zombiess
10 MW
Connectors aren't much of a worry yet. I just need to make sure the darn thing works as I expect it to. Then onto phase 2 
Connectors will be phase 3 or 4.
Connectors will be phase 3 or 4.
Throttle interface for sensitive high power setups
- Thread starter zombiess
- Start date
Neat. This looks like it's working quite well.
I don't see connectors being an issue. This isn't really useful on low power bikes, since a few hundred unintentional watts is barely noticeable. People using several kW setups should be able to solder 3 wires. Plus, it will make it cheaper and easier to stock/order.
I don't see connectors being an issue. This isn't really useful on low power bikes, since a few hundred unintentional watts is barely noticeable. People using several kW setups should be able to solder 3 wires. Plus, it will make it cheaper and easier to stock/order.
zombiess
10 MW
I didn't get to test it on my bike tonight, but not for lack of trying. I'm missing one stupid connector to connect up the throttle (and am not going to cannibalize any of my working stuff for it). Got the one made and installed, hot glued the important parts to the breadboard so nothing comes loose during testing and then couldn't find the female molex connector I use on my throttles. Drove to Fry's Electronics who has them in stock and forgot my darn wallet. Gave up for the night because I got there right before they closed. Should be testing it tomorrow and posting results good or bad after I pickup the correct connector when I'm done with work unless forces conspire against me. I'm anxious to test this out. Might not be as good as a Sevcon with it's awesome throttle I've tried on Farfle's 2Pi, but it's gotta be better than stock at my power level.
If this works I'll get some test units made for others to try out and review.
I'm also wondering if this will help improve WH/Mi when making small speed changes by preventing the amps from shooting up while trying to reach the new target speed. I'm hoping I can get good enough throttle response that I can always run around on high power and not have to worry about switching modes as often as I do, that would be fun.
If this works I'll get some test units made for others to try out and review.
I'm also wondering if this will help improve WH/Mi when making small speed changes by preventing the amps from shooting up while trying to reach the new target speed. I'm hoping I can get good enough throttle response that I can always run around on high power and not have to worry about switching modes as often as I do, that would be fun.
zombiess
10 MW
Finally got the unit on the bike and tested it tonight. Short version, it makes a huge difference in how much throttle control you have, especially in the low range. I had no problem at all starting from a dead stop and taking off smoothly while running 7000 watts. It also gave me a larger throttle range with no dead zones because I could use the trim pots to set the start and end voltage. I don't need to run a power selector switch if I don't want to because the throttle control is that good when starting from a stand still even on 105% throttle. It is also 100% compatible with the 3 speed selector switch settings, worked as expected in every setting.
Now the long version for those interested. The delay function works but not the way I want it to. After letting off the throttle it tends to make the throttle too laggy if it's set too high. I have a solution for this I'll try to implement over the next few days. With or without the delay the mid range is a little on the touchy side still but I have several idea on how to solve that now that the interface is working. It could also be because I didn't have my throttle range set correctly so I need to double check that.
First problem I encountered was that I forgot to set the some variables to zero on initialization so the wheel would start spinning. I fully expected this to happen and had the bike upside down so it wasn't a surprise. Quick program change and that problem was solved. The second problem is the throttle seemed to be glitchy which seemed strange to me since it was so smooth on the bench. I figured it was noise on the throttle line so I added in a 0.1uF decoupling cap from the throttle signal input to the microprocessor to ground on an educated guess and that solved that problem; nice and smooth throttle response from then on. My EB318 controller has a very small voltage range over which it operates. With the controller hooked up the motor starts to turn at about 1.4V and is full throttle around 3.3V so that's where I set my controller to with it inline between the throttle and controller. When I took the interface off the bike and measured the range with it on the bench my throttle range reads 1.85V to 4.42V using the same settings but not connected to a controller. This surprised me, there must be some resistors that place a load on the throttle signal line, my output isn't buffered through an op amp, it's just a digital 10k ohm potentiometer so there might be some interaction going on. I'll probably try adding an LM358 op amp as a buffer to see if it makes any difference. The actual throttle itself when supplied with 5V has a low voltage of 0.83V at the minimum position and 4.29 at the max position, this is the same on or off the bike.
A little more work through next week and I should have something I consider solid to make a test batch of and send out to others to get their thoughts on it. I'm extremely happy with the results so far and with some of the changes I have in mind it should get even better.
Now the long version for those interested. The delay function works but not the way I want it to. After letting off the throttle it tends to make the throttle too laggy if it's set too high. I have a solution for this I'll try to implement over the next few days. With or without the delay the mid range is a little on the touchy side still but I have several idea on how to solve that now that the interface is working. It could also be because I didn't have my throttle range set correctly so I need to double check that.
