motor braking

[E-HP]

Hello all,

I took my e-bike out for a ride yesterday and ended up putting on 48 miles. Perfect day, after not being able to ride for a couple of weeks due to the smoke in the air. 90% of my riding was using pedal assist, and mostly on level 1, since most of the riding was pretty flat, with a few moderate inclines. On the way back, after about 40 miles, I noticed some strange behavior on 3 or 4 occasions. From a standing start or from very slow speeds, pedaling the bike before PAS kicked in feels like pushing against the motor, as though it were braking. At one point I stopped and hand turned the wheel and could feel the resistance. After PAS kicked in, or if I used the throttle, everything went back to normal, and stopping and starting required the same pedal effort as before.

I'm wondering if this is a common problem with an easy solution. The motor, controller and battery remained cool during the whole ride, but I did do 3 275 ft climbs of maybe 10% which shouldn't be an issue based on the bike's observed performance. Battery settled at 51 volts after the ride, but I did see it sag to 48 volts on the last climb up to my house.

Thanks ahead of time for any insights.

Here's my setup:

Motor:
26” Rear Hub - MXUS model: XF 40 1000W 1610 0020
48V, 1000W

Controller:
Sine Wave : KT36/48SVPRK-SLSO2G
Maximum current: 45±1A
Rated voltage: DC36V/48V
Rated current: 22A
18 MOSFET

Display:
KT-LCD3

Battery:
Li-Ion – Unit Power Pack - 52V, 20Ah, 30A BMS

Halfway point in my ride, Yerba Buena Island in the middle of the SF bay. The Bay Bridge has a walking/biking path from my side of the bay to the island.

 

[neptronix]

DD motors all have some resistance, but it sounds like a lot..
Is your freewheel rubbing against the frame or something?

How bad is the drag? like a slight brake rub, death gripping the brakes, or somewhere in between?

 

[E-HP]

DD motors all have some resistance, but it sounds like a lot..
Is your freewheel rubbing against the frame or something?

How bad is the drag? like a slight brake rub, death gripping the brakes, or somewhere in between?

It feels noticeably more that brake rub, but not a full lock up. I'm hoping there's not something damaged. The day before, I did torture the bike a little when I took it off road. The trail turned steep and rutty a while into it, and then got to points where I need to walk the bike (probably would have even with my mountain bike), but the motor, battery and controller never got hot, and didn't have these symptoms that day. I actually think the bike would have climbed those sections if I had off road tires instead of slicks packed with mud (I think saw a post about phase wires shorting and causing the hub to lock up, so I'm hoping there's not an intermittent short of some kind).

Mud makes good cammo...

 

[E-HP]

I finally figured out what was going on with this motor braking stuff. The issue occurs when rolling off of throttle, and when below the 1v lower threshold of the hall throttle. Between 0v and 1v, the motor brakes like with regen, but without the brake symbol or regen on the display. It's subtle right below 1v, but is fairly strong just above 0v; enough to slow the bike a few mph on a 6 percent grade. Hard to keep it in that range, but as soon as it drops to 0v, the braking releases and the wheel coasts freely. If I start pedaling when the bike is in this braking mode, I'll get strong resistance until the PAS kicks in (the original symptom).

I tried upping the regen level in the settings to see if there is an effect on the non-regen motor braking, but didn't notice a difference, but that may be due to the slower speeds I was riding at. I'm going to try testing on a longer hill at higher speeds next. Interesting behavior, but now that I'm getting to know how it behaves, I can work around it or maybe even take advantage of it in some situations.

 

[E-HP]

Still playing around with this, but another observation is that with regular regen braking, the braking takes the bike down to 8 mph then the bike will coast. With this throttle braking, the braking takes the bike down to 4 mph before coasting.

I'm thinking that I can take advantage of this behavior by upping the lowest throttle voltage to something like 0.1V, using a pot (and fine tune for maximum motor braking). I was already going to add the pot to get rid of the dead zone of the throttle, but instead, I'll keep the dead zone, but just change the lower threshold to something just above zero volts. That way, when navigating between foot traffic on a bike trail, I can use throttle only to speed up and slow down at those lower speeds, which may result in more predictable behavior that PAS (that can surge forward) combined with normal braking.

