Yamaha YZ250F

j bjork

100 kW
Aug 31, 2018
Linköping, Sweden
A bunch of money and:


Resulted in this:



2018, It is supposed to have run 105 hours.
There is a little problem though:


Woopsie poopsie! I guess it needs a new engine, huh? :wink: :wink:

Seriously though holy crap wtf happened to that engine ??

Going to be a fun build! Awesome bike to start with
Well, the connecting rod broke. I dont really know why so far.

But first things first, checking weights:



49kg both front and rear.
This is dry weght, no oil, no coolant and no fuel. Well, not even a chain.


"Naked" motor 24kg.
(Complete, but no fuel injection or exhaust)
I was hoping it would have been heavier :?
I want to do a 2 stage reduction on this bike, but I think I will have to do 3 stage instead:


With the output shaft in the standard position there is not much room for a bigger sprocket on the other side, if I just used a shaft that goes from one side to the other.


But if I use the transmission from the bike I can use a sprocket instead of the clutch.
Then it is a different story.
A problem is I dont think there is enough room for the motor in this position, I will have to move the input shaft of the transmission down. That will also mean less room for a big sprocket, but hopefully not too much.


A little left over engine parts :wink:


3:d gear has taken some beating.


One of the shifters too.


I measured runout of the shafts, it is a little high.
I get 0,2mm on the input shaft, and 0,28mm on the output shaft.
But that is from end to end, between bearings it is less.

Third and 4:t gear is damaged, but 1:st, 2:eek:nd and 5:th seems usable.
2:eek:nd looks strongest.

Final gear (between output shaft and wheel) is 1:3,846 (13:50)

1:st 2,14 (14:30) total 8,23

2:eek:nd 1,75 (16:28) total 6,7305

3:d 1,444 (18:26) total 5,55

4:th 1,222 (18:22) total 4,699

5:th 1,04 (24:25) total 3,999

So I think 2:eek:nd may be a good choice, with probably a little reduction in the first stage.
If you could make the box use only 2nd and 5th I think you'd have a perfect combo. Basically deep reduction and direct.
I honestly don't think a 3 stage is necessary. You will just be faced with an overly complicated situation and unnecessary losses.

Due to the lack of the chain guide on the other side I think you might be able to go as high as 15T safely. You can also get a 520 sprocket for the QS motors as low as 10T from AMP sprockets so that's a 1.5 reduction alone.

Keeping the box is just a massive waste of space that could be used to maximise battery space.

1.5x multiplied by whatever normal reduction you keep on the final drive side like 13/48 stock should be more than enough.

What's your target top speed? Would help find a good solution for you.

I would keep the final drive stock 13/48 so you have room in the future to either achieve a greater top speed or better acceleration by just using original sprockets. Could go from 12-14T on the front and 47-52T on the rear in future.

So keeping 13/48 in mind, using the tire circumference and desired top speed, motor speed (depending on voltage/potential field weakening) you can decide on the primary reduction
Sure, 2 stage would be a lot easier.
But I think you are wrong about being able to have a 15toot sprocket on the other side.
There are no chain guide as you say, but instead the chain has to fit between the sprocket and swing arm. And the chain is thicker than the guide.

The transmission wont take up any battery space if it works as I have planned.


The output shaft has to be there anyway, and the input shaft will be in maybe 45 degrees angle down/in front of the output shaft.
The gears on the output shaft are about the same size as the chain sprocket, so I would need some clearance there anyway.


Something like this. The white cage symbolize the motor, but is a little smaller. So it will be tighter..

Even if I was able to run 10/15tooth in the primary, I dont think that is enough. And I am trying to avoid small sprockets, I would like to be in the 20-30:s range, or at least around 15 on the motor.

I am not sure what gearing I will end up with, but on the other bike with a qs 138 70h about 8:1 seems about right in total. That is with a kv of about 60.
According to qs the 180 90h should have a kv of about 55, but with more power maybe I will go for a taller gearing if I get more low end than I can handle anyway. I am not that interested in top speed, around 100kph is enough. It is not intended for the street, I want lots of power at the speeds I am riding.

