Throttle sensitivity High powered Bike

[johnnyz]

Hey Guys

Got a question....Since modifying a Motorino XMR with my Crystalyte Controller, the throttle at cruise or maintaining anything above 37 kph or so results in a very sensitive throttle...this thing has a lot of "kick" now that I am pulling 55 amps vs the 35 it was doing..plus the Lithium pack is 75 lbs lighter..BUT it is jerky when attempting to cruise at 40 kph or so...full throttle is smooth...anything below 32 kph is smooth...Im wondering why this is happening as I didnt feel this when running the same controller in my recumbent trike...perhaps Im feeling this because its got 2 wheels and therefore transmits the power directly to the rider?...I at first thought it might be the cycle analyst but it is not. I thought maybe the power wires were not properly connected but they are...

If this is "normal for a system pulling almost 5000 watts, then perhaps the latest version of the CA can smooth out the throttle ramping...

Does anyone have any ideas?...I guess its possible that the controller is damaged but cant see how...I also changed the throttle for brand new Crsytalyte.

Thanks

John

 

[johnnyz]

Well...I put the speed limit on the cycle Analyst at 40 kph..and it is smoother maintaining this speed than I can do with the throttle...alot smoother although not exactly free of little inconsistencies in speed flucuations which logically rules out some of the other things I though were wrong with the controller...I guess CA v3 here we come?


John

 

[johnnyz]

Im kind of surprised that I didnt get a single response to this post...actually I take that back, Im not surprised


J

 

[Alan B]

This is a common problem with high power ebikes.

A better throttle might help, or a torque mode if you have that option in the controller. A CA version 3 can be used to improve throttle response with various techniques.

 

[johnnyz]

This is a common problem with high power ebikes.

A better throttle might help, or a torque mode if you have that option in the controller. A CA version 3 can be used to improve throttle response with various techniques.

Thanks for the reply. I bought a CA v3 just for this purpose and cant get it to work in anything but legacy mode. When I hook it up so that the throttle is hooked up with the CA then I can see the slider and map the throttle voltages but no power to the motor. Ive emailed the guys at Grin but very very sparse responses and when I answer them in an attempt to troubleshoot I have to wait days for a 6 word answer that so far hasnt worked. Very frustrating and dissapointing.


John

 

[MadRhino]

Yep, various trhrottles will give different response. You will have to experiment and find one that you like. I have had that problem, and since I don't want any display or CA on my bikes, I've had to try many to find one I am comfortable with. Having very high power and 0 to 120 Kph on a third of a twist, is something one has to get used to. Usually, the twitch is at the begining of throttle course where the high power hits the motor on a start, then it gets easier to control as it is gaining speed. I say to those who are trying my bike: I'm gonna lift the tail and you find the "friction point on the throttle", it is like the clutch and the throttle are in the same hand and you need to master the acceleration with throttle only. It does work with some, but some others are f****ing scared of the bike after a first start. After years using the same half twist QO on all my bikes, I am now in perfect control of my throttles.

 

[johnnyz]

Yep, various trhrottles will give different response. You will have to experiment and find one that you like. I have had that problem, and since I don't want any display or CA on my bikes, I've had to try many to find one I am comfortable with. Having very high power and 0 to 120 Kph on a third of a twist, is something one has to get used to. Usually, the twitch is at the begining of throttle course where the high power hits the motor on a start, then it gets easier to control as it is gaining speed. I say to those who are trying my bike: I'm gonna lift the tail and you find the "friction point on the throttle", it is like the clutch and the throttle are in the same hand and you need to master the acceleration with throttle only. It does work with some, but some others are f****ing scared of the bike after a first start. After years using the same half twist QO on all my bikes, I am now in perfect control of my throttles.

Mmmm...well throttle is right there and really too torquey on the start..but its trying to MAINTAIN speed thats the problem...unless i maintain a death grip on the throttle and part of the handle bar just a fraction of a twist cause the bike to lurch ...acceleration is fantastic and smooth as is top speed and everything under 32 kph...worst is around 37 -40 kph...cant keep it steady...

John

 

[nutnspecial]

Hmm, that IS odd. I second everything Rhino said, but this doesn't seem quite your issue. A full twist throttle cut down to an inch or so of grip/twist (even shorter than 1/2 twist), and finished off with a lock-on grip is also helpful for accurate throttle control when there's alot of power and speed available.

