Throttle response on my first e-bike

[LandKurt]

I just put together my first e-bike from kit and rode it twice this weekend.

I used a Grin Tech eZee rear conversion kit on a Marin Muirwoods 29er bike which is a cromoly frame no suspension 24 speed bike with disc brakes. I’m using an 8 speed 11-tooth freewheel, half twist throttle, and a Cycle Analyst. The battery is a 48V 13Ah AllCell lithium NMC from Chicago Electric Bicycles, only 8.5 lbs. It all went together fairly well, though I still need to finalize the placement of controller and battery, currently they are in a trunk bag on the rear rack.

I’m pretty pleased with how it works, but I was somewhat surprised by how the throttle operated. Though I got some hints from playing with the bikes.ca simulator.

In practice there seems to be a rather narrow range of useful throttle input for a given speed. Playing with the simulator for my setup I see that, for example, below 10 MPH 66% throttle and above is the same as WOT. On the other hand, when you're cruising at 20 MPH anything below 66% is the same as no throttle. So if you've just been coasting down a hill with no throttle and want to add just a bit of power as things level out you've got to apply two thirds to three quarters throttle before you get any response at all.

I'm sure this comes as no surprise to any of you. Do all throttle systems and controllers on electric bicycles act this way, or am I misinterpreting my limited experience and the simulator? I noticed discussion of throttle mapping in the CA V3 thread; does that pertain or is it another issue?

I felt it a bit difficult to adjust the throttle to get just what I wanted, but maybe that will come with practice. Does cruise control help with this, or does that simply lock the throttle at a given input setting? I believe I’d have to upgrade controllers to get cruise control. Since I have the CA I may just rely on the maximum speed setting by setting it at a speed I'm comfortable and then riding full throttle most of the time. I tested this on my second day and found the speed response a little slow; set to a limit of 20 MPH it would get up to 22 MPH before cutting the power and when hitting a hill it would drop to 18 MPH before reapplying the power. That 10% plus or minus seemed a little sloppy, but could be livable.

Kurt

 

[jasonf150]

I noticed that strange throttle thing too. I found that limiting the amps intead of the top speed helped. When I was using a battery that was pretty underpowered for my setup, I limited the speed to 17 and had the exact same problem. So I undid that and limited the amps instead, it worked a lot better. Now you don't really want to limit your amps, so this solution might not work, but you could try setting the amp limit to one higher than your controller and see if it helps spread the range.

 

[docnjoj]

Somewhere in all these posts is a way of inserting a 1K ohm pot or resistor to gain throttle range, but I can't remember which throttle wire to put it in? Can anyone help?
otherDoc

 

[neptronix]

You guys both have low amp controllers and are not giving your motors enough juice.

It is like dropping the hammer on a 1 liter 3 cylinder car up a hill. Of course you are not going to get much response.

Dropping major speed on a hill is partially a result of not putting enough amps into a motor. You'll see it when playing with the ebikes.ca simulator pretty easily.

I run 40A on a geared motor that is basically a larger, more powerful, and cheaper version of what you have. It has no problems keeping speed. Now, if i gave it only 20-25 amps? it would accelerate like a slug and i'd drop 10-15mph on a hill.

P.s. you bought a battery that can run at 2C maximum, so your max amps are about 26A on a 13AH battery. I think your stock controller puts out 20A. So you're close to the limit of your battery, but reprogramming the controller for a few more amps would help.

 

[jasonf150]

I am planning on putting a 40A controller on my bike due in part to your advice on my other thread, but I'll still limit it to 25A (unless I want to have some more fun on occasion) due to the range. I need to be able to get 12 miles out of one charge, and I use about 7.5-8aH at 25A. The battery is only 9.2aH. If I ran that at 40A, it would lower my range. Fine for playing around, but for my daily commute I can't do that.

 

[neptronix]

You should really be running 40A full time. Starving your motor for amps at such a high voltage loses you efficiency.
A better setup would be lower voltage, high amps - or a more appropriate motor like a HT Crystalyte.

Play with the ebikes.ca simulator, you'll understand things a bit better... you have a motor that wants high amps, not high volts, for best torque and efficiency.

I always set my controller to have quite a bit more amps than needed to maintain a cruise. This ends up being a power reserve for when i hit a hill, face a headwind, or cross a high traffic road quickly from a stop, statistically increasing my lifespan..

I am planning on putting a 40A controller on my bike due in part to your advice on my other thread, but I'll still limit it to 25A (unless I want to have some more fun on occasion) due to the range. I need to be able to get 12 miles out of one charge, and I use about 7.5-8aH at 25A. The battery is only 9.2aH. If I ran that at 40A, it would lower my range. Fine for playing around, but for my daily commute I can't do that.

