wr250f conversion

[dougf]

Another small update, I got the motor temp working....sort of. It looks to be pretty inacurrate and it looks like the coeff for it is unprogrammable. Chucked it onto the display, will see what it does. I also worked out the current and watts readout from the controller. They were a bit tricky to spot because they are essentially negative when drawing power and positive when regen is flowing back into the battery. Also chucked those onto the screen.

I havent played around with the regen much, mostly because it doesnt really work properly as engine braking. Tends to hang and not re- accelerate when giving more throttle. I though I could make my own version of it by simply reading the throttle ADC and applying the low brake signal when I have > 0 rpm on the motor and my throttle is closed but I dont think that will work either because from some testing it looks like if you activate the low brake, bring the throttle on and then release the brake, the controller wont come out of braking until the throttle is lowered back to 0%....Ill think about it some more.

Does anyone properly understand the difference between "stop back current" and "max back current"? The manual is a bit vague, is it something like the maximum back current is how hard the motor is braked and the stop back current is the limit of recharge into the battery? Recommends having the max setting 25-50% higher then the stop setting. I guess if the motor is generating more power then you want to put into the battery it can just shunt the windings together and heat the motor instead. I could be wrong, hopefully someone actually knows.

I didnt have the balls to try changing the current settings while riding yet. Bit worried it might just really confuse the drive and burn something out. I know from testing that I can safely change the throttle mode (eco, lin, sport) while the bike is stationary without any issues. I could try PWM on the low brake but Im pretty sure that wont work....hmm how much do I want variable regen I suppose. Anyone actually used variable regen on here and care to chime in?

Doug

 

[j bjork]

Another small update, I got the motor temp working....sort of. It looks to be pretty inacurrate and it looks like the coeff for it is unprogrammable.

Same conclusion I made :wink:
I think it wasnt exactly the right sensor you can choose. I think you couldn't set your own temp limits either.

I havent played around with the regen much, mostly because it doesnt really work properly as engine braking. Tends to hang and not re- accelerate when giving more throttle.

That has worked for me since I updated the controller.

Does anyone properly understand the difference between "stop back current" and "max back current"?

What I have noticed is if you set "max" lower than "stop" the brake wont release until you reach that level.
So it will keep braking when you release the brake.

Anyone actually used variable regen on here and care to chime in?

I have, and I think it is very useful. I think it is easier to regulate than the foot brake, and I dont mind that the pads last longer.
I think it is too weak at slow speeds on this controller though.
It was very nice on nucular, and I think I was able to set it to work well on sevcon too.

 

[dougf]

I actually did some digging around facebook and made some changes. Turns out if you do what the manual says and set max back current 25-50% higher it all just works. Max is a safety limit, stop is the actual setting. I will look into variable regen then.

On the dyno today, just did some tweaks to the dyno software because it couldnt handle the torque :lol: . Recharging now but looks like i am pegging at 500A most of the way to 100kph and that means around 30kw at the wheel till about 75kph, then it tapers off to about 22kw at 110kph. Just topping up battery now, will try some different setting and see what happens.

Doug

 

[dougf]

Ok, have had a look at the data I collected this morning. A couple of things stand out. My bike is now making (on my dyno) around 30kW at the wheel, probably from 0 rpm, my dyno cannot measure below 30kph, the brake wont take that much torque at those speeds) up to 70-75kph, then slowly tapers off to about 17kW at 120kph. The last time I had it on the dyno with the APT 96800 it made around 18kW and tapered off to about 12kW so a significant improvement. I did some testing with different battery current limits and as suspected going from 400 to 500A yeilds a near linear improvement in power, but going from 500 to 600 only gives a couple of extra kw. I can also see this in the current draw, basically pegs at 500A but only gets to ~540A with the limit set to 600A.
I am starting to sag pretty hard at more then 500A so with better batterys it should be capable of a bit more, especially if starting at 28S or 30S.

The three traces on my graph, the gray is my initial test at 500A with the depth of field weakening going as deep as 50% @ 150% of rated rpm, linearly.
Orange trace was one of the tests I did with the battery at 600A limit, can see its a bit higher on the up slope. looks abit funky as i had some dyno oscillations.
Blue trace is 500A with depth of field weakening going to 50% at 175% rated rpm. Definitely flattened out the tail a bit.