First problem I encountered was that I forgot to set the some variables to zero on initialization so the wheel would start spinning. I fully expected this to happen and had the bike upside down so it wasn't a surprise. Quick program change and that problem was solved. The second problem is the throttle seemed to be glitchy which seemed strange to me since it was so smooth on the bench. I figured it was noise on the throttle line so I added in a 0.1uF decoupling cap from the throttle signal input to the microprocessor to ground on an educated guess and that solved that problem; nice and smooth throttle response from then on. My EB318 controller has a very small voltage range over which it operates. With the controller hooked up the motor starts to turn at about 1.4V and is full throttle around 3.3V so that's where I set my controller to with it inline between the throttle and controller. When I took the interface off the bike and measured the range with it on the bench my throttle range reads 1.85V to 4.42V using the same settings but not connected to a controller. This surprised me, there must be some resistors that place a load on the throttle signal line, my output isn't buffered through an op amp, it's just a digital 10k ohm potentiometer so there might be some interaction going on. I'll probably try adding an LM358 op amp as a buffer to see if it makes any difference. The actual throttle itself when supplied with 5V has a low voltage of 0.83V at the minimum position and 4.29 at the max position, this is the same on or off the bike.
A little more work through next week and I should have something I consider solid to make a test batch of and send out to others to get their thoughts on it. I'm extremely happy with the results so far and with some of the changes I have in mind it should get even better.
lostrack
100 W
Zombiess, this is fantastic work. Anks for developing something all of us high power users have been crying out for.
I for one will be glad to see one of these if you reach production stage soon. Throttle averaging would be fantastic on the bumps too.mas long as you say, it can release instantly (or you can install a throttle interrupt switch, activated by ebrake or similar, that instantly cuts to zero)
Nice work
Tom
I for one will be glad to see one of these if you reach production stage soon. Throttle averaging would be fantastic on the bumps too.mas long as you say, it can release instantly (or you can install a throttle interrupt switch, activated by ebrake or similar, that instantly cuts to zero)
Nice work
Tom
I am in awe of you guys on here that have this level of understanding of electronics! 8)
I would really like to try one of these on my mid mount when you have a few made, it is only 3Kw but on the bighit chassis with long travel suspension i get squat at the rear on the initial take up as the power does kick in a little more ferocously than i would like unless a lot of care is taken, any amount of care though is sometimes not enough when off roading on rough ground. This makes standing up in lower gears when doing technical riding a bit of a challenge, i shant stop in my quest to make my bike more user friendly untill it becomes the very capable off road bike i know it can be. I think the exponential voltage rise is the answer to an e-bike that "carburates" less than perfectly with the off the shelf throttle and controller setups that we currently use, i have allready trimmed the the throttle voltage to eliminate the dead spot which improved it a lot
I am also planning an RC build and have a cyclone equipped bike that lacks some finess in the initial take up so i would be happy to buy 3 of these off you when you have cracked the design.
Simon.
I would really like to try one of these on my mid mount when you have a few made, it is only 3Kw but on the bighit chassis with long travel suspension i get squat at the rear on the initial take up as the power does kick in a little more ferocously than i would like unless a lot of care is taken, any amount of care though is sometimes not enough when off roading on rough ground. This makes standing up in lower gears when doing technical riding a bit of a challenge, i shant stop in my quest to make my bike more user friendly untill it becomes the very capable off road bike i know it can be. I think the exponential voltage rise is the answer to an e-bike that "carburates" less than perfectly with the off the shelf throttle and controller setups that we currently use, i have allready trimmed the the throttle voltage to eliminate the dead spot which improved it a lot
I am also planning an RC build and have a cyclone equipped bike that lacks some finess in the initial take up so i would be happy to buy 3 of these off you when you have cracked the design.
Simon.
zombiess
10 MW
Here is an update.
I had a problem with the way I was calculating the exponential curve and not getting a true exponential curve causing some over sensitivity in the mid range I didn't like (but way better than stock). I believe I have that solved now. I'm building a version of this that I can install on my bike tonight. I want to start commuting with it this week to make sure all is well before I do the first run of prototype boards.
I haven't done much with it this the past week besides think of solutions to some problems since my Fiance moved in with me last weekend so we have been reorganizing the house so she's feels more at home I'm also studying for a Microsoft cert I have to get really soon so I can get a contract position. Once that's out of the way I'll be a lot less stressed. Haven't been enough hours in the day lately.
