Observed trials bike designing

General Discussion about large electric scooters and motorcycles and other things with no pedals.
bikerpete   1 W

1 W
Posts: 58
Joined: Jan 27 2020 9:09pm

Re: Observed trials bike designing

Post by bikerpete » Feb 06 2020 3:00am

A question that has probably been asked many times before, sorry :

If I run a motor rated for 72V on a say 48V battery, what will be the effect down at the low end of the rev range?

I understand that it will have reduced revs at the top end, but I'm not too sure I understand what happens at low revs.
I'm thinking that at low RPM if you twist the throttle to full, the controller will ramp up the voltage until the amps rise to the required level or limit, but due to the effects of back EMF etc. (of which I have a vague big picture understanding, but no clue on the detail) I'm thinking it will probably get full current at somewhat less than rated voltage (in this case 72V)?

I expect it would limit the maximum power and speed available, but that might not bother me 95% of the time. I never use everything my Beta 300 has available up in it's rev range, and have never come close to it's top speed, but I do use all the low down torque in the bottom third of it's range.

I'm really wondering if it's practical to run the QS 3000W 72V motor on lower voltage in this application? Or even the 1000W 72V if I decide on the Fatbike option?

Ohbse   10 kW

10 kW
Posts: 876
Joined: Dec 26 2013 5:15pm
Location: Auckland, New Zealand

Re: Observed trials bike designing

Post by Ohbse » Feb 06 2020 3:10am

Running a motor on a lower voltage will merely reduce its usable RPM range, plus power in the higher portions of that range. Due to the effects of bemf, you will be unable to drive the same current as RPMs increase vs if you had a higher voltage pack. The low down torque will be largely unaffected however, as this is determined by phase amps and the torque constant of the motor. This has nothing to do with voltage. The RPM range over which the controller can drive that motor to that torque level IS affected by your voltage. Hopefully that's clear enough.

For popping the wheel up, 48v would work just fine.

bikerpete   1 W

1 W
Posts: 58
Joined: Jan 27 2020 9:09pm

Re: Observed trials bike designing

Post by bikerpete » Feb 06 2020 3:38am

Ohbse wrote:
Feb 06 2020 3:10am
Running a motor on a lower voltage will merely reduce its usable RPM range, plus power in the higher portions of that range. Due to the effects of bemf, you will be unable to drive the same current as RPMs increase vs if you had a higher voltage pack. The low down torque will be largely unaffected however, as this is determined by phase amps and the torque constant of the motor. This has nothing to do with voltage. The RPM range over which the controller can drive that motor to that torque level IS affected by your voltage. Hopefully that's clear enough.

For popping the wheel up, 48v would work just fine.
Awesome, thank's for confirming that - it might make battery options a bit more flexible.
Ohbse wrote:
Feb 06 2020 1:18am
It's impossible to replicate the functionality of a clutch, having the ability to dump that stored energy near instantly is pretty killer - but I think you could get pretty close with enough of a reduction ratio.

I've got a Nucular 24f, I think it's absolutely fantastic. Votol/Kelly/etc are absolute trash in comparison. Specifically its ability to 500 phase amps into a dead stalled motor reliably is difficult to replicate in any other controller that I'm aware of. You can precisely configure the ramp rate of current increase. By using a brake cutoff switch in place of a clutch, you can pre-position your throttle, release your brake and instantly burst power. It's how I get the front wheel on my Surron up - works a treat. My bike is capable of ~110km/h on 16s. Were it geared down to half that, it would wheelie on command with ease.

If you've got some detailed motor specifications on the QS you can determine what 500 phase amps is going to give you, customise a sprocket configuration to give you the wheel torque required. With the ability to field weaken quite substantially above base speed of the motor, you'll still be able to go a reasonable speed if you need to move around.

Regarding batteries - I think you're way up the wrong tree looking at 18650's. You're looking for power density and instantaneous response, range not a factor. You should really be looking for power dense Lipo packs - you can buy these from the Aussie Hobbyking warehouse, so supply shouldn't be an issue. Voltage is up to you, anything from 16-20s would be perfect when combined with the Nucular 24f. If you're somewhat close to a support vehicle, building a couple of smaller packs and swapping between them would give you a decent weight saving relative to filling it with enough 18650's to get your desired power output.
I think it's impossible to replicate a clutch, full stop. It's not just what happens when you drop it, but also what happens when it's slipping, and when you pull it in. The more I think about what's going on when I ride my ICE bike the more benefits I find to a clutch.
That's not to say everyone needs one, just that there are some things you give up without a clutch.

Batteries have been so hard to work out!
I started thinking I'd need to go with Lipo packs of some sort to get the density, but then it all becomes a bit of a nightmare joining them up, and the affordable ones like Hobby King don't seem to get a good rap for durability - if I'm going to spend $$ & hours putting it together I'd like it to last.
So then it seemed 18650's might not be so bad really, the volume is manageable and they seem to have more options for building the pack. I have a battery welder from a past life.
Then of course the 21700/20700's dropped into the equation and things became even harder for me to decide.

72V, 200A, 20Ah takes a fair few cells and volume however you cut it.

