Ninja 250 "Clone" Made in China

[peadar]

No links ATM, it was years ago, but I would guess that terms like "DIY LED Headlight" would probably find some of them. I don't recall all the term sets I used back then. I probably saved some of the project pages but that was on an older computer....

there was a tread i saw last week maybe, in the pushbike section ,he used 100w modules for offroad use,he used pwm control

 

[Dui ni shuo de dui]

Hi guys, it's been a while!
Everything still runs fine, the bike is doing well and I haven't done much work on it until a few days ago.

A few months ago, a good friend of mine gave me his old 6000W motor. He told me that the motor was doing some weird noise and had sudden power losses. He gave up on it and bought a more powerful motor instead.

I tried installing it on my bike last week to see.





At first, everything seemed to work very well. The bike was a lot more snappy than with the 3000W so I was super enthusiastic. This lasted for about a day.
Then at some point during a ride, while applying full throttle, I suddently lost 80% of the power and the motor started making a very stange noise. Sounded and felt just like the tire was slipping on the ground whenever I reached a certain amount of torque. It was weird because the motor behaved just normally for the first 1/3rd of the throttle, then it was suddently losing power and making this noise. Kinda feeling like there was some clutch slippage.

Anyway, I took the motor out of the bike, put it on the bench and started investigating.
At first everything looked completely normal, no burnt phases, hall sensors seemed fine, no corrosion, main phases cables wires seemed to be in good shape... Then I started looking at the rotor.
I was not expecting that, but it turns out that the issue was that the magnet glue had given up, all the magnets were free to move!!

The magnets are firmly stuck to the rotor back iron because of their magnetic strength, so they don't fly away, but when applying too much throttle it seems like this force is overcome by the electromagnets and the whole magnet ring starts turning inside the rotor!
I've heard of such stories before but that's the first time I saw it in real life, very interesting!

Anyway, I was super happy to see this, it meant that it would likely be an easy fix!
So first, I decided to remove all the magnets, the plan was to clean everything, apply glue and put it back together.



So I removed all the magnets... But I didn't really think things through and completely forgot to record their polarities, I realized it right after removing them all... Kinda started panicking at this point...
Then I really wasn't sure how to put them back, couldn't find any clear answer on Google so I opened some other motors to see if there was a pattern. It seems like all of them were simply North - South -North -South and so on. Also I noticed that there was no drama at all while removing the magnets, they always kinda stick with their neighbor which was a good indication that they were opposing polarities.
It all seemed to make sense, but I really didn't want to screw up everything, so I thought I'd reinstall all the magnets without glue, put the motor back and see if everything was normal with this magnet arrangement before comitting to glue them.

So that's what I did and it turned out everything worked fine. Great!

Next, I removed the motor from the bike, opened it again and removed all the magnets hopefully for the last time. I then cleaned everything with acetone and glued the magnets back in place using 2 part epoxy.



I also took this opportunity to run 4 more wires through the axle, in order to install temperature sensors, as the motor didn't feature any in the first place.




After that I put the motor back together, installed it on the bike and went for a ride.

All good, it's working great! This motor feels a lot stronger than my old 3000W, the bike feels a whole lot lighter. So far the top speed seemed to be around 140km/h, but I'm certain it can go higher, I haven't adjusted any setting, it is still running with the same parameters I was using with the 3kw motor (550 battery amps and 810 phase amps).
The best surprize so far is that it runs very cool. The 3000W motor often reached temperatures around 120 degree C, but on this motor I could barely reach 70C while riding even faster. That means there should be plenty of room for tuning, that's great!

Now I'll leave the motor like that for a few weeks/months in order to monitor how things are going on, then later I'll surely crank up the power in the controller settings. The 3000W motor was able to sustain more than 22KW, so hopefully a 6000W motor should be able to take 44KW... Hehe I don't think it will, but surely it can take a few more KW right now.

Anyway, that was an interesting problem with a very rewarding outcome, now let's hope this motor will have a long and healthy life!

