Ninja 250 "Clone" Made in China

In regards to the slow acceleration...you can try adjusting the settings for acceleration time under the throttle menu.

Keep this in mind (from the sabvoton manual): The less the “acceleration time” or “deceleration time” is, the faster the throttle will response ,(It is forbidden that the time set over 2000ms).

If you go less than 2000ms the throttle will instantly go to full throttle. Be sure to have the rear wheel suspended off the ground when tuning and testing.
 
jansevr said:
In regards to the slow acceleration...you can try adjusting the settings for acceleration time under the throttle menu.

Keep this in mind (from the sabvoton manual): The less the “acceleration time” or “deceleration time” is, the faster the throttle will response ,(It is forbidden that the time set over 2000ms).

If you go less than 2000ms the throttle will instantly go to full throttle. Be sure to have the rear wheel suspended off the ground when tuning and testing.

It doesn't exactly work like that, unfortunately.
This parameter only sets the throttle response time, which means the controller will determine the average value of the throttle potentiometer input by sampling during this time. It just makes the throttle more reactive, but it doesn't help adding more power. I usually set my throttle response time to 50ms both for acceleration and deceleration.

Anyways, the acceleration is not bad at all, it is still way better than a 250cc gas one, just that it lacks a bit of kick from standstill. I think this is a controller issue, sabvotons controller are not programmed to give full power from zero. I just hope there is a clever way to trick the controller to give more power during this phase.

Changing for a smaller diameter wheel is out of the question, the bike would look weird, it would lower its ground clearance and impact handling. But sure, purely speaking about acceleration, it would work. :)
 
If you guys wanna see the performance on the street check out the videos of Dui and I against a 400cc bike.

Soon another video with the Mobipus 72600. We'll need to find a 600cc for that one!


https://www.youtube.com/watch?v=xRTDp5tUqaY

https://www.youtube.com/watch?v=IUS4IMGhe3E

And this one is for whoever said lead acid batteries are good for nothing:

https://www.youtube.com/watch?v=pFSrPVXyVR8
 
A Chuck Norris kind of fact, yeah
 
If lead acids are so bad, why is it that millions are still manufactured every day? They are not as bad as some people might think.
 
Well diesel fuel and and that tar like bunker oil that ships are using are also produced each and every day. That does not make em good for you or me. Same with the diesel engines. Mass produced but still there are better alternatives.
 
armandd said:
If lead acids are so bad, why is it that millions are still manufactured every day? They are not as bad as some people might think.

Lead-acid Batteries are a huge waste. Especially for EVs.

I mean 250 Cycles@100% DOD and only 30-40wh/kg which gets cut in half if you discharge at 1C or greater, it's just laughable.
Also Lead-acid batteries degrade slowly at above 25C. And in Winter, you can consider yourself lucky if you make it to your local Walmart.

The reason they are still made today is because they can be made cheaply. People then buy Lead-acid batteries without even knowing about the shortcomings, only to dispose of them within 2 or so years.

I'm not saying Lead-acid Batteries don't have their uses. But I do feel sorry for people that put them in their EVs.
 
You only have to look at what lead powered electric cars were like, versus what lithium powered electric cars are like, to understand the implications.
 
Hi guys,

Instead of wasting some nice energy to debate about evil SLA vs good Lithium here, what would you say if we used all this brain power to help me mod my controller here instead: https://endless-sphere.com/forums/viewtopic.php?f=30&t=95846

:wink: :wink:
 
Dauntless said:
Heavens! It's the arguing SLA's that gets their blood going. Now they are ready for full contact Latin dancing on your controller.

You don't know how hard I laughed at this :lol:
 
Lower Kv would give you initial kick but with such ebike it would be disappointing loosing top speed.
I was considering that Sabvoton72150 does it better than Kelly72300. Especially from stand still.
Overall 72150 too small for your hub or was it 72200 than it's kind ok considering price. People saying their Sabvaton fry, mine new Sabvaton fried too, but in my case I think it happened due lack of diode on relay (contactor). There is magnetic electric voltage spike every time you disengage relay so some controllers are susceptible to it and fry, like mine did.
Also you need FF inside your hub unless liquid / water cooling is an easy and an option for you, than it's even better.
Sabvoton software is primitive (basic), there is nothing to change in it to gain your sought initial acceleration. Similarly like any other controller it is designed, programmed to preserve hardware (controller itself and motor) by not pushing Amps too fast when you crank throttle or maybe parts themselves are not enough, but I think it is programmed to deliver considerably subtle wattage in very low rpms. Looking from manufacturer and warranty point of view - softer start, less prone to brake.
In other your thread, I read, somebody suggest that it is possible to change parts: Fets and caps and make it more powerful but even if you would do it, theoretically speaking, it still would be black box (you don't know what inside, what outcome, what you will gain with it). AFAIK you can change parameters which are in their proprietary Sabvoton application only. But I rather live with generic controller than with something like Sevcon. It has tons of parameters to change. It can rape any motor, but one needs be become an expert to understand what it can do and what to change.