Is the controller programmable, and/or could you try a 3 spd switch to just dial in a speed limit for the throttle? < it helps with starts too. CA will add so many possible features, but they are a challenge to hook up for the first time.

Surely you can get help on here with hooking it up from reading and/or participating in Tek's (or some other) CA thread?

 

[Punx0r]

Since it sounds like you have a bad spot in part of the throttle's range of motion, are you sure it's not just a bad throttle? Checking the voltage changes smoothly and proportionately with a multimeter through the full range of motion might work, although testing by substitution is probably easier and more effective.

In terms of general throttle sensitivity, I wonder if ensuring block-time is set to zero in the controller might help?

 

[johnnyz]

Since it sounds like you have a bad spot in part of the throttle's range of motion, are you sure it's not just a bad throttle? Checking the voltage changes smoothly and proportionately with a multimeter through the full range of motion might work, although testing by substitution is probably easier and more effective.

In terms of general throttle sensitivity, I wonder if ensuring block-time is set to zero in the controller might help?

Hi and thanks for the replies

Ive replaced the throttle and it is still doing it. Ive upped the voltage to 92 from 72 and the range at which it does this is pushed up a bit...its a Crystalyte 18 Fet controller...and in my trike I never had this issue (same controller and throttle)...mind you on my trike was the h4080 a motor wound for speed and this motor is def wound for torque

Can you please explain what you mean by "block time"?....If I can get hold of the vendor I can get a programming cable easy enough but wouldnt know how to program it...could it possibly be damaged???

Thanks

John

 

[Alan B]

Regarding getting the CAV3 working:

You can possibly get some help from the CA V3 thread on here, but it is hard for either Grin or a forum to help you with what is likely a simple wiring problem. If the throttle works with the controller, then a properly wired CAV3 should immediately work in pass-through mode. Is there someone local to you who can work on electrical issues? Once you get it working electrically then the forum can be of help to get the configuration correct. The problem may be the standard one that your controller is configured for legacy mode in which chase there are a couple of different ways to correct the situation. Teklektik has documented these choices very clearly in the unofficial manual, basically one way is to modify the internal wiring in the controller to accept the throttle input from the CA connector, the other way is to make a little adapter to bring the throttle wire out of the CAV3 and into the throttle input of the controller. This is what I did originally on the Borg.

I agree with others that it sounds like there is something else going on here with your throttle. A good throttle should give reasonable control throughout the range, but with a powerful motor the sensitivity of the throttle is increased, a small adjustment in the throttle can change the power a lot, though this is usually more of a problem at low throttle settings than at high ones.

The potentiometer throttles are generally more linear and have a wider range of adjustment, such as the Magura and Domino throttles. If they are properly spanned at both ends with fixed or variable resistors the fine control is optimized.

The real solution is to get a controller that has better control, the difference is significant and much smoother than using a high powered PWM type throttle controller which is not a good choice for a lightweight vehicle. Essentially you have a control system that has too much gain because the throttle input mode is wrong for the human interface. There are all kinds of band-aids that folks have invented such as three speed switches, throttle tamers, etc., but if the controller has the right torque control implemented (which cars went to years ago) it would not be the problem you are experiencing. Examples of better controllers are the PhaseRunner, Sevcon, Sabvoton, ASI models, Adaptto, etc. Anything with a true "torque throttle" will be much smoother.

I see you raised the voltage, this causes the power to increase nonlinearly as it goes with voltage squared. This does cause the system gain to be increased considerably.

 

[amberwolf]

Sorry I couldn't help.

previous throttle posts, for reference, for those trying to help:

https://endless-sphere.com/forums/search.php?st=0&sk=t&sd=d&sr=posts&author_id=26823

 

[johnnyz]

Regarding getting the CAV3 working:

You can possibly get some help from the CA V3 thread on here, but it is hard for either Grin or a forum to help you with what is likely a simple wiring problem. If the throttle works with the controller, then a properly wired CAV3 should immediately work in pass-through mode. Is there someone local to you who can work on electrical issues? Once you get it working electrically then the forum can be of help to get the configuration correct. The problem may be the standard one that your controller is configured for legacy mode in which chase there are a couple of different ways to correct the situation. Teklektik has documented these choices very clearly in the unofficial manual, basically one way is to modify the internal wiring in the controller to accept the throttle input from the CA connector, the other way is to make a little adapter to bring the throttle wire out of the CAV3 and into the throttle input of the controller. This is what I did originally on the Borg.