 

[AHicks]

This is the second time I've seen this low throttle setting regen activation mentioned recently. I have a very similar setup, using a thumb throttle, that doesn't have this issue. Like you, wondering about the potential for a little extra control, so watching your progress with interest.

This "surge" you mention as PAS engages is another difference in how our control systems are responding differently. I get none of that, no noticeable surge, as my PAS engages. I'm wondering if this might have something to do with the PAS level we normally ride at? I'm usually in no hurry to get anywhere, so ride mostly in PAS 1 or 2. Thought being the surge may be present, but at a power level where I'm not noticing it.

 

[E-HP]

This is the second time I've seen this low throttle setting regen activation mentioned recently. \

Might have been me. I mentioned it on another thread when I first started noticing it, but then connected the dots and realized it seemed to be tied to the behavior I noticed last year, on this thread. I'm going to confirm that this morning by getting into this braking mode and trying to pedal to see if I get the same resistance that I'd observed last year.

I'm convinced that this behavior is tied to the ability for KT controller to provide variable regen when using the open source firmware. The controller is definitely seeing the signal below 1V.

EDIT: Now that I think about it, this happened with all three half throttles I've tried, I have an old thumb throttle, so I may test that as well.

EDIT2: OK, riding down the hill on my normal route (6% grade), riding up to 15 mph, then applying the motor braking will bring the bike down to about 7 mph in about a half block while still descending. That’s about the same as regen braking set to level 1 or 2. I’ll do some controlled testing, but I can see another use for this since currently I set regen braking between 3 and 4, but 4 kicks in abruptly, which can cause some squirreliness if you need to shed some speed around a wet corner.

On flat ground, when braking with the throttle, I have a lot of resistance if I try to pedal, until PAS kicks in. Currently with my dual hall sensor, it kicks in at about 1/3 of a revolution. When I first had the issue, I was using a 5 magnet sensor, so it was closer to a full turn in that case.

 

[AHicks]

Expand on this please?

"On flat ground, when braking with the throttle, I have a lot of resistance if I try to pedal, until PAS kicks in."

How are you able to peddle at all, without PAS kicking in/engaging? Is PAS = 0? My pedals barely move prior to PAS engagement. 12 magnets I think

 

[E-HP]

Expand on this please?

"On flat ground, when braking with the throttle, I have a lot of resistance if I try to pedal, until PAS kicks in."

How are you able to peddle at all, without PAS kicking in/engaging? Is PAS = 0? My pedals barely move prior to PAS engagement. 12 magnets I think

It's before PAS kicks in, for about 1/3 of a revolution, I can stand on the pedals to fight the motor. With my old 5 magnet PAS, it took almost a full revolution before PAS kicked in. My dual hall is 12 magnets, but still require a short turn of the crank to turn on. I think because it takes several degrees of throttle rotation before power is applied, it's easy to have the throttle slightly rotated into the dead zone without knowing it.

EDIT: I used the motor braking on the way home several times. I think I'm getting used to it, and the variable braking. I navigated past several walkers, with a few walking on the wrong side of the path, and people and bikes travelling in both directions, using only the throttle to speed up and slow down. I also used the braking to slow to a few mph at the intersections before crossing. Basically I was able to get home without using brakes except when riding down my driveway, since I needed to stop to open the garage door.

PLAN: Here's what I'm thinking. I'll need a small two way switch and a couple of resistors, and an adjustable pot used for testing. The plan is to wire a trim pot to my throttle ground, per Tommycat's hall sensor throttle guide, in order to eliminate the throttle dead zone, then take a measurement for resistor #1. Then adjust the trim pot for maximum motor braking; I'll need to do some trial and error to get as close to the cutoff as possible. That will be the value of resistor #2.

After the testing, I'll remove the trim pot, and run the ground wires to the switch and resistors. Position #1 will be normal throttle, with no dead zone. Position #2 will be for motor braking. I probably also need to temporarily hook up my watt meter, so I can confirm whether or not current is flowing to the battery or not, and monitor my controller for heat, until I until I'm convinced that doing this isn't causing any damage.

