Throttle interface for sensitive high power setups
- Thread starter zombiess
- Start date
zombiess
10 MW
Looks like I only 2 left for those who would like to beta test. I would appreciate those who are buying to provide feedback, good, bad what ever, this is open for discussion in this thread and to install the units within two weeks of receiving them.
As long as its not snowing, I'll have it on the bike the day it shows up.
I keep blowing controllers up, so it's only going to be on a 12 FET on the extreme edge. It won't do power wheelies, so I don't know how big of a help it will be just yet.
Pots to dial out the slop in both ends might be cool. I never bothered with that, but it's easy enough to creep the bike at extreme low speeds with this controller. Just requires a death grip on a 1/2 twist. I'll report back findings when I have something conclusive.
I keep blowing controllers up, so it's only going to be on a 12 FET on the extreme edge. It won't do power wheelies, so I don't know how big of a help it will be just yet.
Pots to dial out the slop in both ends might be cool. I never bothered with that, but it's easy enough to creep the bike at extreme low speeds with this controller. Just requires a death grip on a 1/2 twist. I'll report back findings when I have something conclusive.
Tommy L
100 kW
Do you have another available?
I run 40s A123 on a Lyen 4115. I would really appreciate to test for you
Tommy L sends...
I run 40s A123 on a Lyen 4115. I would really appreciate to test for you
Tommy L sends...
zombiess said:
Naah, not that fancy. Mine is an old Nomadio React radio. Took 5 years for everybody to catch up with the same features Nomadio offered. Nomadio left the RC world for military contracts and their drones.
The radio let you create throttle curves. So do the RC ESC. Good stuff for drag racing and variable track conditions. You can also slow down the steering input, if you have a uber fast steering servo.
I just finished riding the Houston Critical Mass. I was pulling a BMX'er for a few miles with the GNG. Doing an average of 15mph and around 500W continuous. Oooh how I wish I had that throttle interface to tame the throttle jitters.
zombiess
10 MW
OK, all interfaces are now gone and will ship next week. Once you receive them I would appreciate your feedback by sharing you experiences in this thread.
zombiess
10 MW
All interfaces shipped out us mail today. Dang I'm busy lately. Working 80-100hrs a week.
zombiess
10 MW
Ugh, I totally forgot to tell you guys which connection does what for the wiring. Spent the time to write a manual but left this detail out.
So the square pads are the +5v line (from controller or throttle), next one is throttle signal, last is ground. All pads have the exact same orientation so it should be easy to figure out, if not I'll post a picture.
So the square pads are the +5v line (from controller or throttle), next one is throttle signal, last is ground. All pads have the exact same orientation so it should be easy to figure out, if not I'll post a picture.
zombiess
10 MW
snellemin said:zombiess said:
Touche!
I also wanted to let everyone know that I setup the pots identical to the settings I've been using on my own bike to give a starting point. The MinV (around 1.4V) and MaxV (around 3.4V) should be pretty close. Delay and buffer will be tuned to your own liking.
I have about 10 minutes of seat time with this installed. No adjustments were made aside from initial min/max throttle voltage. It's about 30ºF outside, so its not the best conditions to be going 50. I will be doing a more detailed review when I get everything dialed in, but this is just my initial reactions.
The board appears to be doing exactly what I assumed it would, aside from the substantial delay. Presumably most of the delay while it is ramping up can be tuned to my liking, but the delay when I snap off the throttle was very obvious right off the first twist. I got more comfortable with it very quickly, but it feels like the bike is going to run away for an 1/8 second. In practice, I don't think its an issue for anyone thats used to it, but it can feel significant. I'm assuming its how the board is set, but the delay in throttle response was making me twist the throttle too far and then back off.
The ramp up is very nice and smooth, and how the bike reacts when you snap the throttle WOT from a stop is great. The controller sounds happy about it. I think I can dial the ramp in response exactly where I want it with some time. Presumably when I get a controller big enough on the bike to effortlessly lift the front end, this could be beneficial. I was getting about 110-120A peaks, probably something in the 85V range, loaded, given the ambients and not terribly high SOC. It takes about 200A before I can't keep the front end down. It will probably be a while until I can test the board with a controller capable of that.
