40kw E-Taljet Dragster. Thoughts?

Hey Guys,

I frequent the bicycle forum. So, many of you may not know me. However, my latest project fits here, not there.

As the title suggests, I will be building an E-scooter within the next 6 months or so.

Here is a Google picture of an Italjet Dragster for those who are unfamiliar with this hot little scooter
https://encrypted-tbn1.gstatic.com/images?q=tbn:ANd9GcSEtVjHZJVSTUVVNnbY400lfka6dNA3zdNrCIb2-1ZQ9_8BiM0NGw

What happened is, a local investor contacted me regarding my E-bike drive systems (I machine high-end brushless drive systems for bicycles). He and I met together and we decided to look at the high-end scooter market rather than the E-bike market because there is an opening in that arena. In an attempt to gauge market reaction (and just because it is flat out FUN) we decided to buy and convert two Italjet Dragster 50s. One will be radical and one more mundane. I will be building the radical one first and that is the reason for my post here.

I will be using this motor
http://www.astroflight.com/motors/brushless-motors/4535.html

And their matching controller
http://www.astroflight.com/electronics/speed-controls/esc-2413.html

This motor and controller are rated at 15kw continuous and (according to the engineers at Astro) well over 40kw burst.
I am an Astro Flight dealer. So, this is a strategic purchase. The motor and controller are already on order.

I would like to run up near 100 volts with two of these modules in series
http://evolveelectrics.com/Enerdel_MP310-049_Moxie.html


The plan is to stretch the rear a bit and lower it a touch. I will also lower the bars and I may run a wider and slightly larger rear tire depending on what is available out there. I want it geared for about 70mph. I will be making a completely new rear swingarm. So, I can configure it whatever way suits my purposes.

What I am curious about is; Would you guys run the stock CVT or direct belt drive to the rear wheel? I am partial to direct toothed belt drive partly because I do not know if the CVT will take the load. I also want a very direct "Hit" when throttling hard. I would run 8mm pitch in whatever width my pulley shop deems correct for the torque load.

Also, I am hoping for relatively decent range (obviously if ridden casually) without adding too much weight. 3kwh should be enough to screw around with without worrying about range.

What are your thoughts?

Matt

Hi Matt,

I have seen a number of miserable failed attempts to use a CVT with an electric motor. I think it would be a mistake. Recently discussed here: http://www.diyelectriccar.com/forums/showthread.php/pgo-500-conversion-cvt-90657.html?highlight=continuously Note the input from member steelneck. And the bad results on a TTXGP bike by squarewave racing a few years ago.

I know you have already chosen the motor, but quickest electric dragsters run series wound motors. But with your tire you might not be able to handle that torque anyway.

Good luck,

major

Matt, The Sora electric motorcycles sold by Lito green motion in Quebec are using a CVT.

Personanly i tried their E-motorcycle on the Racetrack of circuit Gille Villeneuve in Montreal and it was really impressive how well the torque and power are spread over the speed.

http://www.litogreenmotion.com/

Doc

Dragster is just the name of the scooter, right? You're NOT talking about racing?

While certainly you need to review the failures of others, there's no law that you're forced to give up. CVT's are good things, you just have to be one to figure it out. Everything that comes before the success is more or less failure, right?

I find the input of Major more telling.

Unfortunately others don't often document. . . it is nearly impossible to locate. . . searching for _CVT_ comes up empty. It's been a while. I don't recall what I said before.

The typical CVT is designed to overcome the weakness of the ICE, namely the inability to self start (produce torque at zero RPM) and to produce meaningful torque at lower RPM. The electric motor excels in these areas.

Click to expand...

The link meant to discourage you further offered encouragement from another link http://www.teslamotorsclub.com/showthread.php/9982-CVTs-for-Electric-Vehicles :

1) This article validated the improvement in torque off the line and vehicle top speed by modifying a two-wheeled EV with a CVT. Note the negligible range impact in Table 1.
2) The upcoming Lito Sora has successfully applied this concept to a high-performance E-Bike

http://www.fallbrooktech.com/sites/default/files/videos/LEV_CVT_Whitepaperrev10.pdf

Click to expand...

I will say when a CVT is used in racing you gotta keep the RPM up in turns yet below the speed where the centrifugal clutch engages. People who drive CVT cars talk of 'Lag,' probably the same issue at work. Maybe you could do without the clutch.