Seeking for initial kick from 0-10 km/h. Maybe APT would give it. Not too expensivs - ''APT Programmable FOC AE96600''. This one is a brick but 600A phase.

Maybe it does not apply to your hub as yours has wider magnets but:
I had change to run qs273 v2 11kv 28H magnets in 16'' rim with huge tire coupled with Sabvoton 72150, locked version. Within battery and phase amp range there was a point from which amp increase had no noticeable dynamic increase, effect.
 
minde28383 said:
Overall 72150 too small for your hub or was it 72200 than it's kind ok considering price.


It's the "72200", not really an official version since it didn't really exist at the time, but a modified version of the 72150, with unlocked firmware where you can input figures >150A
So far I have it set up at 210Amps battery and 510Amps per phase. It's been like that for more than a year already, driving everyday and it seems just fine with it. Never tried to push it more since I don't know what it the actual theoretical limit of the FETs and I suppose I'm close to it.
It does heat a bit during the hot days of summer in which case I can reach the temp limit after around 15-20 minutes of hard ride, but nothing crazy. And its not even water cooled.

minde28383 said:
People saying their Sabvaton fry, mine new Sabvaton fried too, but in my case I think it happened due lack of diode on relay (contactor). There is magnetic electric voltage spike every time you disengage relay so some controllers are susceptible to it and fry, like mine did.

Never had any problem with mine.
I had problems with my other one, but I think it was due to me welding my chassis without having disconnected the controller (extremely stupid mistake). The controller still works 99% of the time, but sometimes, randomly, it doesn't want to start (the controller turns on, but nothing happens when you turn the throttle). Wait a few hours and it generally will start again. No error messages. Tried everything already, changing a few throttles, hall sensors, re done the wiring... But 99% of the time it still works fine. Super weird problem.
But before that it worked flawlessly for about 3 years. I even inverted the positive and negative terminals from the battery once, it made a gigantic spark but the controler was still working fine. This thing is tough.
I don't really know how people managed to fry theirs, I'm using Sabvoton controllers for about 5 years and never been able to destroy one so far despite being pretty hardcore with them.

minde28383 said:
Also you need FF inside your hub unless liquid / water cooling is an easy and an option for you, than it's even better.

The motor barely heats, for now it is not really a problem
It is probably the bright side of the limited power from 0-10kmh :)

minde28383 said:
Sabvoton software is primitive (basic), there is nothing to change in it to gain your sought initial acceleration.

That's actually the only thing it misses in my opinion. I wouldn't say it's primitive given the fact that it is a no brainer for the user 99% of the time, you just plug everything, click on "test" and the software will determine your motor parameters for you, you don't have to guess or mess up with specs you don't understand it's pretty advanced in that regard.
Other than that the fact you cannot change the startup torque I think the software is good enough, it's easy to use and quite foolproof, any changes you make don't need a reset of the controller so you can tweak in real time which is a huge, huge advantage. You can tweak the settings wile riding the machine if you want to. I don't remember any other controller being able to do that.
Like you, I think that the startup torque is limited to protect the controller and motor from unexperienced people and to improve the controller life expectancy. It's a bit annoying but I can live with that actually, now that I'm used to it doesn't really matter so much to me. It reminds me of turbo engines with turbo lag, not much until you reach 3000 rpm and then BOOM! Except it doesn't arrive abruptly, its more like a smooth wave of torque.

minde28383 said:
In other your thread, I read, somebody suggest that it is possible to change parts: Fets and caps and make it more powerful but even if you would do it, theoretically speaking, it still would be black box (you don't know what inside, what outcome, what you will gain with it).