I agree with others that it sounds like there is something else going on here with your throttle. A good throttle should give reasonable control throughout the range, but with a powerful motor the sensitivity of the throttle is increased, a small adjustment in the throttle can change the power a lot, though this is usually more of a problem at low throttle settings than at high ones.

The potentiometer throttles are generally more linear and have a wider range of adjustment, such as the Magura and Domino throttles. If they are properly spanned at both ends with fixed or variable resistors the fine control is optimized.

The real solution is to get a controller that has better control, the difference is significant and much smoother than using a high powered PWM type throttle controller which is not a good choice for a lightweight vehicle. Essentially you have a control system that has too much gain because the throttle input mode is wrong for the human interface. There are all kinds of band-aids that folks have invented such as three speed switches, throttle tamers, etc., but if the controller has the right torque control implemented (which cars went to years ago) it would not be the problem you are experiencing. Examples of better controllers are the PhaseRunner, Sevcon, Sabvoton, ASI models, Adaptto, etc. Anything with a true "torque throttle" will be much smoother.

I see you raised the voltage, this causes the power to increase nonlinearly as it goes with voltage squared. This does cause the system gain to be increased considerably.

Thanks for the reply...My throttle is a crystalyte which is kind of a cheapie but goes with the controller..."too much gain"...yes thats what it feels like...for example if Im going 60 kph and then let off the throttle by lets say 30 % then when it slows down and reaches the speed to which the controller figures there needs to be power, it lurches for a second then off and then on and off very quickly giving a jerky feel...it smoothens out the closer I reach full throttle and is fine 30 kph and below...Ok...ill research these controllers but want something with some grunt (pull at least 55 amps) and take 100 volts.

Ill check out the difference in the throttles...I mapped it out with the CA and the voltages looked fine and linear but who knows? when doing this through the CAv3 no power to the motor so had to go the legacy route.

The vendor says the controller isnt damaged but can be programmed but dont know how to do it as I cant find a manual...


John

 

[johnnyz]

Hmm, that IS odd. I second everything Rhino said, but this doesn't seem quite your issue. A full twist throttle cut down to an inch or so of grip/twist (even shorter than 1/2 twist), and finished off with a lock-on grip is also helpful for accurate throttle control when there's alot of power and speed available.

Is the controller programmable, and/or could you try a 3 spd switch to just dial in a speed limit for the throttle? < it helps with starts too. CA will add so many possible features, but they are a challenge to hook up for the first time.

Surely you can get help on here with hooking it up from reading and/or participating in Tek's (or some other) CA thread?

I asked Tek for some advice but after a couple of tries, I found his responses to be pretty demeaning. I appreciated his help but he made it look like I was too dumb to follow instructions,and too dumb to ask the same questions none of which he answered. its all good Ill figure it out.


John

 

[nutnspecial]

Yes, the CA def takes some time and effort, on your own. It's a nice tool, but so far I haven't felt it was right for my builds and used it to monitor only. It def could be your controller wiring needs modded to run anything other than legacy mode. There's instructions for that if you wanna try, and it could fix the core issue by running thru it and playing with various configs, although what you describe isn't normal.

The issue, if throttle has been ruled out, is still a mystery. My knowledge isn't helpful past that but if you want another controller I can recommend an em3 18fet as an excellent simple speed controller. Get the 3spd switch and program cable too, imo. Maybe throttle too, but I prefer a simple twist with no other buttons etc.

Sounds like your controller would just need the cable and software to get into programming it, but if there is very little support and/or you can't find much in posts here it might be harder or less productive to mess with the settings. If it's xie chang programming I can send you em3's instructions to help explain some of their settings' purposes.

Really sounds like a weird glitch that the throttle would be so jumpy towards the end of it's range, on multiple throttles!