Safety is important to me too. I can deal with the personal safety issues around the batteries, I'm I'm more concerned about the battery going up in flames when I'm out in the dry Australian bush in summer due to component failure, damage etc. The consequences of that simply don't bear contemplating! Think possibly thousands of hectares of commercial pines and native forest, several towns ... I worry about that on my ICE bike too, but at least that's a commercially sold machine that is well accepted, not some DIY job gone wrong! The round cells seem to have a slight edge there, particularly compared to cheap LiPo's of somewhat dubious quality..

Maybe I should just go Hobby King and be done with it!
I'll look into that option more thoroughly.

SlowCo   100 kW

100 kW
Posts: 1416
Joined: Jan 05 2015 5:43pm
Location: The Netherlands

Re: Observed trials bike designing

Post by SlowCo » Feb 06 2020 3:44am

Why not get or build a 72V battry pack and use a lower gearing (mid drive QS motor, not hub?) to be able to use all the power the motor has to offer?

Altair   1 kW

1 kW
Posts: 318
Joined: May 03 2014 10:35am
Location: Quebec, Canada

Re: Observed trials bike designing

Post by Altair » Feb 06 2020 1:07pm

Thanks Pete, I will answer your questions in the order they were asked.
1- The Nucular will certainly be a better option, I have one on order but still 2-3 months to be shipped. The Kelly KLS-S is a sinusoidal controller, and it does a good job on the Astro motor. Very smooth delivery, quiet, but is big. Well I could have ordered the next smaller one but I wasn't sure it would be enough. Finally, it would have been enough, I ended up running 30 bat Amps and 60 phase A. What I don't like about the Kelly is that it has no on-board means of programming. So you need your laptop to make changes in the settings. The Nucular will solve that.
2- I don't see the slipping of a wheel as much of a problem, when you ride slippery terrain, you learn to feather the throttle and keep lots of weight on the rear, and anyway in those situations, you're going pretty slow. If the wheel slips, the fact that you have much less inertia than on an I.C. bike will help you because as soon as you begin to reduce throttle, torque will also reduce immediately! Lack of inertia can also be a benefit sometimes.
3- The Astroflight is a 7 turns. I was going for minimal weight when I decided on the motor. It was an experiment...
4- The Adaptto had an adjustment for linearity of the throttle, called Thr progr., but you could also create your own curve by "cheating" on the Thr linear setting. The Kelly also has a selection of non-linear curves, and I also have an older trapezoidal Kelly that has this feature. You need a curvature only in the low part of the thr. response curve, with the rest being quite linear, but steeper than it originally was. Good idea on using a standard Domino throttle associated with a throttle box. It then would be easy to generate the kind of curve we want, and it could even be adjustable with a pot.
5- Originally, my sprocket carrier was splined directly on the cassette carrier when I was still using the freewheel. Yon can see the splines in the first pic. Then to go direct drive, I machined an adapter that is bolted to the sprocket carrier, and the other end is machined with drum-shaped protrusions that enter the cavities where the pawls normally go in the hub. Sorry, I don't have a pic of this end, I would have to remove it from the bike. It effectively locks the sprocket carrier to the rear hub.

If you have any other questions, please do not hesitate.
Altair
P1070221-1.jpg
P1070221-1.jpg (494.8 KiB) Viewed 1076 times
P1070222-1.jpg
P1070222-1.jpg (489.3 KiB) Viewed 1076 times

thoroughbred   10 W

10 W
Posts: 99
Joined: Apr 11 2018 6:18pm
Location: virginia

Re: Observed trials bike designing

Post by thoroughbred » Feb 06 2020 6:08pm

I measured the space available on my crf150r to install an electric motor onto the engine case. 100 mm diameter is all the room available. I think that limits the selection of motors pretty severely unless you space it far to the left to gain more clearance.

Another option would be to move the motor forward more and drive the clutch basket with a chain, belt or large intermediate gear

bikerpete   1 W

1 W
Posts: 58
Joined: Jan 27 2020 9:09pm

Re: Observed trials bike designing

Post by bikerpete » Feb 07 2020 8:38pm

I'm really heading toward the e-Fat-Trials bike now. A cheaper and perhaps simpler project to cut my teeth on, it sounds a heap of fun and I can take it into local areas that I would never take a normal motorcycle (electric or not).

Current thoughts:
A Fat Bike with fairly short chainstays and top tube. Rear-set pegs to correct CofG, play with riser bars/stem to get a nice cockpit setup. Currently I'm looking at:
Specialized Hellga Comp women's specific FatBike as it has a shorter top tube, low standover height and is available secondhand at present. The Norco Bigfoot looks like a decent value bike that would work, but only from the 2020 model due to seatstay layout on the earlier ones.
QS 1000W 90 motor (reports suggest it can handle up to 3.5Kw). It's 9.5Kg, so no lightweight, but it should be totally robust in this application. 97Kv perhaps according to one post.
Kelly KLS7212S controller (72V, 50A cont., 120A 30sec phase) - I'm not too sure if I should go to the KLS7218S (80/200A) to leave some headroom. It'd be nice to save some $$, weight and volume, but not worth it if it causes frustration when riding. ?
Hobby King Rhino 6s 5000mah 50C packs, 3S2P. A test of these packs said they must be the first cheap Chinese packs to ever meet their C rating! They got a pretty good rap for capacity, output and low sag, and it seems from various reports that they have a good cycle capacity. That would give me 300A at max C rating, and I expect to be closer to 10C. 10Ah/0.66Kwh is rather low, but I expect to just play about within a few hundred meters of my house (I've got a trials park setup) so hopefully that will be enough to start with. I could go to 3P and take it to 1KwH but as part of the idea around the fat bike is to keep costs down this first time around ... It might not be too hard to add more P cells later (not too much later or the packs will be out of whack).
Probably a Bluetooth BMS.
Have to work out the charger - haven't even started on that yet.

