Ninja 250 conversion: Need some direction

[tostino]

Well i've been kinda away from the forum for the past few months...

Back in October I got hit by a car while riding my bike home from work, my bike has been out of commission since. Still dealing with insurance from that, and I need to get surgery on my knee from that accident.
That kinda took the steam out of me when I was trying to get started on a new project... I want to build a reasonable performance electric motorcycle with a very good range, and nice top speed.
The basics:
I managed to snag a nearly perfect Ninja 250 for $800.

I have mostly built 2x 12 fet 4115 controllers ready for 150v, and have all the parts for 2x 36 fet controllers with 4115's set up for 150v max operation.

I was wanting to do a 36s lipo battery pack using nanotech's, and batch charging with meanwells.

I was originally thinking that I should use two hub motors, lace them into wheels, out them on the front and rear and hope to hell I can make the brakes work with them :p. I even bought two 9c motors to do this with. That was a bad idea now that I think about it lol.


Here are some key points that I would like to outline, or I need help with:

• I would like to stick with 36s and utilize the 2x 36 fet controllers I will build
• I want to stick with a BLDC motor
• I think a central motor driving through the chain is the best bet, but would like your thoughts/suggestions on that.
• Should I use the transmission, or try and directly attach the motor to the chain?
• What motor would you suggest for me to use? Could I split the windings to effectively use both of the controllers on a single motor?

I would appreciate any advice or guidance you could give, and I would like to say that it's really good to be back on the forum!
-Adam

[Biff]

Good find.

What top speed and acceleration do you want? .. that will help determine an appropriate motor.

Other than that I think you are on the right track with NanoTech batteries, and you seem to have the controller portion under control if you are planing on building your own.

-ryan

[tostino]

I was really wanting a top speed of over 100mph. I don't care for much more than that as far as top speed goes.
Acceleration... It would be nice if it was hard to keep the front wheel on the ground :p. I really would like to at least match the gas version of this bike at the very least.

The controller portion should be handled... They are just Infinion 36 fet boards I got from Keywin, with 4115 fets I got from Methods. I need to beef the traces, put better caps in place, solder the fets, and then do the wiring.
The only bad part is that I need to create a cooling solution for the fets, because the aluminium bars weren't available from him when I bought. It hopefully won't be too bad to create something.
I think hitting 250-300 phase amp peaks should be no problem for each of these controllers, which at my voltage should give me a peak of prolly 30kw for each controller. I don't expect any issues with these quick bursts of high amps being an issue with heat.
See previous thread on the subject: http://endless-sphere.com/forums/viewto ... =2&t=22261
I learned a lot from that.

[Biff]

I have a 2003 ninja 250, and can tell you it goes a little faster than 100mph if you gear it right, but not much. The spec of the engine is 27.9 kW, so without doing any significant aero modifications, you will need to put down about 25kW of power into the ground to get you 100mph.

a 25kW peak motor, should give you plenty of starting torque to get your wheel off the ground, depending on the weight of your battery. If you are using a motor controller like you have described (which I believe don't have field oriented control), The torque the motor can produce will drop off significantly as the speed increases. If you modify the controller to use field oriented control, or use an external timing adjuster (like Burtie has made, which allows you to essentially estimate field oriented control) you might be able to keep the torque pretty high up to high speed.

Now the task becomes finding a brushless dc motor capable of the power you want.

This one might come close with a 36s battery (I am assuming that the battery cell voltage is 3.7V nominal, so 36S would be around 130V)
http://www.motenergy.com/me0913.html
It is only rated for 96Vdc, but maybe you could push it with your 36s controller. It is $850. It is a 2 stator 1 rotor axial flux design, and looking at the picture, it seems you could easily connect one controller to each stator separately with hardly any modification needed.

I think the biggest drawback of that motor is the open-to-air fan cooling, I don't think it would handle water splashing or dirt very well.

I know I have come across more motors that I think might work, but none of them as far as I can remember have a price, or are available.

we'll see what other members think of the Mars brushless that I proposed and see if they have any better alternatives. I imagine some will say, put 2 of the Colossus motors in there. But we are still waiting to see if they will deliver the performance they are hoping for.