 

[Alan B]

I've used the CAV3 to convert the PWM throttle to a battery current throttle, or power throttle. It is an improvement over the PWM throttle, but it doesn't solve the low end completely because it doesn't take much battery current to get more low speed torque than you might want. But the high end is where it works well. The throttle is very smooth instead of jerky, and the rotation of the throttle determines how much power to add rather than what speed to run. It does add a little delay to the throttle, which is adjustable, but if set too fast it can oscillate.

It is difficult doing support via a forum for basic electronics wiring. It is best to find a local ebike electronics expert if you don't have electronics expertise. If you experiment you are very likely to damage some expensive gear. 100V is enough to do some serious damage and calls for some safety precautions. One solution is to get your controller professionally wired for the CAV3 and then it will be plug and play. Another is to buy or build the throttle adapter cable as I did. Or just get a quality full range linear throttle like a Magura, that will help some. But you still have to add some resistors and wire it up, so it is not plug and play.

Several solutions to your problem are already documented, they have been worked out and used by many before you. There are a number of choices. It is a common problem with too much voltage and power on a light vehicle for a PWM throttle. Many people just learn to control their wrist or thumb very precisely, which is similar to what is required to ride a high performance motorcycle. I don't know if you saw the viral video of Luke's deathbike that launched into the sky, but that's the same problem, magnified. Many others rode that bike that day, but the one rider in the video didn't have motorcycle experience and he failed to control the throttle. Too many KW per pound, high power to weight (and rotational inertia) ratio systems need smoother controls, or a lot of skill. The trike has more weight and more rotational inertia. Increasing the voltage makes a big difference. Of course the other solution is to drop the voltage back down, but you already know that.

Have fun and be safe,

 

[Punx0r]

Can you please explain what you mean by "block time"?....If I can get hold of the vendor I can get a programming cable easy enough but wouldnt know how to program it...could it possibly be damaged???

I've not used a crystalyte controller, but this thread indicates they can be programmed the same as the xie-chang(Infineon) controllers: https://endless-sphere.com/forums/viewtopic.php?f=2&t=80353&p=1196232&hilit=crystalyte+controller+programming#p1196232

Block time is the length of time the controller will allow a spike in current to the motor before imposing the current limit. I don't know if it's an entirely uncontrolled spike, but a spike in current is a spike in torque. I *speculate* that this will worsen a snappy/jerky throttle. Some documentation suggests the minimum value for this setting is 1 second, users mention setting it to 0. It's something I'm going to experiment with myself in the next few days. It could be a red-herring, though.

But yes, speed/PWM-based throttle is pretty nasty. I wonder if the Kelly controllers have a proper torque-based throttle nowadays?

Good tip on the Magura throttle, Alan. I didn't realise there were compatible potentiometer-based throttles out there, but thinking about it there's no reason why there wouldn't be

 

[nutnspecial]

I was wondering that on the magura pot throttles (and what 'pwm' stands for in relation to speed throttles).

It was my impression they only worked with xie chang when going thru CA? Or on some kelly's. I haven't heard overy impressive things about either of those forms of torque/power control though, for performance. But if a magura could work with xie chang after some simple mods that could be fun to try.

The programming of the crys controller would be the cheapest and easiest (and most natural) place to start imo, and that's a good call on block time. I can't think of much else to try in the programming except to confirm proper power and speed settings, and enable a 3spd switch if possible.

Keep at it John, let us know what you end up with please!

 

[MadRhino]

The recent Crystalyte controllers, those that are made to be hooked with the APM, are weird. Also, the recent Crystalyte throttles are made with a "smooth start" signal curve, that feels weird too. I had purchased a few of them and gave them away because of that. I believe that you will need the APM to reprogram the controller for the voltage that you are now feeding, and/or you should try another half twist throttle of another brand.

If you have a CA, you should read the threads about throttle management program. It is not so easy to program the new CA V3, for it has a much bigger software and it is not as intuitive as the older ones.

Some are using sensorless function with recent Crystalyte controllers. Never do that when feeding higher voltage and power. This is making it very twitchy and could fry a motor.

 

[Punx0r]

I must admit the 3-speed switch is a useful bandaid to the problem

 

[johnnyz]

I must admit the 3-speed switch is a useful bandaid to the problem

Wow...really appreciate all the answers and troubleshooting...just took it for a ride again...if I hold onto the throttle really tight then it smoothens out. A slight twist results in a rush of torque and then when it appears like the controller senses the speed at which that throttle position or voltage is done applying power you get a sudden "off" and its jerky...jerky on and jerky off...