Well, yes and no.
Even if you don't know how the software is programmed, you know for sure that it takes its decisions based on some inputs. Those inputs will mostly be voltage and current values. Those values will come from different sensors located somewhere on the controller. So the idea here is to trick the controller, supplying it with altered values. For example, lowering the resistance of a shunt sensor to make the controller believing that it gives less current than it is actually giving. Or bypassing the current sensors for a similar result.
Actually I'm pretty sure you could do almost anything you want with an external microcontroller, gathering the value of these sensors, then mapping them according to a map of your choice before finally feeding them to the actual microcontroller of the sabvoton.
Let's say for instance that the current sensor detects right now a current of 200Amps on phase A, you feed that value to the 1st microcontroller who turns it into, let's say 150A and then feeds this value to the Sabby microcontroller. The sabby will then think it didn't reach its actual target of 200A, it will widen the PWM duty cycle as a consequence and the FETs will be opened for longer. In the end the limit will be the heat capacity of your mosfets as well as your motor saturation point. At least I think so.
Sure, you'd need a clean code who would run fast enough not to delay the information too much but I think this should be doable.
I'm unfortunately unable to test that because I just have one controller that I actually need everyday to go to work, plus I'm not entirely sure I'd be qualified enough to tackle this challenge.

minde28383 said:
Seeking for initial kick from 0-10 km/h. Maybe APT would give it. Not too expensivs - ''APT Programmable FOC AE96600''. This one is a brick but 600A phase.

600A/phase is not that much different than 510A/phase. Sure It might be slightly better but nothing really insane.
But maybe the power curve is different, I don't know,

minde28383 said:
Maybe it does not apply to your hub as yours has wider magnets but:
I had change to run qs273 v2 11kv 28H magnets in 16'' rim with huge tire coupled with Sabvoton 72150, locked version. Within battery and phase amp range there was a point from which amp increase had no noticeable dynamic increase, effect.

Could be your motor saturating (in which case it should run very hot), but I had a similar issue until I discovered that there are two parameters that work together, I think it was the "rated current" and the "max limit current" or something. Changing one without changing the other wasn't making any difference passed some point, but then changing the other made a difference. I believe that was because the controller was hitting one of the two limits first, so whatever parameter you entered couldnt' be reached in real conditions.

If you post some screenshots of your config page I can help you if you want.
 
What's BMS model?
Mine is ANT 32S 120A 300A. Does you app works with it? What's different in your app?
 
Vortecks said:
I've just revisited this post and I've noticed you're using the same BMS as me.
I wasn't happy with the Chinese app so I wrote my own - https://play.google.com/store/apps/details?id=com.vortecks.vbms
Drop me a PM and I'll send you the latest version if you want :D

Thanks a lot for the offer, very much appreciated!
I'm unfortunately the kind of sucker who still uses an Iphone4 so I just can't use it :(
But if one of these days I get out of the dark side I'll send you a PM, for sure :)
Thanks again and nice work on the app
 
Hello to those who follow my thread :)

A few updates of the work made on the motorbike during those past few months:
Basically 4 projects:
-I've re-built all the wiring harness
-I've upgraded a few components
-I've built some new endcaps for the handles
-I'm building a new battery box

So let's start with the first one, the wiring harness.

The old harness was working fine, nothing major to complain about, but it was still the old wiring harness of the petrol bike. So I had a lot of useless wires, some of them were still "hot" with either 12 or 72V, and I thought it was quite a risk to keep them. Also, none of the plugs were waterproof which sometimes led me to a few small issues during heavy rain (unresponsive throttle mainly, but also unreliable lights).

So I've bought some automotive grade plugs, as well as suitable wires. All of them are properly crimped, insulated with rubber gaskets and heat shrink, and covered with nylon sleeves to protect them from abrasion.
I've also installed a more powerfull DC/DC converter, the new one can deliver 20Amps @ 12V, which will be useful later for my new battery box.

Some pics of the work done:
The new plugs:
IMG_5752.JPG

The motor harness (halls and phases):
View attachment 21

Most of the wiring plugged into the controller:
View attachment 20

The BMS box (3D printed):
IMG_5762.JPG

This new wiring seems to work very well so far, nothing to complain. It does look much cleaner than the mess I had and I feel much more confident about it now.


Next project was to upgrade a few components, mainly the brake system and the rear shock.
There was nothing really wrong with the original brake levers or master cylinders, but I had different ones on both sides and I couldn't adjust the lever position much. So I switched to some radial master cylinders, which, so far, are very nice.
The original rear shock was a real brick, and it did not allow for any adjustments. I bought a gas shock instead, which makes an incredible difference comfort wise. I still need to adjust it to make it a bit stiffer though, I found the spring is a bit too soft, maybe it was intended for a lighter bike, I don't know.