-ryan

[tostino]

Does anyone see an issue with using an Infinion controller at that high voltage with that motor? Will I run into E-RPM issues?

I didn't know anyone had axial flux motors available like that! Very cool =).

[Biff]

Here is a thread from another forum discussing the motor, and it has an RPM/ torque chart, I found this thanks to AmpEater in the Zero 2011 discussion on this forum.

http://www.diyelectriccar.com/forums/sh ... 41426.html


You shouldn't run into electrical RPM problems, it is an 8 pole motor, so at 5000 RPM (which is what the other thread says is its maximum) you would only have
5000 / 60 * 4 = 333Hz electrical. The lowest limits I have seen are 667Hz on the Kelly controller.

[tostino]

Hmm sweet!
Do you have any idea what the rpm is limited by? Is it bearings, is it the magnet glue? Because if I'm running well over 100v, its going to be quite a bit over 5000 rpm... just wanted to see what it was that they think will fly apart.

Do ya think going through the transmission is a good idea (or even possible)? Or would just a straight chain be better?

[Biff]

Hmm sweet!
Do you have any idea what the rpm is limited by? Is it bearings, is it the magnet glue? Because if I'm running well over 100v, its going to be quite a bit over 5000 rpm... just wanted to see what it was that they think will fly apart.

Do ya think going through the transmission is a good idea (or even possible)? Or would just a straight chain be better?

I would bet that it is the rotor integrity that limits RPM. So that could be magnet attachment or the rotor material itself. I think with 130V you might be able to spin faster than 5000RPM at no load, but on the motorbike you probably won't be able to get the motor to spin more than 5000. The load of the bike will cause voltage sag in the battery, loss in the controller and loss in the motor bringing the voltage the coils see down.

Going by the power curve in the other forum, they have an output power of around 7200W at 3100RPM, which would give a torque of 22Nm , which is 200pound inches, so the bottom axis in that chart is pound inches of torque. Peak torque will probably be 2x that, so say 50Nm for 30 seconds.

Based on the torque capability of the motor you should be able to figure out what kind of gearing you want to be able to lift your front tire off the ground. I know I can (barely) lift the tire on that bike in second gear with some slipping of the clutch and lifting pretty hard, so that might be able to help you figure out how much torque you want at the rear wheel. Another way of figuring out the torque you want is based on acceleration and 0-60 time then calculating how much force the rear tire needs to apply. Once you figure out the gearing, you can determine what the top speed would be at 5000 RPM motor speed.

-ryan

[Biff]

To give some comparison, the Mars ME0709, which is a common in-expensive brushed DC motor has a torque up to 240 inch pounds of force (20% more than the me0913), so to get the same force at the tire, you would need to gear the 0913 20% higher ratio (rear sprocket : front sprocket) than the 0709.

The Agni 95-r performance curve goes up to 50Nm so with the me0913 you would have to have a sprocket ration 2x that of the the agni to get the same force out of the rear tire , and since they both have a top speed of 5000RPM, you would obviously only be able to go 1/2 the speed of the agni. That is assuming that the 50Nm from the agni spec sheet is a continuous rating, and not just a 30 second rating. If it is just a 30 second rating, then you are back to about the same continuous torque capability as the 0913 (assuming that the 22Nm from the 0913 spec sheet is also a continuous rating).

[Biff]

Here is a link to someone who is selling a Ninja 250 conversion. Looks pretty decent.

http://electricmotorcycleforum.com/boar ... 71#msg4171

He used a me0708 with a 60:12 sprocket ratio, top speed is around 55mph.

[tostino]

Lots of interesting info! Thanks for that.

Hopefully there will be an easy way to mod the rotor for a higher RPM limit (or it is just an imaginary limit to begin with)... Because the no load speed at 36s should be about 8000 rpm, and under load that should be close to 6000-6400 rpm.

Sadly, I am really bad at the math for all this =/ lol. I do my best, but still fail epically most of the time.

To your knowledge, has anyone ever run one motor off two controllers? Do you know of any interference with hall sensor signal when running it to both controllers, or anything like that?
I know that the throttle was an issue in the past with dual controllers, and that was fixed by using two sets of hall sensors (one going to each controller). What else may I have to look out for?

[Biff]

Lots of interesting info! Thanks for that.