 

[harrisonpatm]

So, now that you happen to have a perfectly good 3000w motor laying around, what are you making next?!?

 

[dominik h]

Very interesting Problem, and I hope your repair lasts for a long time.

I hope my 10kw motor also runs a lot cooler as my previous installed 5kw motor. The 5kw often reached 140°C, with only 630pA anps and 300A from the batery. I will also test with the same settings to see how the new motor reacts and make a few acceleration runs before I upgrade to a bigger controller.

 

[TorgueRPM]

Some wider clamping torque arms might be a good idea now that you've moved up to a motor with more torque.

 

[dominik h]

I had to make bigger ones, my new motor has a bigger axle. The old one was 20x14 the new one is 24x18. I also had to widen the swingarm.
The new 10kw motor does behave nearly the same as the old 5kW motor, but it runs a lot cooler.
My Fardriver ND72680 seems to be to weak for the bigger motor.
I will do a few acceleration runs, and then I will change to the ND721800.

 

[dominik h]

Did a test ride with the 10kW 60H 273 Hub Motor in my scooter.
With the same settings 300bA 630pA with the Fardriver ND72680 there is no big difference in acceleration. 0-80km/h 0,3 seconds faster.
With the nearly maxed out 72680 (340bA) it is 0,8 Seconds faster to 80km/h and 2,5 seconds faster to 100, now it takes 9,7 seconds to reach 100km/h.
After 10 acceleration runs it was only 55C , before i hit 140C after 10 runs.

Now it is time to change to the ND721800 and raise phase current to 1000A and probably above and see if the starting tourque improves.
If the installed JK-B2A24S20P BMS survives a 30 second 500A current test with an electronic load I will raise battery amps also.
I will do this test outside of the scooter with an electronic load. I don´t won´t to risk a fire inside my scooter, while I´m driving.

 

[TorgueRPM]

You may want to make your own thread, then you can add pictures and other good stuff without cluttering this one

 

[Dui ni shuo de dui]

So, now that you happen to have a perfectly good 3000w motor laying around, what are you making next?!?

Hehe Good question, I have no plan so far but it would be nice to make a fun project out of it.

One thing I'd like to try someday is to make some kind of modification in order to fit a different rim to these kind of motors. The QS rims I have are too narrow and are only 16 inches, it makes finding compatible tires very difficult as the only choices I have are either Michelin or Metzeller and there's only one reference existing in their linups as far as I know.
Would be nice if I could fit a custom rim, I'd have more size choices and could fit a larger tire, which would be very beneficial to the bike cornering stability (at least I think so). Would also definitely look a lot better, I think the narrow rear rim and its skinny tire is the only real bad looking spot on the bike, it would clearly look better with a proper wide tire.

Maybe this motor could be a good guinea pig for such experiment

Some wider clamping torque arms might be a good idea now that you've moved up to a motor with more torque.

Yes indeed it wouldn't be a bad Idea. Especially since the 6KW axle is a lot smaller in diameter than the 3KW motor, for some reason I cannot understand.
I've designed a torque arm for a friend a few years ago, it worked well on his 8KW motor so I'll probably just make another.

With the nearly maxed out 72680 (340bA) it is 0,8 Seconds faster to 80km/h and 2,5 seconds faster to 100, now it takes 9,7 seconds to reach 100km/h.

That seems rather slow for such power settings on a scooter, unless maybe its very heavy?
I recall my scooter was doing 9 secs and change with the 150bA /350pA Sabvoton controller and a 2000W motor (85kg scooter+85kg rider)

then you can add pictures and other good stuff

I second this, would be cool to see what your scooter looks like and follow its build

 

[dominik h]

That seems rather slow for such power settings on a scooter, unless maybe its very heavy?
..... (85kg scooter+85kg rider)

my battery weights over 70kg , the scooter weights 183kg. With the 10kw motor and the bigger controller the scooter will be over 190kg.
With the SVMC72150 it took 12 seconds to 80km/h

Thread is created

 

[Dui ni shuo de dui]

Small update, I fitted a new rear tire to replace the ugly skinny one that came with the 6KW motor.
Unfortunately it is VERY difficult to find anything good on 16 inches rims, so I kinda had to go for a no name tire. So far it doesn't seem too bad, but time will tell.