I should note, I'm used to riding WOT. Typically I flip the kill switch, snap the throttle WOT, and snap off when I want to slow down. I do a lot of full throttle cruising, and 95% of the acceleration is WOT. I suppose one could argue I don't need throttle acceleration for that kind of riding.
I think this board can be really great to go slow on >10kW bikes. Presumably some setups can be really twitchy at lower power levels, and it might do wonders for rough terrain, but I didn't test it on that yet.
The board appears to be doing exactly what I assumed it would, aside from the substantial delay. Presumably most of the delay while it is ramping up can be tuned to my liking, but the delay when I snap off the throttle was very obvious right off the first twist. I got more comfortable with it very quickly, but it feels like the bike is going to run away for an 1/8 second. In practice, I don't think its an issue for anyone thats used to it, but it can feel significant. I'm assuming its how the board is set, but the delay in throttle response was making me twist the throttle too far and then back off.
The ramp up is very nice and smooth, and how the bike reacts when you snap the throttle WOT from a stop is great. The controller sounds happy about it. I think I can dial the ramp in response exactly where I want it with some time. Presumably when I get a controller big enough on the bike to effortlessly lift the front end, this could be beneficial. I was getting about 110-120A peaks, probably something in the 85V range, loaded, given the ambients and not terribly high SOC. It takes about 200A before I can't keep the front end down. It will probably be a while until I can test the board with a controller capable of that.
I should note, I'm used to riding WOT. Typically I flip the kill switch, snap the throttle WOT, and snap off when I want to slow down. I do a lot of full throttle cruising, and 95% of the acceleration is WOT. I suppose one could argue I don't need throttle acceleration for that kind of riding.
I think this board can be really great to go slow on >10kW bikes. Presumably some setups can be really twitchy at lower power levels, and it might do wonders for rough terrain, but I didn't test it on that yet.
AussieJester
1 TW
looking forward to hooking the interface up, my new bike just needs chains fitted and its test ride time, very excited...
KiM
KiM
zombiess
10 MW
There is a around a 100mS delay at WOT that I recently noticed as well (I almost never ride WOT because there is no space), it only happens at WOT. I'll just have to tweak the code to eliminate it completely. Removing the 0.1uF cap either on the throttle input or output could eliminate it as well, you can either unsolder it or just snip it off. I'll do some testing on my scope/bike and let you which way to go this weekend.
Start playing with the settings, I sent them out with my own settings in them which is for a 6kW setup in a 20" wheel. It feels completely different compared to a 26" wheeled setup so you will most likely need to a make the ramp rate faster for the delay setting (clock wise).
I also found that when riding my bike now I use the throttle more like a motor cycle. I'm rarely full off the throttle any more, I just reduce it a some so I usually stay in some power going around corners. It does some getting use to riding like this after knowing how violent stock throttle setup can be.
Start playing with the settings, I sent them out with my own settings in them which is for a 6kW setup in a 20" wheel. It feels completely different compared to a 26" wheeled setup so you will most likely need to a make the ramp rate faster for the delay setting (clock wise).
I also found that when riding my bike now I use the throttle more like a motor cycle. I'm rarely full off the throttle any more, I just reduce it a some so I usually stay in some power going around corners. It does some getting use to riding like this after knowing how violent stock throttle setup can be.
zombiess
10 MW
This is irking me now. I can't figure out what is causing the 100 mS delay when slamming the throttle off. It's not noticeable when rolling off the throttle, just slamming it off has a 100mS delay. I'm using 2 timers, one running at 50mS intervals (20hz) to monitor and set the throttle, but speeding it up to 10 mS makes no change. The code in this high priority interrupt takes 600uS to execute. The second timer is variable and is controlled by throttle rate change and also the buffer pot. Changing the buffer setting seems to have no effect on this either. This code takes between 520uS to 620uS depending on what's going on.