I had some trouble with my dyno coping with the large changes in torque so I had to be a bit gentle bringing it in, I also had some oscillations at times. The data is a bit messy. I didnt have much time.

The controller was reporting about 40kW of battery power in most cases, bit higher at 600A. I have some screen records of the phone apps if you want to see them.

My guess is this motor will take more abuse. It got a bit warm after 8 or so runs. I think the depth of field weakening could be stretched out quite a bit more. The controller always calculates a rated rpm of ~2900 rpm but Im pretty sure the KV of these motors is around 50 to 55 which would be a bit higher. Regardless, that will only add power up top, ~30kW at the wheel up to 75kph is not a slow bike .

Doug

 

[j bjork]

Alright, very interesting :thumb:

Too bad you cant get the lower rpm:s, it would be interesting to see what changes to the % could do there.
I dont really understand what your settings are now, did you start with 50% from 4500rpm but then changed to 50% from 5000rpm?
Do you have 100% under that rpm?

Did you notice if the temp reading on the motor seemed to follow fairly?

How about phase A, did you make any observations? I suppose much of those settings are already passed at 30kpm and you start hitting battery A limit almost right away.

Yes, your screen recordings would be very interesting to see :wink:

 

[dougf]

I might be able to modify my dyno to get a bit lower but its a pretty big eddy brake rated for 200kW and down at low speed it doesnt respond like it does faster. I need to program in a switching PID profile with rpm to stabilize it but I use it so little it just doesnt get any attention. Im pretty sure the torque and power shoots up to basically 30kw right away and then tapers off.

See attached screen shot of my %. Basically what I mean is when I get to what the controller sets at the rated rpm, I start to reduce the % linearly at a rate as to be at 50% when the RPM is at 175% of the rated rpm. Setting this up this way has definitely reduced my mid range, I dare say I could stretch it out further but the problem I had was surging and if I pushed it hard on the road I would get stuttering caused by MOE errors. Like this it's nice and smooth, no MOE errors or stuttering and I think its also a bit more efficent, pretty sure I only used about 55-60% battery riding to work today when I usually use about 70%.

Motor temp, didnt watch it too closely, hard to tell if its accurate because Im guessing the heat inside the motor is a bit higher then the outside. The outside casing probably got to maybe 70^C. Uncomfortable to keep you hand on. Still jumps around.

Didnt get as much time on the dyno as Id have liked today, I planned on locking the dyno at each RPM step and slowly upping the % until it didnt make any more power or it started to MOE error. A dangerous game .

https://www.youtube.com/shorts/uypPLTVY_QA

You can see Im being pretty gentle with the throttle, the torque increase rate outruns my dyno at these speeds.

Your YZ with the extra voltage, lower ISR batteries and lower gearing must be a widowmaker :lol: .

Doug

 

[speedy1984]

Hi there,Very impressed with the numbers for d.i.y off the shelf parts build tbh. I do wonder why the phase amps are that low though, at 550 dc i would of liked to see more phase amps than just double ish battery current.I thought it would nearly triple the battery current in phase amps.Also noticed that you are in field weakening very early seeing as the motor is 4100rpm rated.You did managed 4600rpm peak but as soon as you reached your very low rated controller rpm of course you are in fw. Maybe you could make it much more efficient at higher road or flat track speeds if you could get your rated rpm up.Did you auto learn with the controller in D drive mode and with the sprocket off? Im doing some testing with the 108 1800 model atm and can only get 3950rpm on the 138 v3 motor but on my 850 model i can get 7600 rated rpm.Speed ki is on 18 and 10 on this big controller but 9 and 10 on the smaller versions looking at my screen shots.I did manage 11.500rpm max on this controller though lol, but dropped it to a safer 10.000 after.I have also had the 530 model, all been very good apart from regen and on the lower phase amps side so interested on any of your findings :thumb:

 

[j bjork]

See attached screen shot of my %. Basically what I mean is when I get to what the controller sets at the rated rpm, I start to reduce the % linearly at a rate as to be at 50% when the RPM is at 175% of the rated rpm. Setting this up this way has definitely reduced my mid range, I dare say I could stretch it out further but the problem I had was surging and if I pushed it hard on the road I would get stuttering caused by MOE errors. Like this it's nice and smooth, no MOE errors or stuttering and I think its also a bit more efficent, pretty sure I only used about 55-60% battery riding to work today when I usually use about 70%.