I've also still got to work out the drive - I'd like to go super-simple direct chain drive to a really big laser cut rear sprocket, but I very much doubt I'll be able to fit something big enough inside the chainstays. Failing that I'm loosely thinking a first stage chain reduction to the bottom bracket, then on from there. I'll certainly get rid of the freewheel and have a fixed rear sprocket.

For the throttle I'm planning a standard cable throttle twist grip to a cheap throttle box from Aliexpress. I can 3D print new pulleys for the throttle box to create whatever throttle ramp profile I like. Nice and slow down low and much quicker up high is what I'm planning.

I'll put a thumb throttle on the LH bar to use for regen braking to try to extend the tiny battery's range a bit.

Tossing up about a "negative throttle" on the LH lever and a foot brake, or just a LH rear brake with cutout set at fairly hard braking. I find the thought of cutout-on-brake a little scary, I often need to pull the clutch on the ICE bike to stop from flipping it, and I know from painful experience that even a tiny touch of rear brake when coming down off an obstacle can bring the front slamming down hard enough to really hurt the wrists (and that's with suspension). It's also easy to "whiskey-throttle" when things get a bit messy so a LH lever cut-out would be essential in that case. Tricky. Probably try to make the cutoff switch adjustable so I can set it to engage before or after the brake does.

I'd love to put a clutch in there to test it out, but that's a whole other level of complication so it's not likely this time around.
Also like to try a Nucular, but cheap, cheerful and accessible are more where I'm thinking for this build, too much waiting for the Nucular, and more $$$.

Any comments or suggestions gratefully received.

Ohbse   10 kW

10 kW
Posts: 876
Joined: Dec 26 2013 5:15pm
Location: Auckland, New Zealand

Re: Observed trials bike designing

Post by Ohbse » Feb 07 2020 9:17pm

Sounds cool. Not sure you're going to have enough phase amps or response out of the Kelly to snap the front wheel up though, they're known for being underwhelming on low end torque and low speed throttle response.

Altair   1 kW

1 kW
Posts: 318
Joined: May 03 2014 10:35am
Location: Quebec, Canada

Re: Observed trials bike designing

Post by Altair » Feb 07 2020 11:22pm

Pete, I'm using a 1KWH pack and I'm good for 1½ hrs. Depending on your age and weight, I don't see anymore than that needed for a nice trials wring out.
While checking out frames, be sure that they can take a 5.2" tire. There aren't many. Check RSD in Canada. BTW, I ride a Small frame. It's short, hence goes easily over obstacles without catching the skid plate, and it has a low standover and it's got a very good reputation.
Motor- Check LMX, or the Big Block.
Controller- My advise, wait for a Nucular 12F. (Maybe the new 6F will be enough). With Kelly, be aware that the KBS has a slight delay between opening of throttle and motor being powered. Totally unsuitable for trials. Other Kelly maybe.
Charger- With the TinyBMS, you only need a power supply with adjustable voltage and current. I got one from aliexpress. The BMS takes care of everything.
Sprocket- Get a Karting sprocket for #219 chain, I get them from Demon Tweeks in UK. For the motor sprocket, I use a 12T Rotax kart sprocket. I think that LMX has some too. Ratio=8:1 No need for a double stage reduction.
Altair

Altair   1 kW

1 kW
Posts: 318
Joined: May 03 2014 10:35am
Location: Quebec, Canada

Re: Observed trials bike designing

Post by Altair » Feb 07 2020 11:32pm

Rims! Like I said previously, they are Carbonfan 100mm from China, but I forgot to mention that they are tubeless, so no hassle with pinched tubes. The tires "pop" nicely onto the rims, and they allow you to run very low pressures without slipping on the rims. (And tearing the valves as on tubed tires). I run some Orange sealant inside to eliminate the slow leaks.
I even got 2 of those rims here, brand new, that I could sell you. You'll never believe how light they are, and how strong too. Where are you located BTW?

bikerpete   1 W

1 W
Posts: 58
Joined: Jan 27 2020 9:09pm

Re: Observed trials bike designing

Post by bikerpete » Feb 08 2020 12:21am

Altair wrote:
Feb 07 2020 11:22pm
Pete, I'm using a 1KWH pack and I'm good for 1½ hrs. Depending on your age and weight, I don't see anymore than that needed for a nice trials wring out.
While checking out frames, be sure that they can take a 5.2" tire. There aren't many. Check RSD in Canada. BTW, I ride a Small frame. It's short, hence goes easily over obstacles without catching the skid plate, and it has a low standover and it's got a very good reputation.
Motor- Check LMX, or the Big Block.
Controller- My advise, wait for a Nucular 12F. (Maybe the new 6F will be enough). With Kelly, be aware that the KBS has a slight delay between opening of throttle and motor being powered. Totally unsuitable for trials. Other Kelly maybe.
Charger- With the TinyBMS, you only need a power supply with adjustable voltage and current. I got one from aliexpress. The BMS takes care of everything.
Sprocket- Get a Karting sprocket for #219 chain, I get them from Demon Tweeks in UK. For the motor sprocket, I use a 12T Rotax kart sprocket. I think that LMX has some too. Ratio=8:1 No need for a double stage reduction.
Altair
Love the wealth of knowledge!