Hopefully there will be an easy way to mod the rotor for a higher RPM limit (or it is just an imaginary limit to begin with)... Because the no load speed at 36s should be about 8000 rpm, and under load that should be close to 6000-6400 rpm.

Sadly, I am really bad at the math for all this =/ lol. I do my best, but still fail epically most of the time.

To your knowledge, has anyone ever run one motor off two controllers? Do you know of any interference with hall sensor signal when running it to both controllers, or anything like that?
I know that the throttle was an issue in the past with dual controllers, and that was fixed by using two sets of hall sensors (one going to each controller). What else may I have to look out for?

Yes, it has been done, there are a couple threads in the motor technology forum discussing it. There shouldn't be any problems with the hall sensors, as long as you hook the grounds of both controllers together and only use the +5v from one of them to supply the sensors. The throttles shouldn't be a problem, as long as you do the same thing (connect the grounds together, and only use one of the +5 voltage sources), but it really depends on the controller's archtecture. Plan on starting out with one controller, and if that goes well but you want more power, figure out how to get the second one to work.

-ryan

[tostino]

I actually have no idea how I would couple the motor to the transmission if I were to go that route. Have there been any other builds that have used the tranny? I've had a hard time finding that info, but then again, I'm browsing with my phone which isn't always the best for finding info lol.

How large of a battery do you think I'll be able to fit on the bike? I was thinking about 36s 30ah or 40ah.
I saw in another thread on an electric dragster, Luke said to use some copper bussbars and solder thick male bullet connectors for all the packs in parallel, and use the bussbars for series connections too. I loved that idea, because it seemed so elegant, and allowed for any size pack. I was going to use 6s5ah packs and put 6 to 8 (or more if they will fit) in parallel, and 6 in series.

[Biff]

I actually have no idea how I would couple the motor to the transmission if I were to go that route. Have there been any other builds that have used the tranny? I've had a hard time finding that info, but then again, I'm browsing with my phone which isn't always the best for finding info lol.

Check this recent thread.

viewtopic.php?f=10&t=25160

How large of a battery do you think I'll be able to fit on the bike? I was thinking about 36s 30ah or 40ah.
I saw in another thread on an electric dragster, Luke said to use some copper bussbars and solder thick male bullet connectors for all the packs in parallel, and use the bussbars for series connections too. I loved that idea, because it seemed so elegant, and allowed for any size pack. I was going to use 6s5ah packs and put 6 to 8 (or more if they will fit) in parallel, and 6 in series.

36S is pretty high voltage, Most of the 15-20kW motors are 72V systems and would break if you put them on 130V. I think I would go with 30S max. Or maybe 20S to start with, and then increase the voltage if you want more speed.

To begin with I would go with maybe 4 of these 4Ah 45C 10S packs, 2 parallel, 2 series
http://www.hobbyking.com/hobbyking/stor ... duct=14614

That would give you 74V 8Ah (592Wh), but more importantly the capability of drawing 8*45 = 360A (probably about 21kW) for less than $500 and only weight around 10 lbs. That would give you a really good performing bike, with very minimal battery expense, and it would be really easy to integrate into the chassis. Once you get familiar with the motor and all the electronics and the average power draw and the range you want, you could decide how much battery you want, and either add to the pack with the same cells to get the range you want, or design a large battery using cheaper cells. Those high discharge cells end up being about $0.87 / Wh (if their spec sheet is accurate) which isn't bad even if you are assembling a large pack, from my experience the cheapest lithium cells are around $0.5 / Wh, and I have paid as much as $3.25 / Wh, but I haven't looked at battery pricing very closely recently. Be careful when looking at prices / kWh on the internet, most companies lie about the actual capacity / capability of their cells. My bet is that by the time you are done experimenting and what not with a small high-discharge battery, there will be a better / cheaper battery solution available for your full range pack, and you will have a much better idea of what you want your pack to accomplish.

I am interested what do other people think of my suggestions?

-ryan

[tostino]

Quick update... I got some work on the controllers done this week! I also ordered some copper bars to mount the fets to! =)

[grindz145]

Controller is looking good man. I'm following closely.

[tostino]

So I got some work done on the heatsinks last night!