So you used to have higher numbers, but then you got stuttering and lower efficiency?
I have higher numbers from 4000-7000 (70% most of the way there) At full throttle and higher rpm I got stuttering too, I lowered the max Ki/Kp and it helped. I dont think it solved it, maybe this is the problem.
Where are you at Ki/Kp now? I am at low Ki 1, mid Ki 2, high ki 3.
I went from 2 to 1 in low to see if it would help on cutouts I can get when trying to wheelie over obstacles.

Motor temp, didnt watch it too closely, hard to tell if its accurate because Im guessing the heat inside the motor is a bit higher then the outside. The outside casing probably got to maybe 70^C. Uncomfortable to keep you hand on. Still jumps around.

It looks like it is working in your video, I think it went a little over 100 degrees.
Probably works ok for protection then at least.

Didnt get as much time on the dyno as Id have liked today, I planned on locking the dyno at each RPM step and slowly upping the % until it didnt make any more power or it started to MOE error. A dangerous game .

Yes, but do you think you have to hold it long? I mean, it probably will show pretty quick if it starts oscillating.
And try the Ki/Kp settings too :wink:

You can see Im being pretty gentle with the throttle, the torque increase rate outruns my dyno at these speeds.

Yes, it would be interesting to see what the phase amps would look like if you could just hit full throttle at close to zero rpm.
Also what those % settings would do at low rpm.

Thank you for the info and screen recording, there is a lot of interesting info to see there
I think I saw close to 1100pA now, at a little over 2000rpm?
It keeps it kind of steady 8-900 then for a while, I guess it is with fw. Then starts to drop off.
I will have to look a few times more and see how it was :wink:

I wonder if the rated speed might be where max power should be without fw? In that case half rpm would make sense, and I think the kv should be about 60. I dont remember where I got it from, but it probably has something to do with programming sevcon..
I think sevcon started fw at about 3-3500rpm

Your YZ with the extra voltage, lower ISR batteries and lower gearing must be a widowmaker :lol: .

Doug

To me it feels really powerful, but I am not so sure about the batteries now.
I think I saw about 12v sag at about 550A draw in your video, I get that to 22m ohm.
That dosent sound bad at all, but I suppose the battery was warm by then?

I think I got my cells to about 1,2m ohm each at room temp, that would mean at least 36m ohm.
But I guess I would have to test at these kind of loads to see for sure what happens.

 

[dougf]

Hi there,Very impressed with the numbers for d.i.y off the shelf parts build tbh. I do wonder why the phase amps are that low though, at 550 dc i would of liked to see more phase amps than just double ish battery current.I thought it would nearly triple the battery current in phase amps.Also noticed that you are in field weakening very early seeing as the motor is 4100rpm rated.You did managed 4600rpm peak but as soon as you reached your very low rated controller rpm of course you are in fw. Maybe you could make it much more efficient at higher road or flat track speeds if you could get your rated rpm up.Did you auto learn with the controller in D drive mode and with the sprocket off? Im doing some testing with the 108 1800 model atm and can only get 3950rpm on the 138 v3 motor but on my 850 model i can get 7600 rated rpm.Speed ki is on 18 and 10 on this big controller but 9 and 10 on the smaller versions looking at my screen shots.I did manage 11.500rpm max on this controller though lol, but dropped it to a safer 10.000 after.I have also had the 530 model, all been very good apart from regen and on the lower phase amps side so interested on any of your findings :thumb:

Yeah I noticed that they looked low. Like I said, I had to gently ramp up, its possible the phase amps a significantly higher at lower rpms. I will screen cap a WOT launch on the road.