Sounds like I'd be OK (just) with the 0.66KWh pack if I'm careful. I'm 70Kg and as I say, most of my obstacles are at my back door.
You are using the Kelly KLS on your current bike, is it OK besides the PC programming issue? No throttle delay? I'll probably heed your advice on the Nucular, but tempted to charge ahead if the Kelly isn't too obnoxious.

I'm a bit confused with the Big Blocks - it seems it's a generic term for a whole bucket full of motors of widely varying quality/specs?
The LMX at EU280 + freight is getting a bit expensive. The BHT is nearly EU100 less and I saw a good write up on it. Turning to Aliexpress et al it seems pretty hit & miss. I'll take a bit more of a look though.

Fat bike supply in Australia is very limited, and importing can get expensive fast. The Hellga I'm looking at can take 5" tyres, possibly no bigger. It's a medium, but womens so shorter and low standover. If the budget for this starts blowing out I'd prefer to spend the money on the full moto-trials bike.

Tubeless - I've done a fair number of ghetto tubeless conversions to non-tubeless rims and can usually get a decent bead-up. Sometimes it takes a bit of finagling with layers of tape to create a tubeless bead seat, but it eventually works (usually). I've even resorted to gluing the bead to the rim successfully. I looked at the Carbonfan rims - nice, but look out of my budget I'm afraid.

BMS and sprockets - Roger.

Now if my thumb gets better - endoed over the bars doing a stoppie turn this afternoon and my thumb is feeling distinctly worse for wear - not convinced I haven't broken something in there.

Altair   1 kW

1 kW
Posts: 318
Joined: May 03 2014 10:35am
Location: Quebec, Canada

Re: Observed trials bike designing

Post by Altair » Feb 08 2020 1:39pm

I run the Kelly KLS-S which a sinewave controller. It can run the Astro pretty well but I'm not sure it would be useable with the usual motors. You might go with one of their other non-sinusoidal models instead.
My KLS-S doesn't have the throttle delay defect, but I don't know which other one would be similar.
The real Big Block was here: https://lightning-rods10.mybigcommerce.com/mid-drives/
but they don't seem to sell just the motors anymore.
The BHT they sell now is not at all the same (electrically) they had originally, and the shaft is too small and will break.
Get that thumb checked if you don't want it to limit you for the rest of your life...

bikerpete   1 W

1 W
Posts: 58
Joined: Jan 27 2020 9:09pm

Re: Observed trials bike designing

Post by bikerpete » Feb 11 2020 9:38pm

Still trying to find the right Fat Bike to base this all on, but this is a layout based around one candidate (Salsa Bucksaw). I really like the idea of suspension, but it tends to limit the maximum tyre size - I think this bike can squeeze to a 4.5" depending on the tyre.
BS 01 layout.jpg
BS 01 layout.jpg (110.62 KiB) Viewed 914 times
Image

Cyclone 3Kw motor
Micro Nucular 6F controller
Hobby King Rhino 4S 6.2Ah batteries as 16S2P - these are really tight in this frame but there are a couple of options. Building the case will be interesting!
Tiny BMS
BMX zero offset handlebar stem to get the bars back and up to the right place.

I've found the Kawasaki KX 60/65 clutches are the smallest I can find - 110mm diameter and conveniently have the release pushing from the outer end rather than through the input shaft as some do. A bit of work to make a fork and bearing arrangement to work the release, but it seems doable. I don't have one in my hands yet, so still very much concept only. Looks like I might be building a housing and release mechanism so I can have a clutch on the eFat Trial bike after all. I haven't modeled this yet.

Currently planning around removing the planetary gears and replacing with the clutch. The clutch will go straight onto the shaft so the small (10mm) diameter shaft wont need to cope with significant bending moments. Exactly how I achieve mounting the clutch and getting the output shaft connected is very vague until I have the clutch in my hands. Drive will be 219 chain to a 99T rear, front will be whatever works - might start with 11 and go up or down from there.

With this rear suspension design the chain line needs to go though or ideally slightly below the pivot point, with the big rear sprocket this drops the motor lower than I'd like, but doable with a decent bash guard I think. I could put an idler in there, but that's a bit clunky. If the chain is running through the pivot axis (no idler) then when the rear compresses it will move above and cause the rear end to squat. This is actually how moto trials bikes are setup (with chainline above pivot) and I suspect it contributes to the pop you get by dropping the throttle as you dump the clutch. Might well be dreaming.

Looking forward to getting the clutch in my hands to see how it might work!

