Brushless dual 6kw ALIEN Power system - electric longboard

[lone_deranger]

Hi beetbocks et al,
I have been avidly following the development of your design since the beginning and am looking forward to starting my own Alien build soon.
As regards wheels, 75mm is the recommended starting point in the distance skate community. The Seismic 85mm 79a Speed Vent in Black Ops urethane is widely regarded as the fastest, it is also the most grippy of the large wheels due to its slalom-based lip profile and as far as I'm aware the most chosen wheel for cross-continental skates. As far as comfort is concerned, it has mostly to do with the durometer and depth of urethane. 75mm Avilas (especially the 72a Purple Skunks) are regarded as the plushest; they have minimal core and a huge lip. I now ride 75a 97mm Flywheels on my main setup and love their ability to roll over chip seal, twigs and small children but do find the back can break loose if pumping hard on damp surfaces. The extra urethane depth over the 85mm Speed Vents is also a blessing on the feet when it comes to vibration dampening. Nearly bought a set of 107mm eFlys but the idea of pushing that amount of weight up a hill scared me off. Personally I find the urethane depth on 76mm and 83mm Flywheels too thin for comfort due to their large core. 90mm Flys may prove to be the best compromise of urethane depth and weight.
I would love a pulley that would be compatible with my 97mm Flys by the way...

Obvious disclaimer; personal findings, YMMV

 

[simonjook]

I have been rethinking the merits of a dual motor. Occasionally I get a loss of sync. Usually it's because I hit a section of pebbles meaning a lot of rapid acceleration and deceleration for the motor (free-spin/grip/free-spin etc on the marbles). If I am under hard acceleration this quickly turns into hard breaking as the motor gets out of sync which in turn results in being hurled down the road ... or onto it. Now I am wondering with two motors whether the drive of the other motor counters this somewhat.

Can someone with a dual setup comment on whether you get loss of sync and what happens? It usually makes a distinct sound so if you've ridden on without incident just a slight loss of power perhaps you've heard it. Of course maybe all the duallies are using quality controllers and don't have any such issues or maybe the sync loss results in the whole ESC getting confused and stalling both motors.

 

[rf]

I have been rethinking the merits of a dual motor. Occasionally I get a loss of sync. Usually it's because I hit a section of pebbles meaning a lot of rapid acceleration and deceleration for the motor (free-spin/grip/free-spin etc on the marbles). If I am under hard acceleration this quickly turns into hard breaking as the motor gets out of sync which in turn results in being hurled down the road ... or onto it. ...

Perhaps you have a bad connection somewhere. Brushless motors shouldn't have a problem at speed, unless the timing is close to the edge.

Otherwise, it might be helpful to try a different ESC.

 

[simonjook]

Perhaps you have a bad connection somewhere. Brushless motors shouldn't have a problem at speed, unless the timing is close to the edge.
Otherwise, it might be helpful to try a different ESC.

Well it is close to the edge - when you get break and bite with at 90+kg of mass at 15+km/h (motor spins up probably close to 25 than immediately back to 15) it causes quite a load and shock and the current flies up spiking well over the 150A the ESC is rated at... so it is close to the edge. In my experience such events often confuse sensorless controllers but this is also only a budget HK150A so I don't expect too much from it. Please remember though these events are rare and there's a good chance I am pushing my board a lot harder than many others. The main reason I passed up an Evolve was the guys didn;t think it would be up to it considering where I live and I climb hills my 1000W Fiik couldn't climb.

 

[rf]

Perhaps you have a bad connection somewhere. Brushless motors shouldn't have a problem at speed, unless the timing is close to the edge.
Otherwise, it might be helpful to try a different ESC.

Well it is close to the edge - when you get break and bite with at 90+kg of mass at 15+km/h (motor spins up probably close to 25 than immediately back to 15) it causes quite a load and shock and the current flies up spiking well over the 150A the ESC is rated at... so it is close to the edge. In my experience such events often confuse sensorless controllers but this is also only a budget HK150A so I don't expect too much from it. Please remember though these events are rare and there's a good chance I am pushing my board a lot harder than many others. The main reason I passed up an Evolve was the guys didn;t think it would be up to it considering where I live and I climb hills my 1000W Fiik couldn't climb.

Just looked at the Evolve page again ... those guys only know flatland. And their boards are underpowered. Like my nasty old Exkate. But buying a board ready-made is just plain wrong anyway!