Ive done about 4 auto learns, all about 2900rpm. Yes, my current limits were maxed out, removed chain from motor, battery fully charged and in High gear. The motor actually spins up to about 4200rpm during autolearn in both directions. It was my understanding that field weakening wouldnt start until after about 4000rpm. However, like i said I would get stuttering and MOE errors with higher % up around 4000rpm. My PID parameters are very low right now, I will grab a screen shot for you later. I planned on trying to bring them up on the dyno but I believe with soo much rotating inerta it might mask oscillation that would occur on the road. I think with those settings sticking to whats in the manual is a safe bet.

Doug

 

[dougf]

https://www.youtube.com/shorts/x4SVVxAN9LA

Another capture with a higher soc battery.

I wasnt getting any MOE errors offroad, only on the road with road tires. I think for now its working great and Im just going to ride it for a bit. I have some niggling issues with the dash I need to fix. If I really want I can program in setting profiles into my dash for offroad, onroad which just reprogram the %.

Doug

 

[speedy1984]

Hi ya,yeah the Fardriver enters fw after it hits its max rated rpm,you can also see it in the logs.So you hit around 2900ish rpm and the W enters the log(Not trying to teach you to suck eggs btw lol)But yeah its just strange these bigger controllers run such low rated rpm and make the system run more inefficient than it should be.Also like you said easing in the throttle will make a huge difference in phase current but the second log you did hit full throttle voltage very quickly so looked very aggressive id you notice the full throttle voltage fluctuate in the log?Could that be something to do with the surging you feel at full power or max rpms ect?I have not seen mine do that if held fully pinned.Also at 585 amps and only dropping 12v is bloody impressive you must be happy with that for a self build pack,great work.Thinking though there must be more tuning to be had surely as its the average of both A and C Phase current numbers that give you your max reached current.So say its 1000 and you are running the the correct voltage and nearly 600dc i cant see another 200dc maxing it to the full 800dc helping it reach its 1500phase designed or rated numbers. Dont get me wrong this machine must feel incredible with those numbers you have achieved,i just hope you can chase down the numbers i believe this build deserves given the effort you have put in and of course we all enjoy the updates lol

 

[dougf]

Hi ya,yeah the Fardriver enters fw after it hits its max rated rpm,you can also see it in the logs.So you hit around 2900ish rpm and the W enters the log(Not trying to teach you to suck eggs btw lol)But yeah its just strange these bigger controllers run such low rated rpm and make the system run more inefficient than it should be.Also like you said easing in the throttle will make a huge difference in phase current but the second log you did hit full throttle voltage very quickly so looked very aggressive id you notice the full throttle voltage fluctuate in the log?Could that be something to do with the surging you feel at full power or max rpms ect?I have not seen mine do that if held fully pinned.Also at 585 amps and only dropping 12v is bloody impressive you must be happy with that for a self build pack,great work.Thinking though there must be more tuning to be had surely as its the average of both A and C Phase current numbers that give you your max reached current.So say its 1000 and you are running the the correct voltage and nearly 600dc i cant see another 200dc maxing it to the full 800dc helping it reach its 1500phase designed or rated numbers. Dont get me wrong this machine must feel incredible with those numbers you have achieved,i just hope you can chase down the numbers i believe this build deserves given the effort you have put in and of course we all enjoy the updates lol

Yeah I spotted the Weak replacing MTPA. Did a quick sanity check today and made sure the motor rpm was correct, it was.

Are you saying that on identical motors, the smaller version of the fardriver will report a different rated rpm after autolearn?

Maybe its something to do with the weak speed setting and the weak response setting. I can try turning weak response to none and running an autolearn again, same with the speed setting. Il give it a go later.

The bike doesnt surge at full power, it will just cut in and out, and it doesnt do it all the time. Would just occasionally feel it stutter a few times when WOT at around 80kph which is coincidentially not long after the rated rpm and weakening started happening. Doesnt do it at all with my current settings. No I think the surging at constant speed might still be the throttle. Ive made it better with some settings but it still does it a bit. ECO throttle basically fixes it but ECO throttle is horrificly bad. There is an eco throttle setting, maybe it can be improved? Also something to look into. So many settings on these drives that must do something but is not well documented.