Snowmobiles make extensive use of CVT's, maybe a snowmobile forum could teach you something. I'm reading a bit here from those other links and no, people do not understand the CVT. (Steelneck is GUESSING.) It is about keeping the engine at the optimal RPM at all times. You would then be using the CVT to keep the electric motor at whatever voltage you would consider optimal, if you had a 72 volt/50 kv motor you be running the 72 volts to be at 3,600rpm, rather than running the RPM up and down. Sounds good in accelleration, it becomes a different thing when you're trying to go, say, 40mph and stay at that speed.

I read a book about Henry Ford as he's putting together experiments for his planned car. People told him WHY what he was doing wouldn't work. He had to go to the equivalent of a paint store to get a can of the uncommon gasoline. But the maven of electricity Thomas Edison listened to what he was building and told him he had "Just the thing. . . ." (Including about using a gas engine instead of electric, at that time.)

But you'll have to come up with the answers, as noone else has them. Bachelier, Osborne, Mandelbrot, they were all berated as fools. Guess who won out in the end.

http://www.diygokarts.com/vb/showthread.php?t=19290

The term "Dragster" is the name of the scooter, not the type of vehicle.

This is a high performance scooter. It will not be drag raced.

Matt

yes a CVT will and can hold up to this sort of power, maybe not this one in this bike ( is it a 50cc or bigger ? ). I would also say go with the whole CVT , if you go with using the CVT then 40kw of motor power can push you well over 100mph with very very good acceleration... ( just look at the maxi scooters and what there performance is from such small engines). I know you would like to stick to what you know but when I went down this route with using a CVT a couple of years ago it do work very well. Efficiency of a CVT can be just as good as a manual gear box ( if not better for a number of reasons ) so don't listen to much when people start to tell you that they are just not efficient ( the benifits of a CVT out weights a direct drive system and a manual gear box ). 10-15kw of power on a smaller scooter will be pleanty to loop out the scooter and push it to 70mph but its all about setting the complete CVT up to work well with the available power.

Edit:
As dauntless said above , the trick is to learn how a cvt works, something that very few understand ( but they think they do! ).

also there is no reason to have any lag in a cvt system, if there is lag then its by design or lack of good setup.

Matt I am familiar with the Dragster scooters, has to be the nicest scooter chassis there is! There are so many parts available for tuning cvt I would go with it. it may even allow you to use a cheaper controller if you do make this a production conversion.
But which ever way you jump I am sure this is going to be awesome!!

I think the lag they refer to is over the clutch not engaging until the RPM is however high. I raced pocketbikes, we wanted over 8,000RPM and you just had to keep on the throttle as you slowed so you weren't waiting for the engine to wind up. I thought of that as the #1 challenge with those. Since you don't have the motor 'Idling' you don't need the clutch, it'll spin while you're slow and give you power instantly.

http://www.buggydepot.com/buggy-tech-center/read/150cc/0/15/how-it-works-the-cvt-system-on-150cc-gy6-engines.html

I realise that this may not be in line with your project as it seems you want to use astroflight kit, but surely Luke (Live For Physics) and John in CR would recommend one of John's Hubmonsters for this project. You could then replace the unsprung weight of the CVT with his hub motor for what, as I understand it, a lot less money.

On the subject of Dragsters in particular (I have 2, a 50 and a 172 which is a bored out 125), make sure that your front upright has no slop as used ones are often prone to this due to a lack of maintenance.

the draw back with a bike hub motor for this application and getting the sort of performance that is required is the weight and the massive controller that would be needed. By using a rc setup through the CVT you can have the performance, less weight and do not need to go over board with the current capacity of the controller.

CVT would be a huge mistake. Every additional power transfer stage adds losses and failure modes and noise pointlessly.

If you want big power to a wheel in a setup that runs in a reliable way, simplicity will always be the winner. Think of a top-fuel dragster, single speed fixed gearing is where you end up when pushing the limits of performance. Choosing the right electric motor makes it absurdly easy to do as well.

liveforphysics said:

CVT would be a huge mistake. Every additional power transfer stage adds losses and failure modes and noise pointlessly.

If you want big power to a wheel in a setup that runs in a reliable way, simplicity will always be the winner. Think of a top-fuel dragster, single speed fixed gearing is where you end up when pushing the limits of performance. Choosing the right electric motor makes it absurdly easy to do as well.

Click to expand...

we are not talking about a top fuel dragster.. I learnt from what I have done with a CVT that you can get huge performance from very small kw's and I was only playing around with a current limit of 3kw on a scooter that was 65kg with all body panels and lights fitted. we dont what to start going around this argument again , do we ? . I have run 3kw on a fixed geared motorbike ( lighter 55kg ) ( could be geared for the same top speed or the same acceleration but not both ) and it would not come close to what the scoot (cvt) could do, performance wise . Yes It may be very easy to just put a big motor/controller with a direct drive ( at what I would say will be of much higher cost and weight ) but where is the fun in that, with the scoot in question I would say that the total weight if using the cvt should be no worst than 65kg with the target power output, Im not sure if this weight could be matched if you go down the route of a big motor with a big controller.