Anyways, some pics:

All the new hardware:
IMG_5808.JPG

The new shock installed:
IMG_5810.JPG

The new brake lines installed, it was a real pain to bleed the front but I managed to do it eventually. The levers are now rock hard:
View attachment 16

The levers are supposed to be titanium, no idea if this is true but I think they look pretty nice on the bike:
View attachment 15
IMG_5828.JPG
IMG_5832.JPG


Next project was the handle endcaps. I've lost mine a long time ago while riding, and didn't want to purchase new ones since all the ones I could find were big and bulky, for no reason. I don't want to add an useless 3cm each side on my handles since it wouldn't bring anything but inconveniences. So I decided to just create my own using my mini lathe.

I'm not very good at using a lathe, this is actually one of my first projects with it, so forgive me if the result is not up to professional standards, but anyway in the end they fulfill their task and I'm pretty pleased with them

Some pics:

Before:
View attachment 9

Lathe work:
IMG_5850.JPG

Final product:
IMG_5852.JPG

Installed on the bike:
IMG_5854.JPG

View attachment 12

Last but not least is the battery box, which is still an ongoing project.
Right now, my battery box is just a big steel tub, which weights around 5.5kg and doesn't protect the battery from rain, cold, heat, dust or anything.
My plan is to create an insulated, air conditioned carbon fiber box.

The goal will be to keep the battery at the best possible operating temperature at all times. To do this, I will use some Peltier modules, which can provide both cooling and heating, depending on which way you supply them with voltage.
The battery will be insulated inside the box, using aerogel blankets, thanks to the advice of someone here in the forum (sorry I forgot your name :)).
The Peltier modules will be controlled by an arduino nano and relays, which will take decisions based on temperature sensors, main power plug sensor, maybe a time and date module who would know my usual work schedule and some switches.

I'm just at the first stages of the code right now and I quite suck at it, but I have a good idea of how this should work. Basically, the main idea will be to do most of the hard work while the charging plug is connected, and try to rely as much as possible on insulation and thermal mass while unconnected. I'll explain more about the control strategy when the thing will actually be operational.

The device looks like that right now:
IMG_5872.JPG

It has 2 Pletier modules and draws around 6 Amps @ 12V at full power, which is reasonable. The Peltier module will be watercooled in order to make it as efficient as possible. This will add a little to the overall weight, but I will also watercool the controller so that won't make much of a difference in the end.

Next step is to create the battery box, and this is quite the hardest step. First I've removed the battery case temporarily, in order to create a foam mockup:

View attachment 7
IMG_5868.JPG

The mockup is slightly bigger than the battery to accomodate for the insulation. I hope this will be enough because space was a bit limited and I can't do much more without having to cut the chassis:
IMG_5870.JPG

Then a bit of squaring, sanding, and I added this little box on the front which will house the cooling system:
View attachment 3
IMG_5876.JPG

I'm using what is called the "lost foam method" to build my final box. Basically, you create the shape out of polystyrene foam, wrap it in packing tape. Then I'll put a few layers of epoxy coated fiberglass in order to make it a bit more rigid. I'll do a test fitting to make sure everything is okay, then cover by a few layers of carbon fiber, place it under vacuum (I think I'm going to use some vacuum bags for clothes and a vacuum cleaner), and let it cure.
Once cured, I'll create the openings, and dissolve the foam with acetone.

So, the first step was to put the tape all over:
IMG_5880.JPG

And right now I'm laminating the glassfiber layers:
IMG_5887.JPG

It's a bit of a mess to work with, epoxy takes too long to cure so it doesn't stick very well in the corners. I'll do a bit of sanding before the carbon layers. Not really sure if this will work fine, I'm quite experienced with glassfiber and polyester, but glassfiber and epoxy is a quite different story. I've never done vacuum bagging either so I hope I won't mess up since carbon cloth is quite expensive.
Anyway, I'll keep you guys updated on the progress!
 
I like perfection and this is heading that way :D
Coincidentally, I've only this weekend used the same automotive connectors on mine!
 
That looks very nice now!
Good job :bigthumb:
 
John in CR said:
You will accomplish exactly nothing by swapping to a slower wind version of the same motor. "High torque versions" are absolute myth. The reason is the windings are longer and thinner copper, so they handle proportionately less current for the same amount of heat, and heat is our hubbies' true limitation.

