Better than the Stark Varg? 76v 90Ah 35kw Dirtbike Build

Here is my project for the winter:
Motivation and solution:
I ride really gnarly single track trails here in southern Idaho and Utah. The instant power delivery from electric is so much better than gas in my opinion but, due to other drawbacks, I'm currently on a Beta 300rr Race Edition 2 stroke gas bike and loving it. The biggest reason is obviously range. I can put 40 miles of super hard terrain on that bike and have more than half a tank left at the end of the day. Another big reason is suspension. I've tried Surron light bees, ultra bees, x260s, and even other custom builds and all of them have been seriously lacking in these areas and not good on power either, before thousands in upgrades. So my solution is to build a full size bike with the suspension and comfort I love, more than enough power thats tunable to my liking, and enough battery to go on the longer trails I like. Another goal, since the Stark Varg is supposedly the best on the market, is to be lighter and also have more range since both of those aspects are disappointing for that bike. Based on where Im at so far, thats looking very promising.

Frame and suspension components:
-I bought a 2008 KX250F as a starting point. I chose this bike because its one of the only bikes that has the honda style split aluminum frame AND KYB "SSS" forks. SSS is the Yamaha term but same deal. KYBs before 06 were not good from my understanding.
-The SSS forks are the only forks I love every time I ride with them. I have not had good experiences with the older KYBs, Showas, ZF, or WP options.
-The reason for the frame choice was to give me a lot of room in the frame without a lot of weight. Also Im an industrial designer so I have some cool visual ideas that work with that exterior frame tube shape.

Motor and controller:
-The motor I have is a QS138 70H V3 that has been upgraded by Electro and Co. They machine the gear housing from billet aluminum and do some other upgrades to help the power capability and longevity. This is currently the limiting factor as far as power goes in the system with a 35kw capability.
-The controller is the only part I don't have yet. That's because the one I want isn't quite released yet. Its also from Electro & Co. Its the Trumoto Zapper controller. Its capable of 50kw (nominal) at 76V. 600 amps max current. They said it should be released well before spring so Im still planning on using that for now.
-Other options I'm considering for the controller, if that isn't released in time, are the EBMX X-9000 and the EM-260. The x-9000 would allow me to use a 22s config instead of 21, but still results in less KW capabilities (the motor would still be the limiting factor so NBD) Its also about $300 more than the Zapper. The EM-260 would limit me to 28kw and doesnt have an easy to use programming app so that one would be my last choice, though the cheapest.

Battery:
This is a big part of the build. Im building my own battery using Samsung 50S 21700 cells (25 amp output continuous/45 peak)
-The configuration is 21s 18p resulting in 378 total cells (big boy) Based on the weight numbers on the website where I got them (18650 battery store) its about 58lbs in cells.
-Total amp output ability for the cells in this configuration would be 450 amps continuous and 810 amps peak. Sadly I have not had good luck finding direction on how to make connections that are able to take that kind of amperage so I will be doing some testing and posting it here. The max current the motor can handle is 461 so as long as I can get it to take that without any issues it will be a success.


Comparison to Varg:
Weight: Estimated 235lbs - Varg: 242lbs
Energy storage: 7kwh - Varg: 6kwh

Forgot to add: I'm using an ANTBMS 420 amp continuous, 1050 amp peak BMS

Blacksheep Engineering said:

The EM-260 would limit me to 28kw and doesnt have an easy to use programming app so that one would be my last choice, though the cheapest.

Click to expand...

I might be wrong since I never actually owned one, but I do own a EM-150 and I'm getting 23KW from it with my current settings. I've set up 280A in the programming app and I've seen peaks going up to 330A being pulled from the battery during accelerations (BMS screen, which is fairly accurate). It has a lot of power, it's pretty brutal. Doesn't heat much so far either so I guess it's not even maxxed out. So, given what the EM-150 is able to do I suppose the EM-260 can do way, way more than the advertised 260 battery Amps.

The programming interface is indeed a horrible mess. Litterally the worst I've ever used, and by far. But the hardware is absolutely brilliant, these controllers are very well made and the build quality is top notch.
I'd start with this one if I were you, it's cheap so you can experiment a lot with it without risking too much money.