Gwhy- I'm not saying you can't make something that moves around under it's own power using any amount of power wasting series transfer stages and reductions etc. I know you can, many of us have done it, and it can be an effective band-aid to make a poorly suited motor get by in a given application.


I'm saying, if you're doing ground-up design of a drivetrain, don't design-in inherent needs for a bunch of band-aids just to function acceptably.

You can't make something more efficient, more reliable, or higher performance than simply creating the energy conversion from electrical to mechanical in the closest state to it's final usage as possible. In other words, if you're powering the wheel of a 60mph roller blade or something, go direct drive with a 10,000rpm motor to the wheel (or whatever). If you're powering the wheel of a scooter that you want to be as high performance and reliable and efficient as possible, choose something that directly converts your electrical energy into a state directly usable by the wheel, or usable with as FEW as possible power transfer stages in-between your motor and the wheel.

Every power transfer stage robs consumes power and converts it into heat and noise and adds new failure modes. In exchange for that you get to use a motor package size that is much harder to keep cool for hot-rod applications.

It's electric power. The coolest advantage we have is the ability to free ourselves from all the mechanical BS band-aids required to let a 50cc torqueless stinker capable of getting started on a hill. We can actually produce the drive torque we want natively with no power wasting noisy band-aids. Embrace it!

Hi Matt,

I like your scooter choice!

recumpence said:

He and I met together and we decided to look at the high-end scooter market rather than the E-bike market because there is an opening in that arena. In an attempt to gauge market reaction (and just because it is flat out FUN) we decided to buy and convert two Italjet Dragster 50s. One will be radical and one more mundane. I will be building the radical one first and that is the reason for my post here.

Click to expand...

I have some concerns:
I'm not sure if roughly twice the power of a Vectrix (without the CVT) will satisfy your definition of "radical".

I think using a $4k motor and controller you will have a very limited market.

http://www.astroflight.com/electronics/speed-controls/esc-2413.html

The ESC-2413 is a high power, high voltage, high current sensorless brushless motor speed controller,capable of up to 15kW.

Intelligent adjustable startup sequence allows the controller to start the motor under light or heavy loads

Click to expand...

And is a sensorless controller really a good choice (no pedals)?

Matt's question is not a generic question "if he should use a CVT or not".

The question is if with a 15kw sustained and 40kw burst motor he should keep the stock Taljet Dragster CVT.

Two questions:
Can the Taljet Dragster CVT handle 15kw/40kw efficiently and without blowing up?

If it can does it perform well enough with 15kw/40kw without using the CVT. My guess is that if the CVT handles the power well he will be happier with the performance with the CVT.

This should give him some idea of the performance with half the power with a single speed Planetary Gear Reduction:
http://www.vectrix.com/content.cfm?n=products-vx1-inner#page=page-2

110 km/h, 68 mph
6.0 seconds, 0-80 km/h, 0-50 mph
193 kg, 425 pounds
Planetary Gear Drive

Click to expand...

Heavier model (bigger battery):

110 km/h, 68 mph
6.25 seconds, 0-80 km/h, 0-50 mph
209 kg, 460 pounds
Planetary Gear Drive

Click to expand...

http://en.wikipedia.org/wiki/Vectrix

Top speed 109 km/h (68 mph)
Power 20.2 kW (27.1 hp) peak; 7 kW (9.4 hp) continuous
Torque 65 N·m (48 lb·ft)

Click to expand...

I think you should consider using Leaf battery modules (but maybe 3C isn't enough for you). 2s2p 30AH, available for about $125 on ebay.

More info here:
https://www.endless-sphere.com/forums/viewtopic.php?f=14&t=52162

15kW max output (according to Astro's site) at 8,000rpm.

That's 13.2ft-lbs of torque or 2534in-oz of torque (to use Astro's units).

From the data on the page, all motor wind options have identical copper fill and hence identical continuous torque capabilities, they all produce 507watts of just copper resistive loss to produce that 13.2ft-lbs of torque assuming no other losses. 507watts of heat just in copper alone (and keep in mind, if it's running at peak efficiency at that point, it would have an equal amount of core loss heating as well) is REALLY hard to shed continuously (but possible with enough airflow).

If you wish to make 40kW of output from that motor at 8,000rpm, you require 34.3ft-lbs of torque at the motor shaft. Any wind choice will make a staggering 3427watts of just copper heating alone to be producing that torque, assuming it wasn't already saturated at a lower value and can achieve it.