When the current capacity of the controller and wiring harness are taken into account, the limitation of a motor for a given speed range can be totally unrelated to heat. Except to the degree that a motor bogged down way below its free speed, such as you suggest, would turn most of the power delivered to it into heat.
 
Looking good so far, any updates on the battery box? And what was the slow acceleron like with the other controllers, have done a shunt mod to get more amps.
 
Hardergamer said:
Looking good so far, any updates on the battery box? And what was the slow acceleron like with the other controllers, have done a shunt mod to get more amps.

Quite slow progress, I'm experimenting with the carbon fiber.
I tried the ghetto vaccum molding just to get a basic test before doing it on the real battery box.

So I made a little enclosure in order to house the electronics which will control the air conditioning system. First I carved the shape out of foam, then covered it with 2 layers of thin glassfiber. I then simply put the ting in a vacuum bag (this is just a vacuum bag for clothes that I had laying around), pumped out the air with a hand pump and let it to cure overnight.

This is the bag and pump I used, pretty cheap stuff:
View attachment 5

The next day, I removed the cured part from the bag. I was a bit afraid it would be a real mess, but to my great surprise it peeled off from the part extremely easily. I can even reuse the bag, it was not damaged at all.
I ended up with a pretty nice part:
View attachment 7

I had quite a lot of sanding to do though, these bags work ok, but they are not very elastic, so I don't really recommend using them unless you are ready to spend a lot of time sanding or if your part has a fairly simple shape, without much corners. Tight corners won't come out very well, because the bag doesn't stretch enough to reach them, you end up with a lot of epoxy to sand.

Anyway, after a fairly long amount of sanding, it eventually ended up like this:
IMG_5927.JPG

The part was still quite flexible and weak at this point, the glassfiber is not very tough.
So the next step was to lay up the carbon fiber, I used some 3M spray glue to keep it in place, this works very well:
IMG_5930.JPG

Then wet it thouroughly with epoxy, and place it in the vacuum bag again:
View attachment 4

After curing, I removed it from the bag, again, the bag went off very easily:
IMG_5933.JPG

The part is globally good, but there is a lot of pinholes and tiny bubbles in the epoxy. Next time I'll try heating the epoxy a bit and putting it in a vacuum bag for a few minutes to hopefully get better results.

Right now, I've almost finished sanding the excess epoxy and cutting the part to shape. I've only put one layer of carbon, but the thing is way, way tougher already than it was with the two layers of glassfiber. I can probably stand up on the part and I'm fairly certain my weight wouldn't crush it. Carbon made a very noticeable difference in terms of strenght, I won't need to add any more layers.

View attachment 1

Next steps will be to finish sanding, to add more coats of epoxy, sanding again, etc, then once the finish will be smooth enough, add some epoxy clear coat.

This was just a small project in order to make sure I grasp the basics, before attacking the serious stuff: the battery box itself.
For the battery box, I'll get all the proper gear: right vacuum bags, peel ply, bleeder, etc. That was just a quick and dirty attempt (but I think it worked ok anyways).

Here is the little enclosure next to the battery box, so you can get an idea of the sizes:
IMG_5932.JPG

See you soon, hopefully with good updates!
 
I'm impressed (which isn't easy)
I'm looking forward to your results with the proper kit - if the DIY attempt is anything to go by, it's going to be a work of art!
 
The build is going slow, but I've started the carbon fiber battery box.

The little part is done and now looks ok. I've managed to remove all the pinholes and bubbles after a new layer of epoxy, a bit of sanding and an other layer of epoxy.
I didn't want a glossy finish, and I found a technique to get a mate finish totally by accident: I covered the part with epoxy, but at some point my brush failed and lost a lot of hairs... So I wiped the epoxy away using paper towels. I ended up with a nice and smooth mate finish which I kept this way because I found it to look quite nice. I'll probably do the same for the battery box too.

It's not perfect but for a first attempt I think it is a decent result:
IMG_5951.JPG

Now I'm starting to put the first layer of carbon fiber on the battery box.
I found a stickier epoxy, which allowed me to avoid vacuum bagging, at least for the first layer. I'll probably vacuum bag the other layers in order to get a better bond, but I'm not entirely sure that will be necessary. For now, the first layer is almost completed and it looks already quite nice:
View attachment 3
IMG_5955.JPG
IMG_5959.JPG

Hopefully this should take a few more weeks to complete.
 
Back
Top