Thoughts on throttle taming?..

W

webstaff

100 µW
Joined
Jan 22, 2016
Messages
8
Location
North West
I just wanted to ask for peoples thoughts on throttle taming on higher power ebikes..
And where to spend my money next..

This is the bike I'm bothered about :)
https://endless-sphere.com/forums/viewtopic.php?t=81898

It's in the process of being chopped up to form a battery box for an 18650 20AH pack as it was only ever thrown together but is ran great..
I also built up a spare as I have a few of the GNG kits and also wanted to have a play with lasercut housings as I have a laser cutter to hand at the moment that I just rebuilt..

WhatsApp Image 2016-11-21 at 13.06.28.jpeg

But the throttle on the standard controller is to aggressive.. (speed vs power) these are speed based from GNG I think and it's just a right handful. :D
And I think it's causing a multitude of issues not having that fine control to be able to set off at lower speeds.. More so on that cheaper spare bike but both bikes have similar issues with high wear rates and aggressive throttles..
So new controller with power instead of speed based throttle.. or CA so I can use their software / hardware interface to achieve the same throttle effect?
Or both :)
I did look into POTs on the throttle control lines but thats just a PITA as you loose out somewhere or you start getting into complicated circuits/setups or arduino's but again just seems pointless and overly complicated when existing options would have other benefits..
So what's your thoughts
Maybe you know of a cheap way to achieve what i'm after?

All the best Webstaff
 
CAV3 should provide the capability to tweak whatever throttle ramp up you want/need? You could also create 3 different profiles for soft, medium and hard power/throttle responses?

Edward Lyen built me a very simple circuit which installed in-between throttle signal and slows the speed (ramp) which the throttle signal is applied. Trouble is, that circuit also slows down the “off” throttle signal so that simple circuit may just shift your problem to different issue of throttle response?
 
Ykick thanks for the quick reply,
Yeah I did wonder about that with regards to delaying the signal, I take it then with your mod/circuit after you throttle off you have to wait for the throttle to actually drop then.. That might be a slight issue if you forgot and we're close to say a car :D hehe

As I say with regards to a simple circuit I was thinking more of going for an exponential curve on the throttle or at least a wide band on the slow range then narrow banding for say jump to 50% and jump to 100% close to the end of the throttle pull but as I say it gets complicated quickly at which point its easier to do in software and at that point you might as well buy off the shelf stuff as I have better things to do with my time than reinvent the wheel ;)

Guess it's a CAV3 then with speed sensor and shunt on order next :)
Wheres best place to order for for UK shipping, any thoughts?

would there be anything to gain with changing the controller? (like integrated shunt?)
 
webstaff said:
Ykick thanks for the quick reply,
Yeah I did wonder about that with regards to delaying the signal, I take it then with your mod/circuit after you throttle off you have to wait for the throttle to actually drop then.. That might be a slight issue if you forgot and we're close to say a car :D hehe

As I say with regards to a simple circuit I was thinking more of going for an exponential curve on the throttle or at least a wide band on the slow range then narrow banding for say jump to 50% and jump to 100% close to the end of the throttle pull but as I say it gets complicated quickly at which point its easier to do in software and at that point you might as well buy off the shelf stuff as I have better things to do with my time than reinvent the wheel ;)

Guess it's a CAV3 then with speed sensor and shunt on order next :)
Wheres best place to order for for UK shipping, any thoughts?

would there be anything to gain with changing the controller? (like integrated shunt?)
Click to expand...

You're welcome and yes, that’s how it “behaves” - throttle release (ramp down) at the same slower rate that it climbs (ramp up). That in mind, brake cutouts will immediately cut throttle.

Decent solder jockey should be able to insert DP CAV3 using most any controller battery-controller (-) neg internal shunt.

Sorry, I dunno a good source for UK shipping but here’s a list of dealers - http://www.ebikes.ca/dealers.html

This guy in Germany posts frequently and I’ve noticed some cool CA supporting products - http://www.ebike-solutions.de/

All the best...
 
I didn't think many people use the throttle tamer anymore?
OP do you have a partial twist throttle and a 3spd switch?
Imo for high power you might not be happy with just adding on the CA to simulate torque throttle. At least for single reduction maybe. And I have heard torque controllers like kelly are more 'like golf cart' control? idk

3.5kw has the potential to be pretty 'rammy' on a bike's drivetrain.
Also, try turning down phase amps if you can, or try dropping some voltage?
 
Webstaff,

If you are looking for a CA V3, try these guys. They are UK based.

http://www.jozztek.com/shop/meter-gauges/196-v3-dps-cycle-analyst.html

I ordered one a week or so back. It was the best price on this side of the pond and arrived within 24 hours.
 