I did use Molicels because they are a good blend of power density and energy density. The bike is more then powerful enough. Awesome offroad, pretty darn good onroad, it will leave my triumph in the dust probably all the way up to 80kph. Take off from the lights with some pace, look down and your doing 100kph, OOPS :lol: .

Doug

 

[speedy1984]

Hi ya,yeah i got i think 7650 on the 850 model and i think from memory 7500 from the 530 model.Both untouched settings but same auto learn process i.e set in D mode,no sprocket,26s pack at nom voltage and same motor.I have noticed though both the smaller units have a much much lower (speed ki) setting of 9 but the big model is 18.Ive not tested if it makes any difference just yet but its the only difference i can see from the other models that stick out.I only pointed out the throttle voltage on your runs as mine dont do that so may be of interest :thumb:

I used the p42 Molicel myself as i do really rate them even if they are a bit pricey.Mine is only 10p but 28s and at 350-370 i get about 9v sag.My build though is purely for slower Enduro stuff anyway so just short burst hill climbs not mx track stuff,although ive done a couple of events and was fine with low enough temps ect.


This was the first couple of test rides with mid power settings i would say and its a good amount more powerful now with nearly 34kw
https://youtu.be/hR96Low8e0k

 

[dougf]

Hi ya,yeah i got i think 7650 on the 850 model and i think from memory 7500 from the 530 model.Both untouched settings but same auto learn process i.e set in D mode,no sprocket,26s pack at nom voltage and same motor.I have noticed though both the smaller units have a much much lower (speed ki) setting of 9 but the big model is 18.Ive not tested if it makes any difference just yet but its the only difference i can see from the other models that stick out.I only pointed out the throttle voltage on your runs as mine dont do that so may be of interest :thumb:

I used the p42 Molicel myself as i do really rate them even if they are a bit pricey.Mine is only 10p but 28s and at 350-370 i get about 9v sag.My build though is purely for slower Enduro stuff anyway so just short burst hill climbs not mx track stuff,although ive done a couple of events and was fine with low enough temps ect.


This was the first couple of test rides with mid power settings i would say and its a good amount more powerful now with nearly 34kw
https://youtu.be/hR96Low8e0k

Hmm, its probably just me on the throttle while trying to keep the bike stable. Ive also modified my magura to reduce its twist by pitting a bolt into the housing and im probably just twisting the plastic a bit. The magura is definitely a noisy throttle. It was also the same on the apt controller. The cheap hall throttle i had originally was much nicer feeling albiet cheap and nasty. I got myself the matching triumph right hand switch assembly with its electric throttle with plans to use that but unfortunately it wont work with the standard WR front brake MC. It may also be digital as its a 6 wire thing and i havent looked too hard yet because it wont fit.

Your bike looks like it goes pretty good. I think most of the time my motor just converts the extra power to high velocity dirt. Its nice for high speed stuff. I picked the molicels because ive used them on model planes and the deliver better W/hr under load then similar packs like the 40T's or the 35E's and vtc6's. Same for the p26a's, pretty good cells.

Put the bike away for now, roads are getting busy now after the holidays and i dont much like riding in traffic. Ill pull it out this weekend and look into possible variable regen solutions.

Doug

 

[speedy1984]

Hi ya,so im maxing my motor out at 9500rpm and through my percentage range/Ratios in Speed i am 100% 7500 then 90% at 8000 then 40% at 8500 and 5% at 9000.Interestingly if i drop my rpm to any other gear mode i.e low or medium where they are set at max of 100% the rpms pulses like you talk about!In D mode its perfectly stable,this to me at least would suggest its the percentage amount at your max rpm number is set to high.You are still at 63% at 4500 and 52% at 5000 when you are maxed out at around 4700rpm from your log.I would drop your percentages/Ratios in Speed massively on these two numbers and i would bet that sorts the pulsing throttle.