Altair   1 kW

1 kW
Posts: 318
Joined: May 03 2014 10:35am
Location: Quebec, Canada

Re: Observed trials bike designing

Post by Altair » Feb 12 2020 9:02am

This project looks real good!
To limit costs and time, you could try it without the clutch first, you might actually like it, the acceleration is not bad at all. And it's adjustable through software anyway.
And by going with suspension, you won't need the widest tires, so the limit of 4.5" won't be a detriment.
36 spoke hubs and rims will be an asset if you can find them.
Please keep us informed!

User avatar
DanGT86   10 kW

10 kW
Posts: 730
Joined: Sep 06 2012 9:18pm
Location: Saint Louis MO

Re: Observed trials bike designing

Post by DanGT86 » Feb 12 2020 10:20am

Seems like quite a divergence from the original plan but still looks like it would be fun to ride. I agree with Altair about eliminating the clutch for the first try. When you were trying to convert a full weight moto trials bike it made sense. You are so far away from that with a mountain bike that there is no need for it. A 4kw bike geared for 35mph will loop you out from a dead stop. On a bicycle there is just no way you are going to need anymore instantaneous torque than what a motor and throttle can provide. You'd destroy parts or mangle the frame before you'd need the clutch IMO.

If you are looking at the cyclone 3k motor you should check out the LMX 30mm motor. They might still be on sale. Mine came with a 12tooth 219 sprocket and with a 94 tooth rear on 80v it goes about 30mph with tons of torque. I like that its 45kv because you can get to reasonable speeds in only 1 stage of reduction.

Do you use Linkage MTB design software? Its a nice tool for calculating suspension action. The full version is $25.

Maybe take a look at the NS Soda Slope frame. It has a handy bottom bracket concentric rear pivot. That might make thinks a little easier with the chain tension as it wont grow throughout the suspension travel if you use the bottom bracket as a jackshaft.

The FUTR beta frame might also be a good candidate to build on.

bikerpete   1 W

1 W
Posts: 58
Joined: Jan 27 2020 9:09pm

Re: Observed trials bike designing

Post by bikerpete » Feb 12 2020 8:51pm

Altair wrote:
Feb 12 2020 9:02am
This project looks real good!
To limit costs and time, you could try it without the clutch first, you might actually like it, the acceleration is not bad at all. And it's adjustable through software anyway.
And by going with suspension, you won't need the widest tires, so the limit of 4.5" won't be a detriment.
36 spoke hubs and rims will be an asset if you can find them.
Please keep us informed!
Your voice of experience is encouraging! I have been a bit concerned about the 4-4.5" tyre limit but this morning decided it should be OK so pulled the trigger on the Salsa Bucksaw. Glad to now read you feel the same. Now just have to get it over the 1,000km to me!
I definitely agree on the 36h wheel. I figure I'll ride it as-is until either I find something suitable for cheap or blow something up and have to make the change.
DanGT86 wrote:
Feb 12 2020 10:20am
Seems like quite a divergence from the original plan but still looks like it would be fun to ride. I agree with Altair about eliminating the clutch for the first try. When you were trying to convert a full weight moto trials bike it made sense. You are so far away from that with a mountain bike that there is no need for it. A 4kw bike geared for 35mph will loop you out from a dead stop. On a bicycle there is just no way you are going to need anymore instantaneous torque than what a motor and throttle can provide. You'd destroy parts or mangle the frame before you'd need the clutch IMO.

If you are looking at the cyclone 3k motor you should check out the LMX 30mm motor. They might still be on sale. Mine came with a 12tooth 219 sprocket and with a 94 tooth rear on 80v it goes about 30mph with tons of torque. I like that its 45kv because you can get to reasonable speeds in only 1 stage of reduction.

Do you use Linkage MTB design software? Its a nice tool for calculating suspension action. The full version is $25.

Maybe take a look at the NS Soda Slope frame. It has a handy bottom bracket concentric rear pivot. That might make thinks a little easier with the chain tension as it wont grow throughout the suspension travel if you use the bottom bracket as a jackshaft.

The FUTR beta frame might also be a good candidate to build on.
Definitely a divergence! I'm kind of looking at it as a lower cost step in to e-trials before going full-size, and one that might be easily as much fun as full-size anyway.

I'm certainly thinking about running it clutchless first (or with an e-clutch), but adding a clutch later will be a pretty significant re-build that to be honest I'm unlikely to do knowing me as I do!
The clutch has several purposes:
* Experiment with building a clutch onto a motor
* Get a feel for how a small MC clutch running direct drive off the motor works (an MC has a significant reduction from crankshaft to clutch).
* As important as being able to instantly apply power, the clutch (I think) is just as important for instantly cutting power.
* operating a lever with one finger is more stable/controllable/accurate than operating a twist grip.
* potentially a clutch might allow the motor to run cooler and more efficiently, spinning a bit rather than spending so much time stalled.
* I don't see it as necessarily providing more torque, just applying torque on/off more instantly.
* I'm finding it hard to imagine riding a bike with the throttle set to zero ramp (to achieve clutch like torque response) - it would be impossible to ride smoothly. But with ramp then it's impossible to snap on torque - even if it's only modest levels of torque. Happy to be educated that this is incorrect. Maybe it would be OK if I set the controller to zero ramp, and rely on my ramped cam shape pulley on the throttle box to achieve a controllable throttle. No idea how that would work out in real life, I suspect still a bit too brutal.