I dont think it is the same shaft on this motor as the 138 70h, qs called that spline M14 and this motor should have M20.
I havent found anything on what that means, so if someone knows I am interested :wink:

I have a cad drawing, but I dont know cad.. I think I was able to get some dimensions of the motor at least:

You have made some fair points, here's some info that might help in your decision.

If my memory serves me correct the person with the 8:1 on the 138 is running a smaller wheel than a dirt bike, dirt bikes have one of the largest tyre circumferences for a bike. There's a dirt bike here on Endless that has a top speed of 95km/h achieved with a 7.28 reduction.

Do you have any information of the QS 138 90H such as true load max rpm as that's very important to know before you even decide on ratios.

For example a 75-7R motor can do 6000rpm peak but 5000rpm sustained so I will use that figure instead. My dirt bike has a rear tyre of 110/90/19 (CRF) so that's a circumference of 2.14m
My target top speed is 110km/h using sprockets of 13/53T to match its original specs and so that I can adjust up or down in future.

That circumference and top speed means I need 856 rpm at the wheel which is from (110 * 1000 / 60) / 2.14
The stock sprockets have a reduction of 53 / 13 = 3.92
To achieve the 856rpm I need a reduction total of 5000/856 = 5.84
Primary reduction would need to be 5.84 / 3.92 = 1.5
Though 10T front and 15T rear on the primary is pushing the chain limit (& noise) as well as space to maintain original sprocket location for me, a belt for the primary reduction or gears soaked in oil would be a better solution for me.

But as I said earlier, the motor rpm matters alot, I tried to do a search for the load rpm of the 90H but couldn't find a reliable figure, I could calculate it backwards from the successful dirt bike builds using the 90H that are only 2 stage reduction.
This is just planning, I might very well end up just making a new shaft and moving it forward a bit from where the output shaft is now. Then I will get some extra clearance for a not too small sprocket on the other side, and settle for a 2 stage reduction with relatively small sprockets on the first stage. One option might be to run a 219 chain on the first stage.

I am a little hesitant to belt drive, I think it has to be rather wide to take the torque. And space is not easy to find.
It is nice if it isn't too noisy, but not first priority on a dirt bike.
I can also run 428 to the wheel, to be able to get a little extra reduction there without having to go crazy big on the rear sprocket.

On the ktm I run 8.18:1 with the qs 138 70h and have a top speed of at least 120km/h, I havent really tested top speed. But I have good acceleration to over 110km/h at least. That is with over 7000 rpm though if I remember right.

The motor I am looking at is qs 180 90h, not 138 90h. There is very little known data I think, I dont think anyone on this forum has one yet. I think there are some data in the qs sales thread, but I think it is hard to make much sense of their numbers. But I will see what i can find.
QSMOTOR 180 9000W 90H PMSM IPM mid drive motor

For a long time, we are developing and testing our new 8000w mid drive motor, below is the result

For Motor
Item: QSMOTOR 180 9000W 90H PMSM IPM mid drivemotor
·Motor Type: PMSM IPM Motor With Hall Sensor
·Rated Power: 9000W (18000kW peak in few seconds)
·Rated Voltage: 72V
·Speed: 4000RPM peak (basic RPM), 7000RPM with flux weakening
·Max Torque: approx 100N.m peak.
·Max Efficiency: approx 95%
·Continuous Battery Current:200 A (72V)
·Peak Battery Current: 250A (72V)
·Working Temperature: 70-130℃ , peak 180 ℃
·Single Hall Set with waterproof connectors
·Waterproof Grade: IP67
·Cooling: air cool
·Dimension: φ200X280mm
·Phase wire: 22mm²
·Insulation grade: H
·Working system: S9
·Color: Black
·N.W./G.W.: approx 25kg
·Package Dimension: pending piece
PS. For 180 PMSM motor, it’s maximum 135H 12000W motor.