Nice setup you're planning tho, I hope it works out for you, please keep us posted!

Nice project but the power level you are trying to acheive will not happen without serious overheating. Needs water cooling for any serious riding. Also that 45 amp peak rating comes with lot of voltage sag and heat buildup. You also need to cool the battery with battery case.

Impressive project and goals. A Stark Varg seems very difficult to match if you look at the total package.

rivvs said:

Nice project but the power level you are trying to acheive will not happen without serious overheating. Needs water cooling for any serious riding. Also that 45 amp peak rating comes with lot of voltage sag and heat buildup. You also need to cool the battery with battery case.

Click to expand...

Is it just the cells that you're saying will build heat? The peak power the motor will pull is only 10 amps higher than the 450 amp coninuous rating of the p packs.
At that amount of draw it shouldn't ever build heat even if I was hitting those peaks for a long time. In reality it could only ever hit those peaks for a couple seconds and not very often.
Even if I was going to pull closer to the 810 amp peak ability of the cells, it still wouldn't be for long enough to build any significant heat.
From testing I've done of other builds, I should only need to use about 23kw for the power I like for most riding, so even using the full coninuous rating of the battery would just be for kicks and giggles.

If you're talking about the motor, I've had the same concern but talked with electro and co extensively about their motors and they say they run them at 35kw for motocross riding constantly and they hold up great. Kinda taking their word on that one but the riding i will be doing requires a lot less consistent power so I should be just fine.

Is there a good way to cool that motor if needed? It doesn't have any coolant channels in it from what I understand.

Dui ni shuo de dui said:

I might be wrong since I never actually owned one, but I do own a EM-150 and I'm getting 23KW from it with my current settings. I've set up 280A in the programming app and I've seen peaks going up to 330A being pulled from the battery during accelerations (BMS screen, which is fairly accurate). It has a lot of power, it's pretty brutal. Doesn't heat much so far either so I guess it's not even maxxed out. So, given what the EM-150 is able to do I suppose the EM-260 can do way, way more than the advertised 260 battery Amps.

The programming interface is indeed a horrible mess. Litterally the worst I've ever used, and by far. But the hardware is absolutely brilliant, these controllers are very well made and the build quality is top notch.
I'd start with this one if I were you, it's cheap so you can experiment a lot with it without risking too much money.

Nice setup you're planning tho, I hope it works out for you, please keep us posted!

Click to expand...

Thanks for the input on the em-260. Electro and co has a preprogrammed version that they say is great but I would have to still deal with the programming if their setup isn't perfect for me (very likely). I'm definitely willing to spend the extra couple hundred bucks to have the nice app to adjust things, but if there is a nicer way to program these em controllers, I'd love to look into that.

SlowCo said:

Impressive project and goals. A Stark Varg seems very difficult to match if you look at the total package.

Click to expand...

Thanks, I am very excited for the challenge.
You bring up a great point. The "total package"

I want to clarify, the Varg is very good at what it was designed for. The issue is, it wasnt designed for what 90% of people want in a full size electric dirt bike. All of us on full size gas bikes doing enduro, singletrack, and other trail riding that want a good electric competitor want range and agility. Not 80hp. Everyone I've seen attempting to trail ride a varg puts it in the 45hp mode so they can get somewhat tractable power and ride for at least a little more than an hour.
My goal is to make a bike that's tunable from 50hp down to exactly what you want for riding (probably closer to 35-40), and make it lighter, more agile (I need to do a post about my weight positioning), and last a lot longer than the varg. Im sure, if you can keep the front end down, the 80hp version of the varg will beat it in a drag race, that's not part of my goal.

Better think in 24s or more
if you think about so many kw

gustavo88py said:

Better think in 24s or more
if you think about so many kw

Click to expand...

I'd like to hear your reasoning for this. My limiting factor is the controller I plan to use which has a 21s max. Im guessing you just mean because of the amp output? I can handle 450 amps continuous with this setup so it should handle the 35kw (460 amp) peaks just fine. Im making up for 21s with 18p. I am still open to higher voltage controller options for the next little bit until I start the actual battery build or buy that controller. I just cant find anything over 22s with a nice interface for programming

I wonder if you really can hit 35kw output with 21s, if the seller say so I guess it might be possible..
There was some data on a qs138 90h on 20s I think. From what I remember about 40hp was the max they were able to get.
To get further they would need higher voltage.
I would go for at least 28s

I think the varg has about 6,5kwh, not 6kwh. So 7kwh is not a whole lot more. That is with molicel p45b, so they will handle high discharge better. So the difference what you get out is likely even less.