Tried removing the caps, they aren't the culprit either. Will have to investigate it further. Good news is it only lasts 100-120mS which is literally quicker than the blink of an eye and it only occurs at 100% throttle, bad news is I don't know why.
Tried removing the caps, they aren't the culprit either. Will have to investigate it further. Good news is it only lasts 100-120mS which is literally quicker than the blink of an eye and it only occurs at 100% throttle, bad news is I don't know why.
Lebowski
10 MW
zombiess said:
just a wild stab in the dark but it has happened to me so... what could be happening is that when you slam the throttle off the
code gets confused and it hangs itself in an endless loop or something. Then you need to wait for the WatchDogTimer (100 msec
maybe ?) for the chip to reset itself...
zombiess
10 MW
Thanks for the ideal lebowski, I will check that. I think I have the wdt off but I'm not sure until I check my flags. Weird part is its not exactly 100ms. sometimes its 60mS and others its around 120mS. I ran into this in a previous version but don't remember what caused it. I'm going to reflas my old versions and find the culprit, then try to eliminate it. The hang only seems noticeable at wot on my bike, can't feel it at part throttle.
I will find it. Just glad its only a minor issue, this is why it's beta.
BTW if anyone has PIC programming hardware I can email you an updated hex file when I make changes to try out if you want.
I will find it. Just glad its only a minor issue, this is why it's beta.
BTW if anyone has PIC programming hardware I can email you an updated hex file when I make changes to try out if you want.
zombiess
10 MW
Full clockwise on delay will give you almost 1:1 on the acceleration, not much ramp at almost like stock. Full clockwise on buffer gives you extreme filtering on the throttle averaged 32 times over a really long buffer time so poor throttle response. In your video the buffer is working exactly as its supposed to, you have it set way to high.
Try setting buffer back to a really low setting and concentrate on the delay setting first then add in some delay to smooth the throttle. You need to do this and ride, you can't tell much without a real load.
Did you also follow the setup of Minv and Maxv? After they are set its good to use the reset button to ensure the scaling is set correctly, if the throttle isn't at its minimum position when it's powered on it will use the setting the throttle is at as the minimum causing a large gap before the throttle will respond.
What kind of power are you running, volts amps and wheel size?
Try setting buffer back to a really low setting and concentrate on the delay setting first then add in some delay to smooth the throttle. You need to do this and ride, you can't tell much without a real load.
Did you also follow the setup of Minv and Maxv? After they are set its good to use the reset button to ensure the scaling is set correctly, if the throttle isn't at its minimum position when it's powered on it will use the setting the throttle is at as the minimum causing a large gap before the throttle will respond.
What kind of power are you running, volts amps and wheel size?
Turns out the instructions are somewhat useful and fairly comprehensive.
I have the min and max voltage set about 1/8 turn away from ideal.
26" 5404, a little over 100A @ 24S LiCo.
I set the two adjustments to the minimum an it appears to be quite responsive. I should be able to mess around with it tomorrow.
I have the min and max voltage set about 1/8 turn away from ideal.
26" 5404, a little over 100A @ 24S LiCo.
I set the two adjustments to the minimum an it appears to be quite responsive. I should be able to mess around with it tomorrow.
zombiess
10 MW
ZOMGVTEK said:
You mean you put it on the bike and didn't read the instructions :lol:
Now I know why technical writers always seem to want to commit suicide and walk around mumbling RTFM. No offense meant, I work in IT and have been guilty of winging it without reading the manual first.
Hopefully you'll get some better response out of it now as it completely transformed my bike into something that would launch you on your ass from looking at the throttle in high power mode to something almost anyone can ride.
You might need to put a big yellow sheet in there that says RTFM. I assumed no documentation would be included, and ended up ripping the instructions in half opening the mailer. That certainly didn't help my odds of reading it.
