250w and low power experiments...

[NRG]

With all the talk of illegal / legal bikes recently I thought I'd try and setup my Alien GSII so peak power fell in line with the EN15194 regs.

Really this was just an exercise to see what a lower powered bike would feel like.

The E-Crazyman controller I have retro fitted to the bike is programmable, I can set various functions including battery and phase current, however, there are certain limitations in the s/w which meant the phase current would have to remain at the same level I normally use IE: 35 amp

But I can program the battery current down to 7amps. Current control in these controllers is somewhat crude and can be as much as 5amps out, my controller I found is +3amps out, so 7amp programmed means 10amp in reality.

Still it was enough to get a feel for a low power setting. Working backwards the EN regs allow +10% on motor power so 275w.

Motor speed where peak current is drawn corresponds to about half the no-load speed and looking at the ebikes.ca simulator for geared motors I took a guess and estimated efficiency of the motor to be 65~70%.

Losses due to wiring and controller resistance are hard to quantify but at these low currents it would not be too much so I guessed at 20w. So a couple of gross assumptions made.

Taking 65% efficiency, 275w would require 423w in plus 20w for losses gives an input of 443w....call it 450w.

At 36v this meant I needed to set the controller to 12.5amp or in my case due to the +3 overshoot 9.5amp.

On the road I noticed the motor was quieter and speed the same. I picked out a few hills steepest about 10% and the bike seemed to climb just fine at about 8mph.

Checking the Watts Up meter it recorded a peak of 12.3amps and 478w so motor power was probably around ~300w or maybe higher....it didn't seem too bad and could be a good compromise setting between distance and assistance on hills.


I then decided to reduce power further and set 7amps on the controller (10amp real) and tried again. Peak amps recorded where 9.7 and peak watts 390 so at a guess the motor was running at 240~250w.

This was noticeable but the bike was still ridable, pulling away I initially thought something was wrong as the motor was very quiet but it was still working! Acceleration was noticeably slower but the bike still climbed the hill albeit at 6mph this time not 8mph.

No idea what the range would be with these settings but I've set the controller back to 12.5amps and intend to try it one weekend.

 

[Scottyf]

what size battery have you got? The 13amp max controller on my 250watt system was a similar spec. It meant you really had to pedal quite a bit. Which is good because you didn't slow down too much but equally meant you wasn't to slow on the start up but got a decent amount of exercise.

Also meant battery and range was huge. I used to chew up 2.7amp hours at 18miles. So in theory i could do 54miles at that rate it only a 9amp battery.

Like you said didn't really offer much speed but still helps considerably no matter what hill you get too.

But it definetly meant it was an assist.

 

[NRG]

It's rated 10Ah but probably less than that now it's 18 months old.

 

[jbond]

Interesting experiments. I've thought a lot about this area and reached the conclusion that what I want is:-

- Pedelec economy mode. Low current limit, low power but without a speed limit. This acts like a cruise control. As long as I pedal I get some assist to push up average speeds and take the sting out of hills. No control thought is required. It should act like an on off switch. Pedal and get assist, stop pedalling and assist stops. If I can pedal at 20 mph (due to wind, slope, effort) the assist should push me up to 22mph.

- Throttle over-ride. No matter what the pedals are doing, I should be able to wind in up to full power on the throttle. So if the hill is just too much, I just apply some throttle. Or if I want to accelerate away from the lights, weave between the cars or take the roundabout at car speed. Once the short term power need is over, I just release the throttle and go back to economy, cruise pedalling.

I think it might be possible to achieve this with a CA, a small amount of electronics and a PAS that outputs an on-off throttle signal. It might be necessary to integrate the PAS pulses and turn the into a 0-4.4v switch before applying them to the CA power control. It's probably possible to create just the current limit control of the CA in discrete components or even re-program a wattsmeter to do it.

 

[flecc]

Interesting NRG, fits in with my experince that the original true 200 watt bikes of three decades ago were just too weak, even with the 12 mph assist limit at that time.

The nearest parallel on the market now are the Powacycle Salisbury and Windsor models at around 300 watts, and both those climb moderately well though a bit slowly on up to 10% slopes. Their motor setup take them just about to 15 mph, a touch weak beyond 13 mph.