At least it finally looks more like a sports bike tire. I wish it could be a tad wider, but that's the maximum you can fit on a 3.5inches wide rim. If only QS made the rim removable, it would be a lot better.

Had a really hard time to seat this one, not sure why, but I hope I won't have to take it off anytime soon...
Anyway, I quite like how it looks now. Also it seems to make the bike a bit more stable in corners so that's great.

 

[rafalg]

Just an awesome story, starting with a guy skimping on battery and unfolding into an engineering feat. You're dangerous, sir!

 

[Eastwood]

Yeah the new rear tire looks great! Had to go back and look at your old tire, what a difference

 

[amberwolf]

Unfortunately it is VERY difficult to find anything good on 16 inches rims, so I kinda had to go for a no name tire

Do you mean just to find where you can get them, or do you mean finding them at all?

I know Pirelli makes 16" tires for this width (at one time I had some discarded ones that were still better than anything I had ever had at the time; just didn't have wheels I could put them on yet, but before I could get said wheels, I had to get rid of them years ago; wish I still had them).

This one would probably work; I think it comes in up to a 150-wide for rear, 80 front?

SPORT DEMON™ 150/80 r16 | Pirelli

Discover the range of SPORT DEMON™ motorcycle tires with size 150/80 r16 and choose those that best suit your needs

www.pirelli.com

There's others here

Hypersport motorcycle tires | Pirelli

Discover the range of Pirelli hypersport motorcycle tires and choose those that best suit your needs

www.pirelli.com

that might work.

I think I've seen similar in Dunlop and other brands too.

CST and Shinko probably also make them (these are the two brands of tire I've had the best performance from, and least problems with, though I am not using anything for motorcycle performance levels).

 

[Dui ni shuo de dui]

Do you mean just to find where you can get them, or do you mean finding them at all?

Sorry, I mean it's a bit difficult to find tires of this size in China.

I know Pirelli makes 16" tires for this width (

Thanks, I didn't know they had tires in this size. I just checked but I couldn't find any shop that have them here, unfortunately.

CST and Shinko probably also make them (these are the two brands of tire I've had the best performance from, and least problems with, though I am not using anything for motorcycle performance levels).

Yes, there was one tire available from CST. I basically hesitated between the CST and this tire. The reason I chose to buy this one is because it is actually the same tire as my red motorcycle. So I knew it was a very wide tire (for some reason all 140-7-16 tires aren't at all the same actual width). Well, by "very wide" I mean compared to the Michelin I got before on the 3KW motor. My main goal was to fit the widest tire I could.

The Michelin was okay, but it was a bit too skinny and it doesn't seem to be distributed in China anymore (at least I couldn't find it).
I wanted to try Metzeller because I heard lots of good things about it, but unfortunately it wasn't available as well.

Sometimes no name tires can be a good surprise, for example at the front I have a "Cordial" tire which is probably as old as the bike, but I don't really feel the need to change it since it always seem to do a very good job.
Thanks for the suggestions, I appreciate it

 

[amberwolf]

Sorry, I mean it's a bit difficult to find tires of this size in China.

I was pretty sure that's what you meant.

Thanks, I didn't know they had tires in this size. I just checked but I couldn't find any shop that have them here, unfortunately.

Yeah, if they hadn't cost more than twice what the Shinkos I'm using on SB Cruiser do, I would have gone with the Pirelli instead. (I think I was actually looking at them back when I was still using CrazyBike2). But...budget.

Sometimes no name tires can be a good surprise, for example at the front I have a "Cordial" tire which is probably as old as the bike, but I don't really feel the need to change it since it always seem to do a very good job.