Yes, If i can figure out how to get hold of Evens he can supply me with a 3 speed switch and the schematics on how to wire it to the controller, however I would need another controller. When I took it apart, one of the screws holding the FETS on broke and now I cant slide it out of the housing...so either I get another Clyte 18 Fet controller or switch to something else...

John

 

[johnnyz]

The recent Crystalyte controllers, those that are made to be hooked with the APM, are weird. Also, the recent Crystalyte throttles are made with a "smooth start" signal curve, that feels weird too. I had purchased a few of them and gave them away because of that. I believe that you will need the APM to reprogram the controller for the voltage that you are now feeding, and/or you should try another half twist throttle of another brand.

If you have a CA, you should read the threads about throttle management program. It is not so easy to program the new CA V3, for it has a much bigger software and it is not as intuitive as the older ones.

Some are using sensorless function with recent Crystalyte controllers. Never do that when feeding higher voltage and power. This is making it very twitchy and could fry a motor.

Ok this makes sense...when you said they are made with a "smooth start" curve...even though initially it jerks from standstill if you give it any more than 5 degrees of twist, if your easy then it feels easy and the "rush" of torque is not there from the start unless I give it more than a little throttle...its in the "middle" of the throttle band where it gets really torquey.....

 

[Paladin]

I had a similar issue with my CA3, in that the settings I had made felt like the CA3 would get to speed then just cut throttle.
I had to many variables set. Speed/Amps/Watts.
Seems like they work better with less options to limit actions. IE; just Amp limit.
Try running the CA3 with minimal setting at first then add to it one at a time till you get the surging (throttle cutouts) then back it off.

Keep at it these CA3 are a really nice item, you guys that don't use these on hi-power bikes are nutts.
With the throttle ramping there is no wheelies...if you want.
I have not even put the all-important torque arm on my last bike because the CA3 limits any out of control wheel spin.

 

[Alan B]

The block time setting on the Xie Chang controllers turns off the current limiting at the start for a few seconds. It has no effect after that. It makes stronger starts, and causes controllers to blow up if you have enough voltage and low enough impedance in the motor since the current soars without limit. Setting it to zero is a good idea, but it may not turn it off since they have an internal minimum. I had one 24 FET controller blow 5 times and every time it blew it was at the start. Block time was set to zero, but it still had a current surge.

To use a Magura or Domino throttle with a standard controller requires a couple of fixed resistors or trimpots, one at each end, to span the throttle to a range of about 1-4V. The full 0-5V of a bare Magura is wider range than the controller expects. I use trimpots and then adjust them to get the full range that the controller uses. It really helps.

The standard PWM throttle is essentially a speed throttle, and works as is described by the OP in this thread. The controller puts out crazy power to accelerate and then shuts of when the target speed is reached, then it slows down and comes back on, and so forth.

 

[johnnyz]

The block time setting on the Xie Chang controllers turns off the current limiting at the start for a few seconds. It has no effect after that. It makes stronger starts, and causes controllers to blow up if you have enough voltage and low enough impedance in the motor since the current soars without limit. Setting it to zero is a good idea, but it may not turn it off since they have an internal minimum. I had one 24 FET controller blow 5 times and every time it blew it was at the start. Block time was set to zero, but it still had a current surge.

To use a Magura or Domino throttle with a standard controller requires a couple of fixed resistors or trimpots, one at each end, to span the throttle to a range of about 1-4V. The full 0-5V of a bare Magura is wider range than the controller expects. I use trimpots and then adjust them to get the full range that the controller uses. It really helps.

The standard PWM throttle is essentially a speed throttle, and works as is described by the OP in this thread. The controller puts out crazy power to accelerate and then shuts of when the target speed is reached, then it slows down and comes back on, and so forth.

Well this explanation mes the most sense of what's going on...however what is the difference between the Crystalyte throttle I have on it now and the Magura throttle you mention?....do they not both work on voltage (the crystalyte is .086 to 4.25 volts)???...dont understand how using this other throttle would make any difference...

John