The controller is running SMD FET packages? It seems like planning to fail IMHO, but maybe they found the magic trick nobody else has.

Take it from a guy who's done some motors with multiple kW's of copper losses, it is not a trivial amount of heat to shed even when it's coming from a 11" diameter motor let alone one with 4.5" diameter.

The efficient way to make torque for stuff you want to be able to hotrod super hard is to grow motor radius. Pretty much improves everything to do with performance always, and it generally ends up more quiet and compact and less complex.

Hi,

If you wish to make 40kW of output from that motor at 8,000rpm, you require 34.3ft-lbs of torque at the motor shaft. Any wind choice will make a staggering 3427watts of just copper heating alone to be producing that torque, assuming it wasn't already saturated at a lower value and can achieve it.

Click to expand...

Other words about the same as the stock vectrix?

The efficient way to make torque for stuff you want to be able to hotrod super hard is to grow motor radius. Pretty much improves everything to do with performance always, and it generally ends up more quiet and compact and less complex.

Click to expand...

Have any suggestions for good motor in your opinion for this application?

Even in a mid-hub configuration John's motors are too small (about 7kw)?

The purpose of this motor is to have light weight with good efficiency that can sustain 8 to 10 kw while giving very short bursts at 30 kw or more. The majority of the time it will be just cruising around at 4 kw or less.

As an Astro dealer, I get products at cost and I need a way to test new items.

This scooter is just a demonstration of who can be accomplished on a scooter platform at a low weight. Production is a completely different story.

Matt

Hmm.. It says 120kG for the 50cc scooter mass? Can you confirm it's that high?

I think you may find all your needs suited better, with respect to performance, efficiency, continuous top speed, and weight (I know your amazingly high quality craftsmanship can make any design as reliable as DIY things get), by gutting all ICE mechanical power-train related parts and installing one of JohninCR's Hubmonster motors.

The way I see a drivetrain these days has changed from how I saw them years ago. I once believed the right path was to spin a lightweight ultra efficient motor to the moon and build all the supporting systems to get that into some useful form of power to turn the wheel. Now I simply see any portion of a powertrain that is power transfer as non-active-material and power wasting dead weight to lug around.

If you're looking for performance, and you're adding mass to a vehicle that isn't battery/controller/copper/lamination iron/magnets/or aero fairings, it's not helping your performance or efficiency, and it's likely just parasitic.

Think about this. Lets say you want to make a ripping EV drag scooter. Lets say you're willing to add 25lbs of powertrain mass. Whatever system will result in the largest amount of that 25lbs of mass be active material (copper/iron/magnets) that you can keep saturated for long enough to go down the quarter mile (or burst up to 80mph in a few seconds or whatever) will be the system that wins for real-world power output.

From my own many hours of dyno testing and messing with gearing and motors and things under real loads with real instrumentation, I've learned you really can't replace having more copper/iron to saturate. Whatever system lets your vehicle be the highest percentage of it's mass as copper/iron/magnet will be the system capable of the highest performance in acceleration, and typically continuous power as well (though I know that nice fan in the big Astro has got to help continuous power quite a bit, though for short drag racing it would only hurt performance).

It's why bikes like Shawn Lawless's with a 13in GE motor hopelessly slaughter bikes like Kilacycle to an extent it will never be possible to be competitive with out a ground-up redesign. Both machine saturate all the active material they've got for the whole quarter mile, but the big 13" motor simply has way more active material to saturate than the pair of smaller motors kilacycle uses.

If you're making something electric for accelerating rapidly, you want as high of portion of vehicle mass as possible to be active motor materials, and a battery/controller combo that is JUST big enough to make it all saturate. That is the path to expanding the EV performance window.

That said, I read those scooters are a ball with 20hp, and electric powerbands crush gas stuff, so I know no matter what you choose to do, it's going to be a hoot. If you want it to be mind-bendingly quick though, I've given you the secret to making it happen.