Yeah, the first CST tire I ever had was on a random thrift store bike, had never heard of them before, but loved them instantly the first ride on them, and after a while I began looking for them specifically as they were much better than the other bicycle tires I'd ever found (and these days I try to be sure the tubes I use are CST, because they're also better than any other ones I've used).

 

[Dui ni shuo de dui]

While I'm there, here are a few updates about the recent light improvements I've made.

A few months ago, I built a new front light system test prototype for the bike using a big 100W LED chip. (See here if you don't know what I'm talking about).

I tested it over the past 5-6 months and everything is working well. The only remaining issue was the materials I used for it, since it was just a simple prototype I used PLA plastic, which usually starts deforming at about 50 degree C.
The LED does generate a lot of heat. The headsink and copper plate did a great job at pulling away most of it, but the plastic did deform over time even though I limited the LED power to only 30W.
So anyway, the plan has always been to replace them with metal versions at some point.

In order to do that, I'm using a somewhat unusual method: 3D printing + metal plating.

The process it the following:
-3D print the object. Here I use ASA plastic, which can handle high heat without deforming. I successfully used ASA to print components for my Honda motorcycle exhaust and the part survived very well so I know this plastic is pretty awesome. Also it doesn't degrade over time, it is intended for outside use.

-Step 2 is to paint the plastic part with a conductive paint. The paint I used contains very small copper flakes and its conductivity is excellent. This is necessary for the next step.

-Step 3 is to electroplate the part. Basically, you plunge the part in an acid solution, with a pure copper electrode on the positive side (cathode) and the part you just painted on the other side, acting like the anode. Then turn up the current (how much current depends on the parts surface area) then wait. The longer you wait, the thicker the copper layer will be.

There is a small caveat, which is that the longer it's being plated, the more irregularities and surface defects you're likely to get. So if you want a perfectly smooth result then you might have to build thickness in several passes and sand in between.
In my case, I've built a layer of copper of about 1mm all over the part in one single pass.

-Once you're satisfied with the copper plating results, you can proceed to the last step which in my case was nickel plating. The goal here is to protect the copper from oxydation. Here you don't build thickness, you just plate for a few minutes until the copper is protected (nickel electrodes are expensive)

I forgot to take pictures of the process, but here was the the result:



The part on the left has gone through all the steps so that's the final result.
The one on the right went though the copper plating and final sanding and was being prepped for the final nickel plating.
Finishing only depends on how much prep and sanding you're willing to do. For these parts I didn't need good appearance so I did the very bare minimum, but it's possible to get stunning results if you're willing to put some time in it.

The benefits of this method are the following:
-The parts are mechanically stronger. Obviously, the thicker the layer, the stronger it will be.
-The part feels and weights like an actual metal part. If done properly no one could guess it's been 3D printed in the first place. It is actually real, solid metal, only the core remains plastic.
-It can widstand a lot of heat. I once made a small test part, heat it until it was glowing red and it didn't deform or get damaged at all. The plastic inside might melt, but the outside metal shell won't
-It's possible to make electrically conductive paths, you can mask some zones and make copper traces if you need to give current to some device without using wires
-you're using chemicals of various colors (deep green and deep blue) which makes you look like a mad scientist.

The caveats:
-It's a bit work intensive if you need a nice finish.
-The process is intimidating and seems rather complicated the first time
-Copper and especially nickel electrodes can be expensive
-The parts are significantly heavier than regular plastic ones
-Electrodeposition is sometimes tricky, the metal can stick to some areas more than others

For example, I got a big defect on one of the parts, as you can see on this picture:



The copper plating on this part didn't go well in this area (this was my fault, I made a mistake during prep), so it ended up with lots of porosity and some of the copper didn't stick.
Here it really didn't matter for this particular application, so I didn't bother to fix it. I think this picture is interesting because it shows well the kind of thickness you can get from such a process. This was obtained after about 10 hours of plating (I left it overnight, basically). The other sides of this part are nice and smooth.