Little arrows with +'s on the silkscreen might be a reasonable idea if you can fit it, especially since the important pots are reversed. This leaves no worry about future adjustments. 25 turn pots take a whole lot of turning to try and figure out whats going on. I was assuming they were 10 turns.
I might not be the best tester since I consider the bicycle to be extremely safe with the 12 FET controller. I can sit on a steep incline and snap the throttle WOT while leaning back, the bike just gently accelerates. A reasonable test would be to give the bike to an unsuspecting individual with no e-bike experience. They usually dump it immediately. The bike has a lot of mass to grunt up to speed, which I like. It really helps to mechanically smooth out harsh throttle application and most road vibration isnt transmitted to the rider as directly as a very light bike. Plus, the weight distribution is great so its not sketchy at all like my previous builds. I remember my first setup was a mess at a pathetic 2-3kW, but it had 60lb of PB on a rear rack, among other issues.
I'm not terribly sure if its my massive assumptions, or how the board is configured. I assumed it would make it more like a carburated ICE bike. Twist the throttle, and you know its trying to accelerate, but it takes a little bit for full power. It feels more to me just like a general delay. I think some sort of adjustable exponential response might be cool as well. I think it might be helpful to have a small shift past about 1/4 throttle to give better resolution at lower speeds. Typically resolution is not as necessary past 20-30 or so. Anyways, I think any throttle input should have SOME impact on the output irrespective of configuration, this way the rider is provided with some feedback like they would with a typical gasoline equivalent. I'd think always letting a little bit of the throttle input past the filter might be all it takes to make the board never feel like its slow. How the filtering feels right now, is as if it waits however many 'cycles' for the throttle to stabilize, then it snaps to that throttle position. This leaves the rider with no feedback, and a sudden jerk. This percentage to allow to pass would need to be configurable as different setups will require different throttle positions to have some sort of response out of the bike. I just feel like that instant throttle response I liked is diminished with the board set how I like for WOT snaps. I can set it up how I like for part throttle riding, but then its slow when I snap the throttle WOT.
Little arrows with +'s on the silkscreen might be a reasonable idea if you can fit it, especially since the important pots are reversed. This leaves no worry about future adjustments. 25 turn pots take a whole lot of turning to try and figure out whats going on. I was assuming they were 10 turns.
I might not be the best tester since I consider the bicycle to be extremely safe with the 12 FET controller. I can sit on a steep incline and snap the throttle WOT while leaning back, the bike just gently accelerates. A reasonable test would be to give the bike to an unsuspecting individual with no e-bike experience. They usually dump it immediately. The bike has a lot of mass to grunt up to speed, which I like. It really helps to mechanically smooth out harsh throttle application and most road vibration isnt transmitted to the rider as directly as a very light bike. Plus, the weight distribution is great so its not sketchy at all like my previous builds. I remember my first setup was a mess at a pathetic 2-3kW, but it had 60lb of PB on a rear rack, among other issues.
I'm not terribly sure if its my massive assumptions, or how the board is configured. I assumed it would make it more like a carburated ICE bike. Twist the throttle, and you know its trying to accelerate, but it takes a little bit for full power. It feels more to me just like a general delay. I think some sort of adjustable exponential response might be cool as well. I think it might be helpful to have a small shift past about 1/4 throttle to give better resolution at lower speeds. Typically resolution is not as necessary past 20-30 or so. Anyways, I think any throttle input should have SOME impact on the output irrespective of configuration, this way the rider is provided with some feedback like they would with a typical gasoline equivalent. I'd think always letting a little bit of the throttle input past the filter might be all it takes to make the board never feel like its slow. How the filtering feels right now, is as if it waits however many 'cycles' for the throttle to stabilize, then it snaps to that throttle position. This leaves the rider with no feedback, and a sudden jerk. This percentage to allow to pass would need to be configurable as different setups will require different throttle positions to have some sort of response out of the bike. I just feel like that instant throttle response I liked is diminished with the board set how I like for WOT snaps. I can set it up how I like for part throttle riding, but then its slow when I snap the throttle WOT.
Similar threads