Hope this helps fella as if its pulsing like mine did i could not stand riding at max rpm putting up with that lol :thumb:

 

[dougf]

Hi ya,so im maxing my motor out at 9500rpm and through my percentage range/Ratios in Speed i am 100% 7500 then 90% at 8000 then 40% at 8500 and 5% at 9000.Interestingly if i drop my rpm to any other gear mode i.e low or medium where they are set at max of 100% the rpms pulses like you talk about!In D mode its perfectly stable,this to me at least would suggest its the percentage amount at your max rpm number is set to high.You are still at 63% at 4500 and 52% at 5000 when you are maxed out at around 4700rpm from your log.I would drop your percentages/Ratios in Speed massively on these two numbers and i would bet that sorts the pulsing throttle.

Hope this helps fella as if its pulsing like mine did i could not stand riding at max rpm putting up with that lol :thumb:

Not quite sure i follow you there. Are you saying that on your bike, if you change from high gear to low, the bike starts pulsing because your % are now to high? That is to say, changing from high to low scales the % rpms? I dont think it works like that. Pretty sure all the lower gears do is let you program slower and lower power modes.

My bike doesnt surge, pulse, stutter, cut out or anything similar at high load. The subtle surge i feel is at low load constant speed like riding at 50kph. Im pretty sure its related to the pid loops, and given it didnt improve that much by lowering the gain, the gains are probably too low and the loops are poorly regulated. It gets a bit better in eco mode which probably means theres some feed forward in the system. That might be related to the vehicle matching parameter, LM or AN, dont remember which one it is.

Doug

 

[speedy1984]

Sorry I thought you said it was stuttering on the road at full speed and not holding a constant speed.Mine does this when in a lower gear where my percentage rate is at 100.At my full rpm where my percentage rate is at only 5 it holds full rpm without any small pulsing or surging at the motor.Must of got my wires crossed with your issue lol.

 

[dougf]

Update, changed my magura throttle for the triumph one i got. The triumph throttle has 2 independant hall sensors. Looks like one is for 3.3v and one for 5v. I just hooked up to the 5v one. 1v-4.5v swing. Sooooo much better, now silky smooth while cruising so it was just the throttle causing my issues. Only problem with the triumph throttle assy is the kill switch is closed in run. Problem to fix later. It also gives me a few more buttons to play with.

I had to machine a bracket to space the front brake master cyl away from the bars to fit it though. Ill need to pull it off, tidy it up and powdercoat it at some stage.

While i was digging in the wiring i sorted out my key situation. For ADR compliance the main contactor must open when the key is turned off. I was initially going to fit a different key with some aux contacts but it turns out that the stock WR key actually has both a NO and NC contact. Just needed to solder the wires to the back of the cylinder 8) .

New pcb drawn and ordered for the dash. Ive added a few more IO pins and another uart with the idea to have it work in a passthrough mode when i want to connect with my phone. Ive also added some ADCs for variable regen assuming it works and i put a DAC on there for possibly passing the throttle into the dash and having programmable throttle curves. This one i will maching a nice enclosure for and make it look proper. My lash up has served its purpose.

Put some new plastics on. Might start hunting on ebay for another tank. I think the fardriver might actually fit inside it without having to cut away the side shroud top mounts. Would look a lot nicer.

Doug

 

[j bjork]

Nice that you got the throttle problem sorted

Did you test any more settings? I dont know if weak response or weak character might be something, I dont think I have messed with them at all.

 

[stan.distortion]

For the regen, does anyone do add-on regen modules? The circuit looks relatively simple, a 3 phase rectifier with DC current regulation, I could probably even put one together within my (very limited) electronic hacking skills. Failsafes might be tricky but if something like that would be possible I'd imagine there's a reasonably big market for it, pretty much every controller with good price/performance only comes with on/off regen.

 

[amberwolf]

I don't know anyone that makes a module to do this, though the concept has been discussed a few times here on ES. One potential way (untested) is using a three-phase rectifier and a brushed motor controller with it's battery inputs fed by the rectified output (but with it's "ignition wire" (to run the LVPS) powered by the main battery so it remains "on" even when the motor is not spinning), and the motor output feeding the main battery charge port. (note that this method of regen can use the charge port of a battery whether or not it is tied to the discharge port, since the controller is not drawing any power from that port unlike the one that runs the motor).