So, next I had to make a plate to hold the LED chip, this was made with an actual pure copper plate (4mm thick):



Finally, I assembled the parts together:











The only visible part of this assembly is at the front, so I only tried to get a good finish there.

I then installed this into the bike together with the old prototype plastic one (I'm curently making a second unit to replace it asap).
Now that it is metal plated I can feed it with more power. So far it is set up to 50W per light and the results are impressive, it is extremely bright. It does heat a bit but so far it seems to be ok. I forgot to tke pictures, but there is a small fan blowing air onto the heatsink fins whenever the light is switched on.

Next step is to make the metal parts for the second light, then I need to find a good way to modulate the power. I'd like to use the starter lever to be able to set the light intensity, this way I can adapt it to the riding conditions and traffic anyway I want.

 

[MorbidlyObeseKoala]

In order to do that, I'm using a somewhat unusual method: 3D printing + metal plating.

Really interesting and incredible results as usual from your workshop. Is there a video describing the process you used to do some trial runs? I don't think I have seen an amateur workshop produce a part like this before (havent been searching for it but impressive to be sure).

Obviously the first few attempts take longer while working out kinks in the process but how long did it take you to go from printed plastic to a strengthened metal shelled part? Excluding final finishing for making it pretty.

 

[Dui ni shuo de dui]

Really interesting and incredible results as usual from your workshop. Is there a video describing the process you used to do some trial runs?

Thanks!
Sorry I don't really have the ability to take videos unfortunately, but some great videos are available on Youtube that describe the process better than I could ever do it:
I saw this video a few years ago and this is what convinced me to give it a try:

Obviously the first few attempts take longer while working out kinks in the process but how long did it take you to go from printed plastic to a strengthened metal shelled part? Excluding final finishing for making it pretty.

Actually it wasn't the first time I tried, I already made many, many parts using this process. It's only the first time I make a part like this on my electric motorcycles though.
For example, I built this remote controlled tractor for my little nephew, all the shiny/gray parts were made using this very process:




It's actually not very time consuming, I can make a part from start printing it to a finished product in about 24 hours. A few minute/hours to print the part, maybe an hour of sanding or more depending on the desired finish, then usually 8-10 hours of copper plating, another hour of sanding/polishing, then 30 minutes to an hour of nickel plating and a final brushing/polishing. There are a few steps of cleaning with different chemicals in between the steps, but they don't take much time.

It's a bit work intensive because of the sanding, but aside from that it's mostly just waiting.

 

[Dui ni shuo de dui]

No special work done on the bike as it is pretty much a finished project now (I still have to make the variable power supply for the front light but aside from that it's 100% finished).

So it was time to tune up a little bit the controller since now I have the 6KW motor instead of the old 3KW one.
I tweaked the max phase current up to 880 amps, 550 battery amps, and changed a few settings so it has more punch at startup. I also increased the top speed but I still don't know by how much, I need to find a longer road to test it.
The motor heats a little bit more but nothing very alarming, it's taking the power like a champ so far!

Spent a wonderful afternoon riding the bike in the countryside nearby since weather finally warmed up a bit, it was a nice opportunity to take a few pics of the bike outside the shop

 

[harrisonpatm]

You are using ANT bms, yes? I'm wondering how reliable it's been for you. I need to pick out a 28s bms for my conversion

 

[dominik h]

40-50kW peaks from a 6kW QS273 motor , very impressive.
My 5kW QS260 40h motor was maxed out with 630pA and 300bA @77V in my heavy scooter.
In the summer I often got to 140C winding temperature.
The now built in QS72V10000W 273 60h motor stays cold with 30kw @1400pA 400bA.

 

[Dui ni shuo de dui]

You are using ANT bms, yes? I'm wondering how reliable it's been for you. I need to pick out a 28s bms for my conversion

Yes, I'm using the ANT BMS for a few years now.
It is important to note that I'm using the very old versions, I can't say for sure if the new ones are still as good now.