 

[lemmy]

It has often occurred to me how nice it would be to have a configurable controller so that I could control the maximum and minimum assist speeds and power, power in and power out points.

It'd need a usb socket and maybe configurable using javascript or vbs, something that would give maximum flexibility rather than just a series of choices.

One thing I'd like, for in town, is to have the power cut out at 5mph, in other words, power assist only for pulling away from lights.

 

[flecc]

One thing I'd like, for in town, is to have the power cut out at 5mph, in other words, power assist only for pulling away from lights.

Sounds like an application for kinetic energy Lemmy, KERS style.

Use a spun up flywheel or spring wind to perform braking, then release that energy to kick off the start.

 

[Deleted member 4366]

Like I was saying before. If you have a powerful bike, you can always turn the power down to how you want it. For my next project, I'm going to have a go at building a programmable motor control based on a Genie chip, which has loads of analogue and digital inputs and outputs. My idea is to read the pedal sensor, speed sensor from the speedo and throttle, process them and send the output to throttle input on the controller. My present setup can draw 28 amps at full throttle, which is too much for most circumstances. I want to have several modes, butone must be more like my other bike, where the power gradually reduces towards 15mph, but with quite a lot at 10-12 mph.
Hopefully, I can power it off the throttle feed.
Rapid Electronics - GENIE 20-pin microcontroller project kit
I got the idea from the Haibike, which must be doing something similar. Also, after looking at the motors on the Ezee bikes and others, I'm sure they are much more powerful than what they're programmed to give.

 

[banbury frank]

Hi

Can I suggest that you start with the infineon controller as it already has all the connections for the Cycle annalist which controls the amps drawn from the controller

http://www.ebikes.ca/store/photos/C3640-NC.jpg

Frank

 

[lemmy]

Sounds like an application for kinetic energy Lemmy, KERS style.
Use a spun up flywheel or spring wind to perform braking, then release that energy to kick off the start.

Would you mind knocking me up such a device at the weekend, Flecc? I'll buy you a drink for your trouble. Also, please don't patent it, there is no need, I won't steal it, honest

Seriously, I wonder whether the gyroscopic effect would be a problem?

 

[flecc]

It may be, though weight would be a problem with a flywheel solution, even with ultra high speed centre mass flywheels.

So clock spring wind-up would be the preferred option to store that short term energy. I envisage it speeding up a commute in this way. First, by maintaining one's speed to as late as possible on red lights approach with hard storage-braking to store as much energy as possible. Second, the spring assisted takeoff accelerating one back up to speed more quickly.

This would be a perfect adjunct to internally geared hub motors, which are not suited to standing start takeoffs for nylon gear stress and current consumption reasons. As the end of spring energy unwind is reached, it could operate a motor-on switch.

 

[Deleted member 4366]

A lot easier to do electrically with either some of those super-capacitors or an additional lipo battery. I think someone else had that idea to have an extra battery, that at the push of a switch would come in series with the existing battery to give an instant boost. You just need a powerful motor and controller that can han handle it.

 

[flecc]

Yes, very easy to implement as you say d8veh, maybe with a braking-dedicated regenerating front hub motor like the latest Panasonic one on trial in Japan.

 

[NRG]

All motors have the potential to provide more power. eZee motors are no more powerful than the equivalent Bafang. Motors are all power hungry little beasts and will eat whatever can be fed to them and then demand more, the difference is the point at which something bursts into flames or fails mechanically...this differing between each motor.

What stops the motor from self imploding is the controller, it also protects the power source from being devoured by the motor.

The current limit setting in the CA and also the controller is there to protect the battery. It effectively strangles the battery and the side result is to limit motor power.

However, within the limits imposed by the CA or controller the motor can still deliver rated power and more because the controller is a very efficient voltage to current amplifier/converter. It will trade volts for amps to ensure the motor can still provide power up to a certain point.

Limiting battery current in this way is a very crude way of controlling motor power. The best way is to control power is via phase current and the Kelly controller is the only one that I know of that truly controls motor power.