Dont get me wrong, I think it can be a really nice enduro bike

21S or 24S setup does not make a big difference in current.
A step up is 28S or even more.
35kw from 21S = 460A
35kw from 24S = 400A
35kw from 28S = 345A

Fardriver ND961000

j bjork said:

I wonder if you really can hit 35kw output with 21s, if the seller say so I guess it might be possible..
There was some data on a qs138 90h on 20s I think. From what I remember about 40hp was the max they were able to get.
To get further they would need higher voltage.
I would go for at least 28s

Click to expand...

Do you know of a good controller that can do 28s? Id love to do even higher voltage that that but from my research the only controllers that are considered reliable and have relatively intuitive programming are limited to 22s or less. I am open to options if they exist.

Blacksheep Engineering said:

I just cant find anything over 22s with a nice interface for programming

Click to expand...

I would recommend taking a look at 3shuls offerings, they make 30s/40s offerings that are vesc based, which should very well suit that "nice interface" request haha. The advantage of stepping up to a higher voltage is you generate less heat for the same kw output as you are running lower amps, this allows you to sustain those peaks for longer periods of time.

Blacksheep Engineering said:

Do you know of a good controller that can do 28s? Id love to do even higher voltage that that but from my research the only controllers that are considered reliable and have relatively intuitive programming are limited to 22s or less. I am open to options if they exist.

Click to expand...

Heres a thread I posted a bit ago doing a high level breakdown of the more common higher voltage offerings Tronic X12 or similar high-voltage controllers?

thebes said:

I would recommend taking a look at 3shuls offerings, they make 30s/40s offerings that are vesc based, which should very well suit that "nice interface" request haha. The advantage of stepping up to a higher voltage is you generate less heat for the same kw output as you are running lower amps, this allows you to sustain those peaks for longer periods of time.

Click to expand...

Had not seen that before. Looking into their options now. Thanks!
They even have downloadable CAD files!
Looking at this one right now.
At 30s, 350 amps should be enough. Its significantly cheaper than the one I was looking at. I will need a new BMS to do 30s but thats ok I can use this one for something else. Let me make sure I can find a new BMS that can do 30s.
This would make connections a lot easier. Let me do the math on how many output amps I could have with roughly the same size battery (less P packs)

Noob question: I havent used Vesc, I have very little coding experience. From my quick research that looks like a coding software. Is this something I can tackle without diving deep into coding? By nice interface I meant they have a nice little app where I can drag bars to adjust power and power curve and stuff. If there is a good starting point and I just need to hop into code to adjust some values, thats probably doable other than I cant do that on the trail.

Thanks again for pointing out this option, that is a big help.

I think the cl700 is too small. 700pA is a max value, in reality you will only be able to run 5-600pA. I think you should get a cl1000.
I dont know if I can say that they are very reliable, but the v4 that they have now is supposed to be better than earlier versions.
The motor parameters are a bit tricky, but you dont need coding. It is just an option to be able to make your own functions.
I remembered wrong, 40hp was a 138 90h with 24s:

j bjork said:

I think the cl700 is too small. 700pA is a max value, in reality you will only be able to run 5-600pA. I think you should get a cl1000.
I dont know if I can say that they are very reliable, but the v4 that they have now is supposed to be better than earlier versions.
The motor parameters are a bit tricky, but you dont need coding. It is just an option to be able to make your own functions.
I remembered wrong, 40hp was a 138 90h with 24s:

Best controller for QS138 V3 70

If 150 battery amps were a limit for the motor .. why not 1500 amps at 10% duty cycle? For an oversimplification, you're not entirely wrong. This is why motors (and batteries) often have continuous power ratings as well as peak power ratings; my 8000w rated hub motor can output 14,000-16000w...

endless-sphere.com

Click to expand...