I have installed 4 units on my own vehicles so far, and a few ones on friends rides.
They are generally reliable, but I had a few who did die at some point. One had a problem with the bluetooth module, it couldn't connect anymore. The one you see on my bike did also die a few months ago, I don't know why. The bike was charging overnight and when I came in the morning the BMS was bipping continuously with a "Discharge Mosfet error" message. Turns out the discharge mosfets died, for no apparent reason. I ordered new ones replaced them, which solved the issue for a week, then it happened again. So I replaced the whole mosfet board with another one I had around (the one from the BMS with the bad bluetooth) and so far so good. I'm still not sure what happened.

It's not completely bulletproof as you can see, but so far the failure modes seemed to be good. It's not unacceptable that a cheap BMS fails, as long as it fails in a safe manner and alerts you. So far that's what happened so I can still say it's a pretty good BMS that does its job.

40-50kW peaks from a 6kW QS273 motor , very impressive.

To be fair, the actual power feels closer to low 30KW, I haven't checked but I don't think the battery discharge actually reaches 550Amps just yet. It's the limit set in the controller, but I think there's a bottleneck somewhere else. It's probably closer to 400A. I'm not sure yet, I'll have to check. It's a bit complicated to explain, but I made a shunt mode experiment on my BMS so that makes my dashboard readings innacurate. I need to fix that before I can get real readings.

The bike is very lightweight (130 kg), so I don't need too much power to accelerate it.

My 5kW QS260 40h motor was maxed out with 630pA and 300bA @77V in my heavy scooter.
In the summer I often got to 140C winding temperature.

Yeah that's way too hot indeed. I would be afraid to ruin the magnets at these temps. Max I did was 120C, but that was with statorade.

The now built in QS72V10000W 273 60h motor stays cold with 30kw @1400pA 400bA.

Awesome, that must be stupid fast. I think you're limited by your max battery current though.
Be sure to upgrade the brakes, frame, tires, etc.

I wish I could install a better motor. Ideally I'd fit a 17inches QS273 10KW , with 6 inches wide rim and put a big tire. The battery wouldn't be able to keep up, but it would look great and the motor wouldn't be stressed.
Maybe someday if I get the opportunity.

 

[dominik h]

Awesome, that must be stupid fast. I think you're limited by your max battery current though.

The controller drwas 400A from 30-100km/h, above I reduce the current in the controller settings. Topspeed is limited to 120km/h.

My battery can handle 800A continious.
I'm limited by the choosen BMS, the brakes and the weak scooter frame.

 

[harrisonpatm]

Yes, I'm using the ANT BMS for a few years now.
It is important to note that I'm using the very old versions, I can't say for sure if the new ones are still as good now.

I have installed 4 units on my own vehicles so far, and a few ones on friends rides.
They are generally reliable, but I had a few who did die at some point. One had a problem with the bluetooth module, it couldn't connect anymore. The one you see on my bike did also die a few months ago, I don't know why. The bike was charging overnight and when I came in the morning the BMS was bipping continuously with a "Discharge Mosfet error" message. Turns out the discharge mosfets died, for no apparent reason. I ordered new ones replaced them, which solved the issue for a week, then it happened again. So I replaced the whole mosfet board with another one I had around (the one from the BMS with the bad bluetooth) and so far so good. I'm still not sure what happened.

It's not completely bulletproof as you can see, but so far the failure modes seemed to be good. It's not unacceptable that a cheap BMS fails, as long as it fails in a safe manner and alerts you. So far that's what happened so I can still say it's a pretty good BMS that does its job.

Thanks for your info. I will be using contactor-controlled BMS (bypassing the BMS for discharge, but not for charge), so basically I just need something that has completely reliable LVC and HVC. Maybe balance too (it balances, right? As far as I can tell from Aliexpress page descriptions I am able to get some ANT models with active balance). I won't be using SOC, display, discharge current, ect.