The electric motor has inherently greater torque at low RPMs.
This causes the jerkiness you refer to... since only low torque is required from starting off.
As speed builds you need more torque to overcome rolling resistance and wind.
So the controller needs to factor in the speed or at least incorporate a soft start in the throttle algorithm.
 
https://endless-sphere.com/forums/viewtopic.php?f=28&t=42785&start=3500#p1172502

ebMCJxb.jpg


small $35 controller. always adds a smooth ramp-up to your throttle input. the side effect of which is that you don't get instant power, it always goes through the ramp-up which takes about 1.5 seconds
 
Anything that uses a time delay is 'sleazy'..
A ramp is good but that's part of what a throttle is anyway.
The best way to 'soften' the throttle is to incorporate road speed into the algorithm.
Better still some feed-back from an accelerometer... since acceleration is what is ultimately required.
Something like artificial inertia such as used in model railway locos would be good also.
 
Alan B said:
The best way to tame the throttle is to use a torque proportional throttle. Give a little throttle, get a little torque.
The simple PWM throttles cause tremendous torque spikes on powerful setups. It is cheap and simple, not good.
Click to expand...

I agree. With a torque command throttle, you don't need any taming. You have to go full throttle to get full torque. With cheapo controllers, a small but sudden throttle increase can produce maximum power.
 
Trouble is... a small amount of torque at low revs causes much greater acceleration than the same torque at higher speed
 
rogervize said:
Trouble is... a small amount of torque at low revs causes much greater acceleration than the same torque at higher speed
Click to expand...

Newton's Law

http://theory.uwinnipeg.ca/physics/rot/node5.html

equation 11 shows that rotational acceleration times rotational mass equals torque. Rotational velocity doesn't enter into it. The rotational acceleration is affected by the rotational mass and the torque.

This is the rotational equivalent to the linear: force = mass times acceleration. The acceleration is affected by the force and mass of the system, not the velocity.

Torque is proportional to motor current. So the trick is to make the throttle control the motor current.
 
At low road speed the amount of torque required for acceleration is less than the torque required at higher speed for the same acceleration.
Reason is .. at higher speed more force/torque is required to overcome wind and rolling resistance... leaving less torque available for acceleration.
I am not in conflict with Newton...
 
How fast are you riding?

Ability to accelerate does decrease with speed, but not primarily due to rolling resistance or aerodynamics at ebike speeds. The back EMF (due to increased speed) reduces the ability to drive motor current and therefore generate torque. So as speed increases the available maximum torque falls off. The controller's algorithms for controlling torque (throttle mapping) can deal with this in different ways, but there is far more torque available at low speed than at high speed due to this power conversion in the controller. This occurs even in a frictionless ebike.

But the overall expectation of the user with a torque throttle is met, this is considered normal for vehicle controls. Unlike the PWM (voltage) type throttles that are very jerky and hard to control on a powerful ebike.

Which torque mode controllers have you worked with? The several I've tried have been vastly better than the usual PWM (voltage) type controllers.
 
Alan B said:
How fast are you riding?

Ability to accelerate does decrease with speed, but not primarily due to rolling resistance or aerodynamics at ebike speeds. The back EMF (due to increased speed) reduces the ability to drive motor current and therefore generate torque. So as speed increases the available maximum torque falls off. The controller's algorithms for controlling torque (throttle mapping) can deal with this in different ways, but there is far more torque available at low speed than at high speed due to this power conversion in the controller. This occurs even in a frictionless ebike.

But the overall expectation of the user with a torque throttle is met, this is considered normal for vehicle controls. Unlike the PWM (voltage) type throttles that are very jerky and hard to control on a powerful ebike.

Which torque mode controllers have you worked with? The several I've tried have been vastly better than the usual PWM (voltage) type controllers.
Click to expand...

On my first build I used CA3 as a torque controller and it works OK, but I would like to run my second build without CA and instead use torque controller.
I'm hoping that my next build will be more streamlined.
Can you share you source and experience with torque controllers and different kind of throttles that you used?
This will help a ton.
 
The CAv3 is a very sophisticated piece of gear, and I own quite a few. They do go a long way towards "throttle taming".

It is not capable of "torque throttle" in any configuration that I am aware of.

It can be configured for "battery current" or "battery power" control, but that's not the same. It is a worthwhile improvement in control.

The controllers that I've used on ebikes that have "torque throttle" capability are the Sabvoton and the PhaseRunner. I have test driven various models of Zero electric motorcycles which also have torque control mode throttles, most likely using Sevcon controllers.

I would recommend the PhaseRunner as the best for most ebike applications.

My most recent experience with PhaseRunners is on the "Bonanza Bulldozer" which is documented in a thread linked in my signature, where I used a pair of PhaseRunners for 2wd operation on a mountain bike. The performance and control are excellent.