The current product shell is a more industrial design, need to update the picture
Below are pictures for previous sample 180 135H mid drive motor, we plan to do some appearance updates for final sample, i will show you pictures when the sample finished.

1.jpg (16.21 KiB) Viewed 4782 times

Here is drawing for whole 180 mid drive motor, if you need high-definition picture, please kindly ask me.
Please check the installation dimension, especially the diameter and width for threaded holes.


2.Match with sprocket SN520 13teeth, plan gear ratio 1:6.5
520-13 teeth

5.jpg (31.02 KiB) Viewed 4782 times

Test report
180 90H MAP


180 75H 7500W 99N.m



180 135H 12kw
with EM300 72V 350A, approx 102N.m peak


with EM300 82V 350A, approx 100N.m peak, 24kw peak


12.jpg (50.45 KiB) Viewed 4782 times

Heavy Electric Motorcycle, e.g. Harly Scooter
E-motorcycle weight, approx 180kg+2*80kg
Speed: continuous 100km/h, peak 120km/h
Gear ratio: 1:6.5
Climbing slope: 50°@260kg
Battery: 72V80Ah
Tire: 0.6138 130/70-17

15.jpg (36.07 KiB) Viewed 4782 times



For Controller
We suggest to choose sinusoidal wave Motor Controller EM72200SP with standards CANBUS,
EM72300P should also work (without CANBUS).
functions available for speed modes, anti-theft, cruise, reverse etc.
refer to below hyperlink.
http://www.cnqsmotor.com/en/article_rea ... s/837.html

17.jpg (68.9 KiB) Viewed 4782 times

18.jpg (69.91 KiB) Viewed 4782 times

I tried to copy and paste the info in the qs sales thread, but the pictures were lost.
You can find it here: https://endless-sphere.com/forums/viewtopic.php?f=31&t=65972&start=1200

Note that it is 3 motors they are mixing the specs of.

This is the "base" batteries I plan to use. Then it will depend on the controller, but it would be nice if I could use 24s (2x12s modules)
These are hard to reconfigure, the terminals are welded on the cells. And the cages around the cells are welded together too..
The cells are about 30Ah I think, I have done some tests but I dont remember exactly.
I think I will beef it up a bit with 5-10Ah of high c-rate lipos in paralell to get some more juice and as low ir as possible.

The plan is to have the battery removable so I can have 2-3 and change between rides.

It looks like they will fit, but it is pretty close on the rear one. And they will probably end up a little higher than they are now, but hopefully not much.
j bjork said:
QSMOTOR 180 9000W 90H PMSM IPM mid drive motor

Is this really for sale now? Any idea on the price?

They have been speaking about it for a while but nothing.

By common logic this should be able to give twice the power compared to a 138 70H
c70r said:
By common logic this should be able to give twice the power compared to a 138 70H

How can you tell, rotor surface area?

It would be really interesting to know how it compares to me1507 and zero 75-7.
I havent seen numbers on the rotor size on those, but it seems the 7 in 75-7 stands for 70h ?

The qs seems to be very close to 200mm in diameter with the 180mm rotor, and the me1507 220mm without fins. So I guess the rotor can probably be 190-200mm dia.

Qs claims to sell, or be ready to sell the 180 90h. They said they would make some next month, I dont know if is the first production run.
How can you tell, rotor surface area?

Same max rpms given from QS for both motors
Power proportional to rpms*torque
-> only torque diff needs to be found

Torque proportional to airgap surface * lever given same flux and materials ->

But a lot of assumptions need to be correct for this to be close to the truth. I guess it can be a bit higher since longer motors have relatively less ”ends”
Looks like an better Basic then the KTM, because of the oval Frame design and space for Battery. :bigthumb:

This thread might be something for your plan using the gearbox https://endless-sphere.com/forums/viewtopic.php?f=12&t=65731

ALIEN POWER SYSTEMS https://alienpowersystem.com/
could be worth a look too, they produce a high amount of differnt motors that could fit into your Gearbox.
They do custom motors aswell.
Thanks larsb, makes sense :thumb:
So it seems like the 138 90h should be able to make about 30% more torque and power than the 130 70h, interesting.