The acid test to me is watching bikes such as the Electric Motion 5.7 (electronic "clutch") next to eTrials bikes with "real" clutches. They always ride quite differently. Not badly, but definitely different and not the way I generally try to ride. Certainly I've been well and truly out-ridden on my Beta by riders on EM 5.7's, so it's not that they can't perform better than I'll ever be able to, it's a quality vs quantity thing.
I accept that I'm very likely to break the bike if I get enthusiastic with the clutch. I'll try very hard to restrain myself, and keep power levels down when using it!

The LMX 30mm looks nice and the larger shaft would make life a little simpler. Sadly the new model isn't on sale so it gets pretty pricey by the time it lands in Oz. Also possible compatibility issues with Nucular as the site says "Not working with sine wave controllers". A quick read around and it sounds like that might be fixable by moving the halls, but certainly an added disincentive.

Chain length - certainly an issue to be dealt with, but it works OK on motos with a spring tensioner, so hopefully it'll pan out OK on this one.

Bike purchased.
Clutch purchased.
Now to start ordering the electronics.

User avatar
DanGT86   10 kW

10 kW
Posts: 730
Joined: Sep 06 2012 9:18pm
Location: Saint Louis MO

Re: Observed trials bike designing

Post by DanGT86 » Feb 12 2020 11:26pm

Clutch or not I think you will be pleasantly surprised by the abundance of smooth controllable torque an electric motor has to offer.

Just fyi, the lmx seems to get along nicely with the nucular. Here is some amazing work with the lmx and nucular!

viewtopic.php?f=6&t=103637

Congrats on ordering parts and a bike! Looking forward to seeing what it turns into.

bikerpete   1 W

1 W
Posts: 58
Joined: Jan 27 2020 9:09pm

Re: Observed trials bike designing

Post by bikerpete » Feb 13 2020 12:47am

DanGT86 wrote:
Feb 12 2020 11:26pm
Clutch or not I think you will be pleasantly surprised by the abundance of smooth controllable torque an electric motor has to offer.

Just fyi, the lmx seems to get along nicely with the nucular. Here is some amazing work with the lmx and nucular!

viewtopic.php?f=6&t=103637

Congrats on ordering parts and a bike! Looking forward to seeing what it turns into.
I saw that build - amazing job!
Given the work they did on the motor I'm not sure it's really typical of an LMX. Looking through the LMX thread it's a bit hard to tell if that's the new LMX or the old. LMX said it was the old, but the guys who built it said it was definitely the new one.
I've sent LMX a message asking if the old one can be made to work on a Nucular, and if so what might need doing. If it's reasonably trivial I'll definitely seriously consider it. Also asked them how much difference there might be between the two versions in this particular application (low speed, low RPM, high torque).

Just when I was ready to start ordering electronics, I see Nucular are getting ready to release a new 6F controller:
VasiliSk wrote:
Jan 16 2020 3:08am
No much, been working on new 6F version. after launch will switch to bms
Patience is a virtue they say.

Whatever happens, I'm determined to be pleasantly surprised - sure beats being disappointed :D

thoroughbred   10 W

10 W
Posts: 99
Joined: Apr 11 2018 6:18pm
Location: virginia

Re: Observed trials bike designing

Post by thoroughbred » Feb 13 2020 6:49am

I'm a yes vote in favor of the clutch. :thumb: :thumb: :thumb:

If it works it will multiply the number of controllers that you can choose from. No need to wait for expensive unobtainable ones.

bikerpete   1 W

1 W
Posts: 58
Joined: Jan 27 2020 9:09pm

Re: Observed trials bike designing

Post by bikerpete » Feb 14 2020 6:33pm

Altair wrote:
Feb 07 2020 11:22pm
Sprocket- Get a Karting sprocket for #219 chain, I get them from Demon Tweeks in UK. For the motor sprocket, I use a 12T Rotax kart sprocket. I think that LMX has some too. Ratio=8:1 No need for a double stage reduction.
Altair
Guy, I was about to order an LMX motor (I think) but got a bit concerned about the reduction ratio.

Your current setup (AstroFlight 7T at 51.8V, 5:1 primary reduction, 8:1 final drive = total 40:1) looks like it runs a theoretical max no load wheel RPM of 251, top speed 36kmh (no load, based on my best approximation of the tyre OD).

An LMX 32mm motor will give me 3552 motor RPM. An 8:1 final drive will give a no load wheel RPM of 430, top speed of 61kmh (no load). That's nearly double what you appear to run, and getting up around what the enduro style riders are doing on this motor - I can't imagine it working for trials without killing the motor!

If I put 9/99 sprockets on the LMX (smallest & biggest I can find) it still gives me 46kmh, 1.27 times your top speed.
9/99 isn't a good option for a bunch of reasons.

On your BHT setup you said you were running 8.75:1 and I think the BHT is fairly similar to the LMX (?). Was it ideal ratio for trials work?

I'm perfectly happy pottering along at 30-40kmh (on my Beta when I'm going "fast" on fire roads I am at all of 11-1200 rpm, idle at 950) and would sacrifice any amount of top end speed for nice trials performance.

So it seems to me I should be aiming closer to 15:1 overall reduction to get the LMX working nicely down at slow speeds. 15:1 would give me a top speed around 34kmh (which is fine by me). Does that sound sort of reasonable based on your experience?