 

[dogman dan]

Ebike throttles are pretty lame. Get into the 40 amp ballpark with 72v, and it's like off, off, off, then it spins the tire. Takes a really fine touch to get out of the gravel driveway without an accident, to where you can straighten out on the road and open it up.

It took me a long time, years in fact, to really learn to ride anything but WOT. It's just really hard to adjust for anything less than a 25% change. A visual indicator of your watts helps some, such as a Cycleanalyst. You can see a change on the CA that is too small to feel through the pedals.

I'd love to see a new throttle come out with a much finer controll, but still in the 20 buck price range of course. Something with a few more hall sensors inside it?

 

[Tench]

There is a previous topic about twitchy throttles with lag in them here, Fetcher explained to me how to improve mine, it worked a treat and was easy to do, the explanation is towards the end of page 2 and on page 3 of this thread linked below;

http://endless-sphere.com/forums/viewtopic.php?f=2&t=37190&start=30

Simon.

 

[LandKurt]

It sounds like the finickiness of the throttle is a recognized issue with e-bikes. I think I'm on the right track using the speed limit function of the CA. Next I'll try the advanced setup menu and see if adjusting some parameters improves the response time. Time to play with PSGain, IntSGain, and ITermMin. Since 50% throttle is as good as nothing around 20 MPH I can probably raise ITermMin to about half of ITermMax.

So far I'm not disappointed by the 20A controller. It's a good match for the 2C 13AH NMC battery. I'm use to straining up 8% grade hills at 5 MPH on my own power alone. The fact that my e-bike needs a little help from me to haul my 275 lbs up that same hill at above 15 MPH is not a problem. Anything below 5% doesn't seem to slow it down much from the 20 MPH speed limit. Luckily we don't have any extreme 20 or 30% grades around here.

Because of my size I was more interested in torque than speed, so I went with the geared eZee. Now a 36V 250 watt store bought e-bike would have been a joke for me.

 

[dogman dan]

They seem to be a joke for nearly everbody on a steep hill. Do try to keep your speed up to at least 15 mph up the steeper hills, so the motor will make less heat. Too slow and much of your power goes into heat, and too much of that melts motors. For most normal speed windings, 15 mph is enough to keep the heat waste reasonable.

Some of what you describe, if I understand you right, is simply because you won't feel anything if you feed less watts into the bike than it needs.

Trying to describe this right, say you are going 20 mph already. Coasting, or pedaling along, whatever. It takes about 400w to go 20mph. With your gearmotor, if you are going 20 mph, but apply 50% throttle, you likely are dishing out less than 400w. Since you are going 400w speed already, the motor does spin, but you'll need to apply 401 watts before the motor spins faster than it's already freewheeling. So you don't feel anything till you apply around 450 watts, or even more.

Since the motor freewheels, it does feel quite a bit different from the direct drive motors. It feels like the throttle doesn't work at all except at WOT, if you are already rolling.

 

[LandKurt]

You're right, it takes about 400W of output to go 20mph on the flat, but I don't think the motor would be using anything near that at 50% throttle. At least not according to the ebikes.ca simulator.

The simulator parameters: eZee 26"/700c, 50V battery, 20A controller, 700c Wheel, Mountain bike posture, 150 kg gross wt

With the throttle at 60% the simulator shows the thrust, power and efficiency curves zeroing at 18.8mph. At anything above 18.8mph it shows motor power and torque at 0; battery power is at 34W. I interpret that as meaning the controller is using 34W to spin the motor with no load. There is no load because it isn't keeping up with the freewheel as you coast down a hill at 20mph.

Admittedly that 34W is completely wasted just because you didn't bother to let off the throttle (or set the CA ITermMin fairly high). If I play with the ITermMin I'll have to see if the wattage behaves that way when I'm above the set speed limit of the CA. It might be worth wasting a bit of power if raising ITermMin means a faster recovery of the throttle when I hit a hill. Supposedly a 30% throttle would only waste 8 watts freewheeling above 9 mph.

I will be careful not to let the motor bog down too badly on steep hills. My own 200 watt contribution can make a big difference according to the simulator.

 

[dogman dan]

Yes, that's exactly right, untill your motor spins faster than you are coasting, a half throttle setting that might normaly pull 200w would merely spin the motor no load and use about 30w.

The point I was trying to explain is that untill the motor feels a load, it would seem like the lower half of the throttle is not working as you cruise, basicly coasting at that point. Transition to uphill, and then it will feel like the lower part of the throttle works again. There is a bit more feel to it with a dd motor, compared to a gearmotor that has no feel to it while it freewheels.