I dont believe I have a way to gauge the phase amps I need without testing. The 21s limited controller I was planning on getting can do 1200 phase amps. Maybe that option is better than these. As far as your comments on not getting 35kw from 21s. Its just limited by amps. Though it will take more work in connections, I can get the amps necessary to give the full 35kw to the motor.

Blacksheep Engineering said:

Noob question: I havent used Vesc, I have very little coding experience. From my quick research that looks like a coding software.

Click to expand...

Vesc is a open source firmware that is intended to run motor controllers, vesc tool is the counterpart to this, it is the tool used to interface with controllers.

Blacksheep Engineering said:

Is this something I can tackle without diving deep into coding?
By nice interface I meant they have a nice little app where I can drag bars to adjust power and power curve and stuff.

Click to expand...

Yep 100%, while it does it has the option to add your own lisp scripts to it for extra/custom functionality it is mainly entering values and selecting options.

I would also suggest the cl1000 as the cl700 isnt really enough to fully saturate a 138. Also I cant speak on the reliability of the cl1000 or older versions of the cl700, but my v4 700 has had zero issues so far and has been running solidly for ~8 months now.

Blacksheep Engineering said:

Is it just the cells that you're saying will build heat? The peak power the motor will pull is only 10 amps higher than the 450 amp coninuous rating of the p packs.
At that amount of draw it shouldn't ever build heat even if I was hitting those peaks for a long time. In reality it could only ever hit those peaks for a couple seconds and not very often.
Even if I was going to pull closer to the 810 amp peak ability of the cells, it still wouldn't be for long enough to build any significant heat.
From testing I've done of other builds, I should only need to use about 23kw for the power I like for most riding, so even using the full coninuous rating of the battery would just be for kicks and giggles.

If you're talking about the motor, I've had the same concern but talked with electro and co extensively about their motors and they say they run them at 35kw for motocross riding constantly and they hold up great. Kinda taking their word on that one but the riding i will be doing requires a lot less consistent power so I should be just fine.

Is there a good way to cool that motor if needed? It doesn't have any coolant channels in it from what I understand.

Click to expand...

Yes even at the continuous current the heat will build up. I have a. 26s14p and at 300 amp there is heat coming from the pack. Mine is cooled via cooling fin and thermal pads and the case becomes hot. The motor will heat up as well but it is probably not a problem. The controller is. I got a cl1000 and at 300 amp it overheat all the time. This with 2 fan on it. You a powerfull controller that you use about half of its rated power for no overheating.

rivvs said:

You a powerfull controller that you use about half of its rated power for no overheating.

Click to expand...

This is usually my philosophy on my builds. Sure there are restrictions: getting a controller that can output twice the power that you actually need costs more, and takes up more space on your build, and is heavier, but for me, longevity is more important. If your priority is speed or performance, your preferences may vary.

rivvs said:

Yes even at the continuous current the heat will build up. I have a. 26s14p and at 300 amp there is heat coming from the pack. Mine is cooled via cooling fin and thermal pads and the case becomes hot. The motor will heat up as well but it is probably not a problem. The controller is. I got a cl1000 and at 300 amp it overheat all the time. This with 2 fan on it. You a powerfull controller that you use about half of its rated power for no overheating.

Click to expand...

Thats a pretty bad review of the CL1000. Its rated for 500 amp continuous and its overheating all the time with 300 amp peak? Thats not an acceptable deviation.

Does anyone have a good review of any 3shul controllers taking peak amps that are somewhat close to what they list as the continuous ability without overheating?

Blacksheep Engineering said:

Does anyone have a good review of any 3shul controllers taking peak amps that are somewhat close to what they list as the continuous ability without overheating?

Click to expand...

Not that I'm aware of. Many people share wild claims about these controllers but so far I've seen nothing very impressive coming from them.
I'm not saying they are bad controllers, but there is very little actual unbiaised feedback so far.

If you want something that is proven to work then you can go for the fardrivers as they are extremely powerful and nowadays many people are using them. I'm using a ND841800 track edition (1800A/phase and 1300A battery), it's definitely a great controller.
I think that the maximum voltage they offer is 108V (the most powerful version is the ND1081800 track edition (1200A battery and 1800A phase @128V maximum which is insane, in theory that's like 150KW).

There is also a 2600A version of the far driver ND962600