The only way I can get this kind of reduction is with 2-stages. If I rework my clutch layout back into a standard motorcycle one I could use it's reduction as the first stage.
The LMX becomes much less attractive if I don't use the clutch and reduction as I'll need to build an unwanted idler stage. This would make the geared Cyclone a far better option.

It seems the Cyclone would be best with the direct drive, the LMX best with the clutch (larger output shaft).

I had hoped to put the clutch directly on the motor shaft, but that seems like it's becoming unlikely now. Cyclone+Planetary+clutch is just too long, and direct off either motor is looking too fast.

EDIT:
I just found this video of "Trials style" on an LMX bike. I think these have 7.5 or 7.8:1 reduction.
No way would I even vaguely consider this "trials" riding! Nor does it sound like the motor would have a hope of performing adequately at the sort of speeds/manoeuvres I'm intending - way too big gearing.

Altair   1 kW

1 kW
Posts: 318
Joined: May 03 2014 10:35am
Location: Quebec, Canada

Re: Observed trials bike designing

Post by Altair » Feb 14 2020 9:06pm

That video cannot be called "trials", obviously.
One thing to keep in mind is that you will never attain the theoretical no-load RPM on the bike. Today, it seems that all we see are low pole count motors, because theoretically you can go faster with them. But for trials, I would much prefer something like 18slots/16poles because it will make the power delivery much smoother,(with smaller cogging steps) and when you want to go fast, a modern controller like the Nucular will be able to produce the ERPM necessary to turn the motor at high RPMs. One other advantage of a higher number of poles is that the torque will be higher, for the same motor diameter & length.

The original BHT was a good combination at 8.75:1 ratio, and it was "only" a 18slots/12poles. This gives 36 steps per revolution. The LMX has 12s/8p, which has 24 steps per revolution. A 18s/16p would have 144 steps per revolution, much smoother. Refering to the video of the Electric Motion EM 5.7, you can see the result of a low pole (and slots) count. At slow speed, you can hear the motor go tock-tock-tock... so, not ideal.
I've never tried the LMX, so I can't comment on it. But I think that for trials, a "slower" motor (higher # of slots & poles would be advantageous. But those motors would be more expensive to make, so that's why we don't see them.
I don't know either about the cyclone, the integrated gearbox would compensate for the low pole count, I assume. But without its gearbox, ???

For a #219 chain, I wouldn't go below 12 teeth, if I want a good sprocket & chain life. Yes I ran 11t on my first E-fatbike but then you become also limited in the diameter of the motor shaft, introducing reliability issues. 12t allows a shaft diameter of 19mm, which is pretty beefy, even if it is a hollow shaft to save weight.
However go-karts that ran the Yamaha KT-100 direct-drive had 11t on the engine. That was quite extreme, because of the very small radius the chain had to make around the 11T sprocket, and the RPMs going between 12,000 and 15,000. Needless to say, no oil or grease was able to stay on that chain for more than a minute on the track. We changed sprockets often, they were wearing fast. Another advantage of a larger sprocket will be a reduction on chain noise.

So ideally, I would go with a motor with 18slots/16 poles, sprocket of 14 to 16 teeth, and a rear sprocket of around 100 teeth. Note that an advantage of a larger sprockets is that the pulling force on the frame will be smaller than if the same reduction ratio was accomplished with smaller sprockets. When running bicycle frames instead of real motorcycle frames, this becomes important for the lifespan of the frame.

bikerpete   1 W

1 W
Posts: 58
Joined: Jan 27 2020 9:09pm

Re: Observed trials bike designing

Post by bikerpete » Feb 14 2020 10:52pm

Thanks Guy.
I'd still be interested to hear if you think my calculations of your bike's top speeds was about in the ballpark? Pretty rubbery I know, but at least it gives me a bit of a feel for things.
Altair wrote:
Feb 14 2020 9:06pm
One thing to keep in mind is that you will never attain the theoretical no-load RPM on the bike.
Yes, just used that figure for "apples to apples".
If it does 20kmh in real life I'll be happy enough.
Altair wrote:
Feb 14 2020 9:06pm
The original BHT was a good combination at 8.75:1 ratio, and it was "only" a 18slots/12poles. This gives 36 steps per revolution. The LMX has 12s/8p, which has 24 steps per revolution.

I don't know either about the cyclone, the integrated gearbox would compensate for the low pole count, I assume. But without its gearbox, ???
The pole count issue makes me think things are looking even worse for the LMX direct drive than I thought!

So it's either the clutch driven as designed in the original motorcycle or the geared Cyclone.
Or find another motor, which seems unlikely without going way over budget.
Altair wrote:
Feb 14 2020 9:06pm
So ideally, I would go with a motor with 18slots/16 poles, sprocket of 14 to 16 teeth, and a rear sprocket of around 100 teeth. Note that an advantage of a larger sprockets is that the pulling force on the frame will be smaller than if the same reduction ratio was accomplished with smaller sprockets. When running bicycle frames instead of real motorcycle frames, this becomes important for the lifespan of the frame.
Cyclone and LMX are both 12 slot.
Not sure about the QS 1000w 90 70H - all I can find so far is "Magnet Height: 80mm, 5 pole pairs" so perhaps that means it 10 pole? It's certainly a possibility for this build, but it's length is a big disadvantage, weight also and if it's 10 pole then yet another.