So a hub is out of the question and different make of motor is out, now you are down to 2 choices use the CVT or not.
the choice is obvious if you what that top speed of 70mph+ and loop out acceleration the it will have to be the CVT.
But there will be a big learning curve to understand how the CVT works (roller weights/ramps/torque spring and clutch )
you will need a rpm onto the front variator of around 10krpm maybe even upto around 12k will be better ( if its a stock 50cc scooter )
12k should take you up 70mph without having to adjust the final secondary gearing to the wheel. ( small gearbox in the rear wheel )
The clutch will need the weakest springs possible, I removed the springs completely and this gave a positive engagement at around 1.5k rpm at the front variator, ( will be even be better if this positive engagement rpm can be brought down further, stock it will be around 6k rpm ). You will need to tune the cvt ( rollers and torque spring ) to set your max current pull under acceleration and this is the tricky bit.

if you are going to try to push 30-40kw peak through a 50cc cvt then I think you will be in danger of snapping drive belts ( the maxi scoots 125cc+ have wider belts than 50cc scoots ) but I think upto 20kw peak will be fine ( and more than enough to meet your needs)
If you dont go with the cvt then I think you will defo need the 40kw+ peaks to satisfy your needs , but this means higher discharging battery's and a controller than can take these 30-40kw peaks for at least 8secs to make it reliable.

Great project, please keep posting about progress. Enerdels are great batteries but I doubt you'll need the MP modules, the ME's should work fine and are cheaper.

Thanks for all the input, guys. I truly appreciate it.

I guess I have been against hub motors for the issues of rotating weight affecting handling and lugging around a hugely heavy motor on a vehicle that will be mostly cruising with the occasional high burst. I also like being able to mess around with gearing. That being said, I have been so focussed on doing things the way I do them that my axiom has become narrow (at least in this regard).

The battery pack choice is flexible. However, that pack is good for 15kw continuous and 45kw burst, and the NMC chemistry has a very long life expectancy with high resilience to overdischarging. I am open to changing that, though, if I see a need to.

As for the motor and controller, I have a deal worked out to get one of each really innexpensively to familiarize myself with them as I am Astro's largest industrial motor dealer. I have spent a huge amount of time on the phone with their engineers about this. Yes, heat has been an issue for them. That is the reason they use a large fan. Also, the rotor is hollow to aid core cooling. They insist there are a number of other areas they addressed to increase air flow for better cooling. We shall see. The controller uses four 200 amp FETs in parallel. I was told the controller has a "Bottomless" current output in relation to motor saturation. We shall see on that too. I am not against running sensors and a different controller. That is another learning curve, however. I can tell you I want to minimize drive system weight as much as possible while containing the juice for 60 to 80mph power wheelies.

Luke, I may begin a PM conversation with you about some of the tech involved here. I need to learn more. I know what I know, but one who relies only on what THEY know is setting themselves up for failure. :wink:

Matt

Really cool project.Keep us up informed.

gwhy! said:

So a hub is out of the question and different make of motor is out, now you are down to 2 choices use the CVT or not.
the choice is obvious if you what that top speed of 70mph+ and loop out acceleration the it will have to be the CVT.

Click to expand...

Holy chit. Gwhy, you think because he ordered a motor he is married to it? Matt is a sharper and more forward thinking guy than to feel trapped like that. Lol

He could simply choose an appropriate traction motor and skip all the power wasting nonsense, save money, save weight, make radically improved performance.

You can choose a motor that makes the torque you want over the speed range you require. At some point you realize it becomes pretty silly not to.

liveforphysics said:

gwhy! said:

So a hub is out of the question and different make of motor is out, now you are down to 2 choices use the CVT or not.
the choice is obvious if you what that top speed of 70mph+ and loop out acceleration the it will have to be the CVT.

Click to expand...

Holy chit. Gwhy, you think because he ordered a motor he is married to it? Matt is a sharper and more forward thinking guy than to feel trapped like that. Lol

He could simply choose an appropriate traction motor and skip all the power wasting nonsense, save money, save weight, make radically improved performance.

You can choose a motor that makes the torque you want over the speed range you require. At some point you realize it becomes pretty silly not to.

Click to expand...

The OP mentioned the motor he would be using and the controller he would be using, he also said later in the thread that he will be using astro...

ok , he can go for a ( different brand ) big motor and big controller and do it the same way that everyone else with bottomless pockets do it , but its nothing new or radical ( and it will be heaver than the original gas version of the scoot ) so I cant see how it can be impressive when its nothing new ( even more so if the price tag is mentioned ).

I doubt matt will feel trapped, after all he already set the choices out in the OP , if he is this sharp forward thinking guy that you say he his then he will not be influenced by me or anyone else..

I also did say later in the post :

If you dont go with the cvt then I think you will defo need the 40kw+ peaks to satisfy your needs , but this means higher discharging battery's and a controller than can take these 30-40kw peaks for at least 8secs to make it reliable.

Click to expand...

Luke is just yankin your chain. :wink:

I am against a hub motor for this for the pure reason of unsprung/rotating weight.

Anyway, I may even build multiple drive units for it that can be swapped out for testing purposes. We shall see.

Matt