Hopefully we will see more "torque mode" controllers, but at the moment the PhaseRunner is probably the best game in town for ebikes that I'm aware of.
 
Alan B said:
The controllers that I've used on ebikes that have "torque throttle" capability are the Sabvoton and the PhaseRunner. I have test driven various models of Zero electric motorcycles which also have torque control mode throttles, most likely using Sevcon controllers.
Click to expand...
If building your own, the Lebowski controller brain chip does torque throttle, too, AFAI have been able to find out (haven't actually found a specific "feature list" so I am still reading every thread about the various builds of it to find out what I can).


Hopefully we will see more "torque mode" controllers, but at the moment the PhaseRunner is probably the best game in town for ebikes that I'm aware of.
Click to expand...
AFAICT, yes, as long as you have the budget for it. ;)
 
Yes Alan B.. you are correct.. and the characteristic tapering of winding current aggravates the problem.
I was thinking of my battery as a current source which would be true if it is sufficiently high in voltage to overcome the motor internally generated volts.
My project uses a car alternator and if I was clever I would control the field current (analog/linear) rather than just turning it off or on.
This would enable control of flux and give additional fine control over torque.
Otherwise I would suggest a current feed-back loop to achieve current drive behavior.
Not sure how ev manufacturers deal with heat losses at low RPM/high torque conditions... this would seem to me a big issue.
I'm sure that a mechanical gearbox/torque-converter is needed or else multiple windings with torque sensing.
 
Controllers that have Torque Mode Control are generally already measuring actual motor current, so they have actual torque information to feed back on and use to tame the throttle. How they map the available torque vs the throttle position and speed determines the system response. Physics determines the amount of torque available at different speeds, from semiconductor limited at low speed to voltage limited at high speed. They can do different mapping in between. One thing a few of them do is to add speed to the mapping so that dramatic changes in load doesn't spin the wheel up to maximum, which otherwise may occur when the wheel leaves the ground or hits some ice. This is often referred to as "torque plus speed control" where the torque control generally dominates but under low loading conditions the speed control element reduces spin-up.
 
Thanks Alan B.. Have you seen these algorithms executed using analog with combinational logic?
I am driving the windings at base frequency in the order of the phases. My motor/alternator is 14 pole.
Have you any ideas on using a car alternator for this purpose?
My plan is to use a friction drive to the rear wheel on bicycle. The drive disc is 160mm.
I am making a separate disc with magnets placed to provide position sensing.
...
 
rogervize said:
Thanks Alan B.. Have you seen these algorithms executed using analog with combinational logic?
I am driving the windings at base frequency in the order of the phases. My motor/alternator is 14 pole.
Have you any ideas on using a car alternator for this purpose?
My plan is to use a friction drive to the rear wheel on bicycle. The drive disc is 160mm.
I am making a separate disc with magnets placed to provide position sensing.
...
Click to expand...

I haven't seen that. ICs, micros or FPGAs are generally used. Motor current is generally sensed with current transformers.
 
Alan B said:
The CAv3 is a very sophisticated piece of gear, and I own quite a few. They do go a long way towards "throttle taming".

It is not capable of "torque throttle" in any configuration that I am aware of.

It can be configured for "battery current" or "battery power" control, but that's not the same. It is a worthwhile improvement in control.

The controllers that I've used on ebikes that have "torque throttle" capability are the Sabvoton and the PhaseRunner. I have test driven various models of Zero electric motorcycles which also have torque control mode throttles, most likely using Sevcon controllers.

I would recommend the PhaseRunner as the best for most ebike applications.

My most recent experience with PhaseRunners is on the "Bonanza Bulldozer" which is documented in a thread linked in my signature, where I used a pair of PhaseRunners for 2wd operation on a mountain bike. The performance and control are excellent.

Hopefully we will see more "torque mode" controllers, but at the moment the PhaseRunner is probably the best game in town for ebikes that I'm aware of.
Click to expand...

Thank you for your quick reply and recommendation. I really appreciate it. It looks like the PhaseRunner is what I was looking for.
 
You must log in or register to reply here.

Similar threads

J
throttle replacement problem
Replies
16
Views
582
J_T
J
J
Ultralight DIY mobility scooter hub motor and controller - AKM-Q75SX?
Replies
0
Views
215
joeka101
J
riding_on
The throttle signal to a motor controller (specifically: hacking it)
Replies
9
Views
416
amberwolf
amberwolf
O
Help identify old 8fun hub motor (bfswx02-p)
Replies
9
Views
167
AGS
AGS
ManicMods
Fardriver ND72260 Inconsistent Throttle from 0 RPM
Replies
21
Views
564
HumboldtRc
H
Back
Top