To an extent I'm trying to balance sprocket size/ratios/driveline.
A big rear sprocket is good as you say, but a smaller one improves the motor position whilst keeping the chain alignment with the suspension pivot. I'm willing to compromise more on alignment, but am concerned if the chain gets too far above the pivot the bike will squat badly under power with soft suspension settings.

Running the clutch as a primary reduction might help solve a bunch of issues:
  • easy to get high reduction ratios.
  • easier to build a hydraulic clutch release mechanism. This was looking somewhat complex with direct mount clutch.
  • allows for an outboard support bearing on the motor shaft, so shaft dia is less critical. (Cyclone easier to use)
  • possibly improve ground clearance whilst maintaining good chain/suspension alignment
  • allows for slipping the clutch at low speeds/stall to compensate for cogging/pole count.
  • Might make sealing the clutch housing for oil easier perhaps.

User avatar
DanGT86   10 kW

10 kW
Posts: 730
Joined: Sep 06 2012 9:18pm
Location: Saint Louis MO

Re: Observed trials bike designing

Post by DanGT86 » Feb 14 2020 11:17pm

My lmx 30mm is 45kv. It came with a 12 tooth 219 sprocket. I run a 92 or 94 tooth rear on a 24" wheel on 20s so 75v nominal. This gearing does 48-50kph and has enough torque to wheelie on a long wheelbase bike when set to 4kw max.

If a human rider on a trials bicycle can side hop over a meter with human legs why wouldn't your motor be able to do the same? Even if its geared a little higher than you prefer it's still stronger than human legs. Once there is enough torque to flip the bike over then you aren't using the rest anyway. Once you add in a clutch and flywheel like you want it's going to be even better because you aren't stalled at zero rpm.

Regarding sprocket sizes the only way to get durability on a bike frame with bike chain or 219 chain is to have a large reduction. When you get into smaller dia rear sprockets and less reduction the force on the chain gets crazy. Look how beefy mx chains are. Even 219 needs to spin fast to work at high power.

If you must use a clutch and need more reduction then I think it would be clever to use the reduction on the clutch gear like a gas bike. That would provide a large reduction and really help get the motor spinning higher rpm.

The new lightning rods motor is 67kv and has an ipm rotor design like the qs3000. That should make it respond well to field weakening to get the high speed for cruising without sacrificing down low. Probably be an awesome match for the nucular at 8000rpm if you end up with 2stage reduction.

bikerpete   1 W

1 W
Posts: 58
Joined: Jan 27 2020 9:09pm

Re: Observed trials bike designing

Post by bikerpete » Feb 15 2020 12:13am

DanGT86 wrote:
Feb 14 2020 11:17pm
1. If a human rider on a trials bicycle can side hop over a meter with human legs why wouldn't your motor be able to do the same?
...
2. Once there is enough torque to flip the bike over then you aren't using the rest anyway.
...
3. That would provide a large reduction and really help get the motor spinning higher rpm.

4. The new lightning rods motor is 67kv and has an ipm rotor design like the qs3000. That should make it respond well to field weakening to get the high speed for cruising without sacrificing down low. Probably be an awesome match for the nucular at 8000rpm if you end up with 2stage reduction.
Numbering added for ease.

1. Because it doesn't have the skills, and nor does the rider! :lol:

2. Not quite true. Wheelying has almost no relation to the back wheel moves on a moto trials bike.
I can wheelie my Beta 300 from close to idle in 4th gear no problems by just rolling on throttle.
However I also rev it to a few thousand rpm (it's got a vastly bigger flywheel than an enduro or MX bike) stationary in 1st, 2nd or 3rd gear, dump the clutch and don't loop out. And I can do that on coarse granite with soft sticky trials tyres at 3psi - no wheelspin at all. And I'm not even a good rider! Watch the gaps that Toni Bou and those guys do from a standstill! Waaaay more power than needed to flip the bike.
Of course doing something like that on a bicycle frame and running gear would end badly!

3. That's true and is what I think I'll need to do. To retain a big rear sprocket to keep chain forces down I'll need to run a big front sprocket - eg to get a 15:1 overall reduction with the clutch primary reduction in the train, with a 90t rear I'd need a 21 tooth front. Looks like I might be making my own sprocket!

4. LR's motor seems nice but on checkout: "Unfortunately one or more items in your cart can't be shipped to your location. Please choose a different delivery address." I might be able to get past that, but shipping to Oz from USA is poisonously expensive so I'd expect that motor to end up around a cool AU$720 by the time I've paid shipping, exchange & taxes! Well out of my budget. The tyranny of distance.

Altair   1 kW

1 kW
Posts: 318
Joined: May 03 2014 10:35am
Location: Quebec, Canada

Re: Observed trials bike designing

Post by Altair » Feb 15 2020 8:28am

Yes Pete, your calculations for speed were correct.
The QS 90 70H with 10 poles has very probably 12 slots, which will give 60 steps per revolution. Not bad. But it is indeed long.
BTW, the torque of a motor goes linearly with its length, but goes with the square of the diameter, so for the same volume, the narrow & large diameter one will have more torque than the long one with small dia.

Post Reply