08/08/19 More testing: I can do a couple tests on my descent each morning, getting up to speed, then testing the motor braking. Tested twice this morning, getting the bike up over 20 mph, then motor braking. Pretty consistently brings the bike from over 20 mph to about 7 or 8 mph in about 2/3 of a block. I might get that to about 1/2 block once I add the right resistors. I also started reading through the Sabvoton controller thread, and this phenomenon sounds very similar to "slide" braking, so I'm going to do more reading.

08/09/19 I'm now skeptical that the trim pot / resistor mod will work, even for eliminating the throttle dead zone. Actually the pot would work for the dead zone, but it will have a negative effect on PAS. This is because of the other odd behavior where any amount of throttle while PAS is in effect, will act like full throttle. So, even with my normal PAS level 1 riding, it seems that any throttle voltage over 0V will just override. I thought about it last night, and tested this morning . I'm going to still try it, but right now I'm not optimistic.

 

[E-HP]

I just did about an hour of testing on this motor braking thing. Had a few observations that will alter my plan.

First, the maximum braking isn't occurring in the throttle dead zone (below 1V), but just above it. I used my pocket knife to score a couple of lines on the throttle to see where things were occurring. That means my resistor circuit won't work the way I wanted, due to how PAS functions on my controller when any amount of throttle is applied (with ANY voltage above 1 V, PAS will ramp up to maximum, regardless of the PAS level (except 0). With no throttle, PAS 1 will ramp to 250W-300W normally, instead of ~1300W or so when slight throttle is applied. So using this "feature" for navigating around/through foot traffic wouldn't work, unless I set the PAS level to 0.

Another observation is that the motor braking is pretty effective. I did 6 runs on flat ground, riding up to 20 mph, the applying the throttle motor braking. The control run was riding up the 20 mph, the releasing the throttle. The bike coasts approximately 120 yards before speed drops down to 4 mph. Using the motor braking, the bike slows to 4 mph in 25 to 30 yards; or about the distance I would normally apply brakes before the intersection. No signs of regen are happening though, since under normal regen braking, the battery level meter scrolls and the battery voltage rises, neither of which occur with this motor braking.

After the testing, and practicing on the path and in a parking lot, I navigated through several congested parts of the path (lots of walkers, joggers, and bikes out on the weekend) easily using throttle and throttle braking only. That's going to be the main application for me, since there isn't a corresponding benefit of regen in other situations. I'm going to still dial out the dead zone though.

08-18-19 After a lot more testing on longer hills, I can now see that this motor braking via the throttle IS also providing regen. Although the battery level icon doesn't indicate regen, the real time voltage is definitely going up. I also did a comparison between regen at the different levels in the setup, against the throttle braking and found that the throttle braking near 0V is as strong as the highest regen setting, and much greater than the setting that works best for me. I tested by riding down the same segment several times, and seeing the mph speed when passing the same point, while applying either the regular braking regen (trying not to apply actual braking) and the throttle method. I think I need a much better throttle to take full advantage of this though.

 

[E-HP]

I've figured out an application for the behavior of my KT controller and hall throttle. I'll be using the dead zone elimination circuit from Tommycat's guide, but setting the trim pot so that it is just above where it starts providing power, not below. The trim pot will be bypassed by a normally closed switch, so normally not used. The normally closed momentary switch will be mounted near my left thumb.



Based on the behavior I'm observing, the push button switch will function in the following manner:

• While pedaling with pedal assist engaged, pressing the button will immediately apply full power/throttle

• When using throttle only, or when pedal assist disengages, pressing the button will allow easily using the throttle down to full regen braking when no throttle is applied (just above 1V)

Even though I was thinking I'd use the button most while navigating in between foot traffic, using throttle only, I can see using it for normal braking/slowing, because it's variable, so less abrupt, but stronger than the regen braking setting I'm using.

The switch and trim pot should come tomorrow. I had to get 6 switches, but I'm using a couple to convert my cheap (but 10-times-better-than-the-crappy-kit-levers) Avid brake levers, to use on my e-bike.