I'm trying to come up with a slightly better throttle algorithm, but I might just release it as is for now since I want to get this out there. Just converting the input to an exponential curve makes a huge difference in how easy it is to control your bike when using a Xie Chang based (Lyen, ebikes.ca, etc) controller.
I own both EB2XX controller and EB3XX controllers and this is compatible with both.
I had a problem with the way I was calculating the exponential curve and not getting a true exponential curve causing some over sensitivity in the mid range I didn't like (but way better than stock). I believe I have that solved now. I'm building a version of this that I can install on my bike tonight. I want to start commuting with it this week to make sure all is well before I do the first run of prototype boards.
I haven't done much with it this the past week besides think of solutions to some problems since my Fiance moved in with me last weekend so we have been reorganizing the house so she's feels more at home
I'm also studying for a Microsoft cert I have to get really soon so I can get a contract position. Once that's out of the way I'll be a lot less stressed. Haven't been enough hours in the day lately.I'm trying to come up with a slightly better throttle algorithm, but I might just release it as is for now since I want to get this out there. Just converting the input to an exponential curve makes a huge difference in how easy it is to control your bike when using a Xie Chang based (Lyen, ebikes.ca, etc) controller.
I own both EB2XX controller and EB3XX controllers and this is compatible with both.
zombiess
10 MW
Got the circuit built into a much more stable platform instead of on the bread board. It's now built on perfboard so I can install it on my bike for further tweaking. Next up is a quick batch of demo boards I can send out to a few people to test after I ride with this on my bike for a few days to check for bugs. This is a product I also plan to continue to improve on if I can find ways that make it nicer. I should have it installed tomorrow night and ride to work with it on Thursday.
Any feature requests? Right now these are the features:
Exponential throttle curve so there is more range at the low end and mid range
Set low voltage output minimum
Set high voltage output maximum
Set averaging filter & delay that can be turned off, (combined together and I'll probably limit the delay to a max of 0.5S, but test riding will let me know)
Settings are done with 3 trimmer pots to keep setup simple
Dual buffered throttle outputs so you can run two controllers off a single throttle, no more need to hack things together and splice wires.
It's pretty basic but it works much much better than stock. I can hand my 12KW bike to someone on the high setting and they have no problems starting off.
Any feature requests? Right now these are the features:
Exponential throttle curve so there is more range at the low end and mid range
Set low voltage output minimum
Set high voltage output maximum
Set averaging filter & delay that can be turned off, (combined together and I'll probably limit the delay to a max of 0.5S, but test riding will let me know)
Settings are done with 3 trimmer pots to keep setup simple
Dual buffered throttle outputs so you can run two controllers off a single throttle, no more need to hack things together and splice wires.
It's pretty basic but it works much much better than stock. I can hand my 12KW bike to someone on the high setting and they have no problems starting off.
Paul_G
10 kW
If possible you could build it right into the throttle and sell throttles at a far higher price/profit.
If you do that just remember I hate a 1/2 size throttle but thumb or full is fine....hehe
If you do that just remember I hate a 1/2 size throttle but thumb or full is fine....hehe
zombiess
10 MW
Paul_G said:
I'd have to have custom entire throttles built with enough room for the circuitry (even a surface mount version), that's not going to happen. Seen way too many bad things come out of China and money wasted (that I don't have to spare) on first and second attempts. First version is going to be all through hole components. Probably going to leave it up to the end user to enclose it, at least initially. An epoxy potted version would be good as well. All stuff to worry about after it's working the way I want. One step at a time.
binlagin
100 W
I want in on this!
zombiess
10 MW
Ran into a little bit of a snafu in the design. It appears my circuit is unable to feed the full WOT voltage to the controller when I use the op amp buffer due to the voltage drop associated with it and the fact drawing any power from the throttle line drops the 5.0V supply down to 4.3-4.6V. This only allows me to output about 3.1V from the op amp and I need to get 3.35V in order to reach full throttle. To fix this I have two solutions. Run a separate power supply for the op amp or just eliminate it. The controller works perfectly fine running off the resistor so I'm most likely going to take this approach unless I find a low voltage drop single power supply op amp. I like to buffer signals, especially those coming from a resistive output, but it works OK without it and drops my part count by one IC. It also loses the ability to drive two controllers, but that's not that big of a deal to me, just thought it would make a nice extra. I don't think it's going to be worth adding the extra complication since resistive throttles have proven to work just fine.
I also swapped from a 10k potentiometer to a 50k to see if there was any difference and the performance is the same, so that's good news.
Was hoping to ride it tomorrow, but now I have to wait and make the decision on what to do for the design.
I also swapped from a 10k potentiometer to a 50k to see if there was any difference and the performance is the same, so that's good news.