Take a look...

IMG_20110220_154024.jpg

The copper bars after I received them.

(146.32 KiB) Not downloaded yet

I cut some of the copper to attach to each bank of 6 fets on the high side

After I got the holes on one of the bars lined up exactly I used that to copy them to the rest

IMG_20110220_203448.jpg

Half way there!

(243.21 KiB) Downloaded 94 times

Yay!!

IMG_20110220_205617.jpg

The fets in the slots

(154.98 KiB) Not downloaded yet

Thoughts and opinions?

[tostino]

Ahh, I guess I hadn't explained my controller plan in this thread yet...

I was planning on attaching the three banks of 6 fets on the high side to their own 1/16" thick copper bars directly without any insulation between the metal on the fet and the bar (which is what is pictured above). That should allow all of the fets to heat more evenly than if it is insulated right from the start. Then those bars will be insulated with kapton tape from the 1/8" thick copper bars that will have 1/4" OD copper tubing soldered to it in a loop.

All of the low side fets will just be directly attached to the 1/8" copper bar without the need for any insulation at all. I hope that this will allow some sick currents to run through these things without anything blowing up

[tostino]

Yay! I got the tubing in last night!

Also got some project boxes that fit these controllers PERFECTLY. Pics to come later.

[Doctorbass]

Tostino, The copper is a nice idea!
This is great to see we are at least two persons here who want to test these 36 FET beast!

I would recommand you to ensure that the mosfet are not too thightened because it is know that the TO-220 tab bend easy and make small deformation that lift a bit the area right under the mosfet where the silicium chip is located. This area is exactly where the heat os the most important.




Using screw on the To-220 is more technical than it appear. The copper thread you have must not deform the flatness of the copper bar when you thighten the screw. Just few micron of space will destroy your good thermal resistance between the fet tab and the copper flat bar. Copper is soft metal and on 1/8 thick it can deform easy the faltness under the FET. Often you can't see it with eyes but there is real gap



That is why i think i blown many controllers.. there is only need of ONE FEt incorectly thightened to weak the entire controller

This time i prefered applying pressure on the mosfet instead of using screw. we'll see..




If you use Kapton, use very small qty of thermal paste to fill the little air gap of the imprefect flatness of the surface. Kapton is not soft enough to fill these gap.



Here is the link i use as a great information reference

http://sound.westhost.com/heatsink3.gif


I look forward for your results

Doc

[tostino]

So due to LFP's awesome video, I have decided that I would have to be retarded if I didn't end up soldering the fets directly to the copper heat sinks. I ordered some of this stuff: http://cgi.ebay.com/ws/eBayISAPI.dll?Vi ... 1995wt_905 and will see how it works.
I need to get some higher temp solder for the low side bar's tubing on the back, because I don't want that detaching while I try and solder the fets in place.

[tostino]

Yay! I got some high quality low esr caps in from Digikey! 160v 1000 microfarad! Also got in some 4w 0.001omh shunts.

Pictures later =D.

Edit: On another note: I signed up for my motorcycle endorsement class today for the weekend :p. Hopefully it won't be too bad.

[Doctorbass]

Yay! I got some high quality low esr Panasonic caps in from Digikey! 160v 1000 microfarad!

I bought the same for my first 18 fets lyen controller. I succeded to put 4 of them in that case!! 22mm diameter each if i remember.

I also compared them:

Infineon 1000uF 100V Rubycon: cap: 998uF ESR: 110mohms
Crystalyte 1000uF 100V DTDZ : cap: 1063uF ESR: 260mohms
Panasonic 1000uF 160V Panasonic:cap: 740uF ESR: 73mohms

http://www.endless-sphere.com/forums/viewtopic.php?f=2&t=10312&p=163099&hilit=panasonic+esr#p163099

http://media.digikey.com/photos/Panasonic%20Photos/TSHC%20SERIES%2040,22D.jpg


Doc

[tostino]

Hmm, hopefully I didn't make a mistake with my cap choice...
I ended up with these: http://search.digikey.com/scripts/DkSea ... 93-1171-ND

[tostino]

Just a little update... I took my motorcycle enforcement class this weekend, it took about 11 hours a day both Saturday and Sunday. I passed!