Then the throttle for the brushed controller is the braking power control--no other control is required. As long as the brushed controller and the rectifier can handle even the worst-case highest voltage spikes that could possibly come out of the motor phases under any conditions, and both can handle the worst-case current flow that will happen under regen, and the battery and all of it's interconnects and BMS can handle the charge current from it, then it should work.

The big disadvantage is that it won't actively brake below the point at which BEMF-rectified is below the battery voltage (meaning once the speed drops below a certain point that is different for each system); for that you'd have to be using a three-phase controller actively switching the currents in the phases to counter rotation, whcih uses some power rather than regenerating it.

Anyway, it's a simple enough experiment if you have the rectifier and brushed controller that meet the voltage/current requirements.



But any (properly designed/programmed) FOC controller should handle variable regen gracefully, and not just have on/off regen, since regen is just negative current flow and should be handleable by the same functions that drive the motor with positive current flow. Some modification of the functions should be able to actively drive the current in a way that doesnt' regen but actively fights the BEMF of the spinning motor (perhaps something like as if it were trying to drive it in reverse) to create active braking that will force it to slow it even faster.

Non-FOC (especially cheap lookup-table-sine or trap) controllers don't implement variable regen very often, and even those that do don't usually brake all that well except at higher speeds, since they don't usually actively brake, they just rectify BEMF from the motor into DC voltage and thus can't really brake once below the battery voltage...some do FET switching to momentarily short the windings with the right timing to create spikes with the flyback effect in the phase coils, but most of the cheap stuff doesn't seem to do this (not the ones I've used, anyway).

 

[dougf]

Nice that you got the throttle problem sorted

Did you test any more settings? I dont know if weak response or weak character might be something, I dont think I have messed with them at all.

No, i havent had much time riding, just in the shed. I spotted some mention of the weak settings on facebook, can probably dig up some info from there

For the regen, does anyone do add-on regen modules? The circuit looks relatively simple, a 3 phase rectifier with DC current regulation, I could probably even put one together within my (very limited) electronic hacking skills. Failsafes might be tricky but if something like that would be possible I'd imagine there's a reasonably big market for it, pretty much every controller with good price/performance only comes with on/off regen.

Its definitely possible, but i dont know if it's worth it. I do like the regen braking, i basically only use my brakes to turn on the brake light but i think there are a reasonable selection of drives on the market that support some form of regen that it isnt an issue. The old APT drive had 2 levels of regen, throttle release and brake signal. This was pretty good because i find i dont want large amounts of engine braking while riding but when i brake i want the motor to do most of the braking. Fardriver doesnt have this, only the option to choose how regen is initiated.

I will try both PWM on the brake low signal and writing new stop brake values directly into the controller while braking. I doubt pwm on the low signal will work but writing new values in just might.

Doug

 

[stan.distortion]

@amberwolf, Thanks, it looked simple from looking at controller schematics but I hadn't come across discussions on it before. Still surprised variable regen isn't a more common feature on these newer controller brands (wasn't aware the bigger APTs had 2 levels), maybe not too far off though, they're improving at an impressive pace. I guess it's not too unnatural, infernal combustion only really offers one level at off-throttle but it never felt right to me.

Hopefully the smart display can force better behaviour but it's a shame it's necessary, maybe one day there will be open source firmware or can bus control options for direct control rather than having to add an additional layer to try an overcome issues with the manufacturers configurations.

 

[amberwolf]

If you're willing to use one of the various open-source controller systems, there is already OSFW (dont' know how well-documented), some of which already offer variable regen of various types.

VESC has a number of high-power variants, especially here on ES, though you would likely have to build the controller for the ES versions since they are not (AFAIK) commercially available. (most use VESC in the title, but there are also the MESC and HESC, etc).

If you can build a suitable powerstage, or hijack it like my "zombie honda ima" thread, you can use the Lebowski brain to run it.

 

[dougf]

Machined up an aluminium case for my display. Couple of mistakes and had issues with my mill but it will me ok. Im going riding tomorrow so ive reassembled. I actually remembered to take pictures 8) .

Doug