Thanks Soap_62, but I am not looking to find a motor that fits inside the housing.
I want a big motor that can make big power for some time.
I prefer an inrunner that is watertight, and I have decided to go for the qs 180 90h and made a partial payment for it now.
I will have to wait though, they will make it after the chinese new year.

So now I will try to find someone to make a 3d printed model of it :wink:
I see.

I cant help you with printing but with CAD if you need something.

Wich kind of battery are you going to use on this one? Thats something iam struggling right now with my build...
j bjork said:
Thanks larsb, makes sense :thumb:
So it seems like the 138 90h should be able to make about 30% more torque and power than the 130 70h, interesting.

Thanks Soap_62, but I am not looking to find a motor that fits inside the housing.
I want a big motor that can make big power for some time.
I prefer an inrunner that is watertight, and I have decided to go for the qs 180 90h and made a partial payment for it now.
I will have to wait though, they will make it after the chinese new year.

So now I will try to find someone to make a 3d printed model of it :wink:

I can probably help you with a 3D printed model.
Vart i Sverige befinner du dig?
Haggan, du har pm.

Soap, I will use the EV modules that you can see some of in a picture when I tried if they fit.
I will probably parallel with some rc lipos or something to get some extra Ah and lower ir.

The ev cells are very good, but not the most energy dense. So they make more sense on a daily driver.
But I was able to get my hands on those, so I will use them.
This will be great. looks like these guys did a very similar build, maybe even with the same cells? Honda?

Similar bike, but I will have 2 or 3 stage reduction so I get the sprocket in a better position and get better gearing.
I will not have the battery modules open like that, I will build some kind of case for them.
I dont want my batteries wet and muddy..

My modules are from vw passat gte.
This man has interesting way to use the trans for gear reduction https://altaownersforum.com/threads/a-little-bit-about-my-2008-yz250f-electric-conversion.2055/#post-26642
It looked a little interesting, but it was hard to see much of it on these small pictures.
(You had to be logged in to get bigger pictures)

I wonder if it is better than just using 2 sprockets and a chain, I guess I will have to do some calculations to find out.
I guess the final gear can be a little bigger than a sprocket, when it dosent have to be room for a chain around it.
But a stable mount for the middle gear will be critical. And not that easy to do, if there isn't room for a shaft.
One advantage with gears instead of a chain is that I dont need a tensioner or a way to adjust chain tension.

At the moment I am leaning towards just using a 2 stage reduction with chains and sprockets, I think that if I move the output shaft 10-15mm forward I will be able to fit a 22tooth 428 sprocket on the primary side. That way I could get 2:1 gearing in the primary if I run 11tooth on the motor.
With the standard 13/50 (3,846:1) to the wheel I would get a total of 7,692:1, and I dont think I would have any reason to go any lower than that. To go higher I could use 12 or 13tooth on the motor and get 7,05:1 or 6,5:1 maybe even 14 tooth and get 6:1.

Qs suggest that I should use the APT 96800 controller. I have asked for info about, and if they sell the 961000. But they dont seem to want to answer that question.
Anyway the the 96800 is already a beast (when it comes to size at least)

I made a cad model of it (cardboard aided design) and tried on the bike:


It wont work like this..


I cant even get it up under the seat..


Back where it started, it might be possible with some cutting in the fender and careful positioning to put it where it is not in the way of the tire or rider. But it will be tight.

I am also considering splitting the motor phases and run dual controllers, then I would be able to have controllers that fit where the radiators sat. Good cooling, not messing up the design..