhoon Home build (recumbent pedelec) with two Astro Brushless 3220motors/twin castle Phoenix ICEHV 160/ Cycl

Iccarus
Iccarus's picture
Offline
Last seen: 6 years 9 months ago
Joined: Saturday, December 20, 2008 - 08:05
Points: 347
Re: BCU Software Version 30.00

Having a variable regen with a brake lever would be difficult to implement. It would require a rather specialized sensor at the brake lever.

You mean like this?

DSC01310.JPG

You set your BRK/SW regen at what is comfortable for you. Then set the BRK/AN (brake analog) to 0%-100%.
This way you keep your hands ready on the handle grips while working the variable trigger. A slight pull on the either brake handle gives you gentle braking , then you can adjust with fingertip ease up to 100% (with index finger). I've been using this for a couple years now and I really like it. I wish I could patent it. That's a 0v-5v thumb throttle. A more professional installation would look better but it's a prototype, so it doesn't have to be pretty.

Bill

2008 XM3500li Mods/Kelly KBL12251/84v 28cell 40AH pack/ Variable regen brake trigger on left brake handle/Givi/Cycle Analyst/Homemade BMS

KMX Typhoon Home build (recumbent pedelec) with two Astro Brushless 3220motors/twin castle Phoenix ICEHV 160/ Cycl

Log in or register to post comments


Who's online

There are currently 0 users online.

Who's new

  • xovacharging
  • stuuno
  • marce002
  • Heiwarsot
  • headsupcorporation

Support V is for Voltage