Was hoping to ride it tomorrow, but now I have to wait and make the decision on what to do for the design.
zombiess
10 MW
Made the decision to drop the op amp from the circuit. Rode the bike tonight on just the resistive setup. Immense improvement in throttle control down low, but the mid range is still quite sensitive if you let off the throttle an then re-engage it at the same speed. I'm going to go ahead with the first prototype circuit boards, but I'm going to work on some code to hopefully allow smoothing when re-engaging the throttle after coasting. Of course it could just be that my muscle memory is so use to the normal throttle that I just don't know where the right spot is. I'll try to think up an algorithm for it over the weekend, I already have a few ideas in mind. Might require an adjustment of 4 potentiometers total. Either way, I'd still rather ride my bike with this interface than not, low speed control is soooooo easy now.
binlagin
100 W
I appreciate this work your doing Zombiess!
But... I feel the mid range application of throttle is really the most important.
Getting on the throttle mid-corner right now is a butt-puckering experience... and that's really where I feel I loose out on a lot of speed. Not being able to apply the throttle to re-distribute weight and hit the apex through the corner.
Currently... with my crappy suspension and this twitchy throttle.. I slide my rear on any corner with speed(very dangerous). My suspension replacement is in the mail.. so hopefully that'll help it, but I really feel the motor could turn on little smoother through the corner.
Just my 2 cents... I'm sure anything is better then what we have to deal with now!
But... I feel the mid range application of throttle is really the most important.
Getting on the throttle mid-corner right now is a butt-puckering experience... and that's really where I feel I loose out on a lot of speed. Not being able to apply the throttle to re-distribute weight and hit the apex through the corner.
Currently... with my crappy suspension and this twitchy throttle.. I slide my rear on any corner with speed(very dangerous). My suspension replacement is in the mail.. so hopefully that'll help it, but I really feel the motor could turn on little smoother through the corner.
Just my 2 cents... I'm sure anything is better then what we have to deal with now!
zombiess
10 MW
Binlagin, I completely agree and have the exact same issues as you, the mid range is where it needs most of the work. I'm happy with the low and high ranges, now I just need to tweak the mid range. I'm working on it.
zombiess
10 MW
Still working on this and making good progress. Made a ton of changes this weekend but it's going to be a little while until I can wrap this up. At first I thought this project was going to be a simple conversion of the throttle input to a exponential output, some filtering, maybe a delay and it's done. It turns out that was pretty easy to do but does not solve all the issues I want solved. I've thought up some creative ideas on how to solve several issues I'm trying to implement now. Hopefully the wait will be worth it. I'm going to be working on this all this week, so much for a quick simple program. What looked good on my simulator worked just as expected, but not exactly the way I wanted once I tested it in actual hardware.
This is why things always need to be field tested. The original version that converts the throttle input to an exponential curve and does some filtering is on my bike now and I ride with it so I'm still field testing and the results have been great. No hiccups with the hardware, behaves just as expected on my EB318 controller.
This is why things always need to be field tested. The original version that converts the throttle input to an exponential curve and does some filtering is on my bike now and I ride with it so I'm still field testing and the results have been great. No hiccups with the hardware, behaves just as expected on my EB318 controller.
Lebowski
10 MW
you need to get started with my controller IC, then you don't need this
Am working on a PCB for my controller...
Am working on a PCB for my controller...
zombiess
10 MW
Lebowski said:you need to get started with my controller IC, then you don't need this
Am working on a PCB for my controller...
yes, I do need to start working on your controller because it's how I'd like to drive my mega power stage (which also needs to a PCB to be finished, I also need to build and test out my driver setup and have the PCBs for that done), but I need to start making money too and this is a useful device to many people so it's got my priority at the moment. I don't have the most secure job situation and now I have two people living on my income and my fiance/very soon to be wife is not legally allowed to work yet, so I have to maintain some priorities.
So many cool projects to work on, just need more time (and now some of my free time is spent with her, just as it should be).
I just passed my MCTS for Microsoft Hyper-V Virtualization so I should have some more time now since I needed a cert to land a part time contract gig I want. Maybe some time will free up in the near future and I can make even more progress.
lostrack
100 W
Any news zombie?
Alan B
100 GW
lostrack said:
He got married and received a shipment of Cromotors, so he's been busy...
Edit - apparently I misunderstood, sorry.
zombiess
10 MW
Not married yet, but soon. Just started a new part time job, got laid off from my regular one and converted to contract. Ive been so busy at the new part time job they havent even been able to give me an orientation and ive been there 3 days now.
Project is on hold until i can get some stability back in my life.
Project is on hold until i can get some stability back in my life.
Similar threads