---------------------------------
Posted using Tapatalk

 

[Orraman]

Greetings d8veh,

Your Post # 8

I had not heard of the Genie microcontroller and followed your link to Rapid Electronics and then on to the Genie site.

In recent years I have 'moved' from conventional microcontrollers to the Picaxe range which although downloading programs in the same easy style as the Genie are, to my understanding, much more competent.

If you are currently using the Genie system and consider it superior I would appreciate any information you would care to share.

I use a Panasonic powered bike now for it's torque sensor, I need a measured amount of exercise for heart problems but am unable to walk unaided for more than a few feet.
Previously I had lashed a Cyclone motor to an old recumbent that I once enjoyed but using the less than perfect throttle I found it impossible to strike a balance between my effort and the motor before the gears needed changing. Although it was exhilarating when up in top gear.

Torque sensing chainsets are expensive and I had thought of a Panasonic style of push button selection used for assistance level but in this case thought it might be possible to select 'power' setting that would manipulate the throttle signal before sending it to the controller.

Comments would be appreciated.

Dave

http://www.rev-ed.co.uk/picaxe/

 

[Deleted member 4366]

Greetings d8veh,

Your Post # 8

I had not heard of the Genie microcontroller and followed your link to Rapid Electronics and then on to the Genie site.

In recent years I have 'moved' from conventional microcontrollers to the Picaxe range which although downloading programs in the same easy style as the Genie are, to my understanding, much more competent.

If you are currently using the Genie system and consider it superior I would appreciate any information you would care to share.

I use a Panasonic powered bike now for it's torque sensor, I need a measured amount of exercise for heart problems but am unable to walk unaided for more than a few feet.
Previously I had lashed a Cyclone motor to an old recumbent that I once enjoyed but using the less than perfect throttle I found it impossible to strike a balance between my effort and the motor before the gears needed changing. Although it was exhilarating when up in top gear.

Torque sensing chainsets are expensive and I had thought of a Panasonic style of push button selection used for assistance level but in this case thought it might be possible to select 'power' setting that would manipulate the throttle signal before sending it to the controller.

Comments would be appreciated.

Dave

PICAXE

The Genie chips are basically Pic chips with their own software in that makes them very easy to program with a visual flow chart interface. We use them at school and the kids don't have too much trouble understanding it. You can see the values of the sensors and all the other inputs/outputs in real time, which makes handling sensors very simple. The only problem with the Genie form of the chip is that there's no analogue outputs (I just realised after ordering them). I might be able to overcome that by using a pulsed output, otherwise it means some extra bits and a bit of head scratching.
My bike is very powerful, so what I want to do is intercept the throttle signal and use the Genie to feed a power profile to its return, so that I've got a tour mode that will ramp down the power from about 400w at 0 to 5mph to zero watts at 15mph, then a climbing mode that will boost the power from 8 to 10 mph to about 650w, and finally a 15mph mode that'll give about 400w up to 13 mph and then ramp down rapidly to zero at about 16mph. I should be able to program any profile I want if these turn out to be no good. You could also connect it to heart monitors, temperature monitors or anything else as long as you can tap in to their outputs. I don't think that this would work with the smaller 250w motor/controller because there's not enough power in the first place,

 

[Orraman]

d8veh,

Thank you for the information about Genie, I understand that some Picaxe chips now have a 32 level digital to analogue converter but have not used it.

In the past there had been efforts to use a spring loaded arm bearing down on the chain behind the chainwheel to measure the deflection due to chain tension as an indication of torque but this proved unreliable.

I had considered using a switched arm in that position to engage the controller at the first pressure on the pedal for immediate response as with the Panasonic motor. A current sensor in the battery lead could allow the micro to regulate the throttle signal to maintain the selected ' power' setting with an automated removal of 'power' when pedal pressure is removed.

This is where I would see the desirability of the modes that you propose, my hill climbing mode would require that power would be maintained between pedal strokes to prevent surging and loss of assistance at dead centres.

I see this as being an adaption of the Cyclone system that would make no permanent changes and be reasonably easy to implement.

I wish you success with your endeavours.

Dave