150amps does seem marginal. R/C companies rate their stuff extremely close to the edge. A 100amp ESC for a 100amp motor is just nuts -- buts that's what they use. When you're upping the load by a few thousand percent (and carrying a human), reserve becomes way more important.

In my limited experience, brushless motors WITH sensors and good, COMPATIBLE ESCs are solid, from zero to top RPM. My two direct-drive ebike hub motors run smoothly at fractional rpms (standing stop) to full speed with NO glitches. Disconnect their sensors (they are designed to run either way) and they don't start nicely and must be helped to a couple MPH. But run very smoothly after that. Anyone that has a sensored brushless motor that isn't solid has a broken one. I'm not ready to give up on sensorless. But maybe sensored systems make more sense. (Yikes.)

The R/C market has a number of very different motors. Inrunners, outrunners, high-RPM, low-RPM, plane, car, boat. Seems like you COULD make an ESC that would accommodate them all -- but I'm not so sure. Outrunners seem mostly targetted at planes, which have very simple throttle control needs. Not so much cars and other wheeled models that need much finer control AND BRAKES.

Stuttering like you're experiencing might never be noticed by R/C users. Things like that need to be accounted for when you've got a human on top. Design should take on a whole different attitude. Simple things like nuts and bolts work the same. Brushless motors, ESCs, controllers, etc., are not so simple. How should they respond when errors occur? Not nearly as important with a model car.

As an ESC, how do you respond when you lose track of what's going on? Reset? Start over? With R/C -- probably. With human on board -- UNACCEPTABLE. What happens if you lose power and then it comes back? Bad connection or something. The ESC starts over at zero miles per hour? Not too good for the human currently traveling at 20mph. ESC design needs to take this into account. And much more.

Losing power is one thing. Doing the wrong thing afterwards could be a much bigger problem.

Tools like an IR thermometer and a recording ammeter (R/C vendors have them) might help insure you're not going beyond your hardware's limits. Considering the state of R/C equipment and the massive stress we're putting on things, I wouldn't build an eskateboard without them.

Current carrying ability of an ESC can be increased some with better cooling. Bigger heat sinks. R/C guys have those too, and bigger ones are available elsewhere. Plenty of room on a skateboard for a little more heat-sink. (Or get an aluminum deck and use the whole thing.

Still waiting for my parts to arrive. :-(. A 50-60 270KV Propdrive (max 90amps), and a 200amp ESC which I hope will be compatible with the open-source SimonK ESC software. Then I'll be able to create my very own face-plants! And perhaps create an ESC more suitable for carrying humans.

What we're doing is a reach. Be careful.

 

[yelnatsch517]

How much more complicated would a sensored setup be as compared to a non-sensored setup?
It sounds like a sensored motor could increase torque and reduce power consumption, which is optimal for an e-board.
How accurate are the hall effect sensor kits, ie. http://endless-sphere.com/forums/viewtopic.php?f=35&t=48770?

Also, there is a project on endless where someone used a skateboard bearing as a tensioner.
I think that would work wonderfully as it significantly increases teeth in the mesh. Only thing I wonder is how much torque is loss.
I'm guessing it shouldn't be too much as skateboard bearings are fairly efficient.

 

[rf]

simonjook, Thinking about your suggestion that two motors provide more than just power advantage. So true. Something I'm often harping about with microprocessor use. The unreliability factor, due to complexity. Processors crash, have bugs, etc., for too many reasons.

Processors in key systems in cars and planes are always backed up with redundant systems. Airliner redundancy is often four and six deep.

Eskateboard wheel-spin is much less likely to dump the rider if you have another powered wheel smoothing things out. Too bad it doubles the cost of the drive system.

 

[simonjook]

Just looked at the Evolve page again ... those guys only know flatland. And their boards are underpowered. Like my nasty old Exkate. But buying a board ready-made is just plain wrong anyway!

Not true. The 200W is a misdirection - there is no way when you are hammering the throttle it's drawing 200W - felt more like a 800+ bushed motor. It's fast and has loads of torque right from zero... but it's not geared as a hill climber.

The HK150A is rated at 1080A peak (hahaha - but that's the claim). The problem is I can't think of another use where you would get the sudden break of load and then reapplication of it in such shirt spurts so its no wonder I get an exception in the behaviour - certainly not on an RC car (much lower load) nor ebike (impossible to load and unload the wheel so rapidly). The motor and ESC stay reasonably cool and the current does get about the max sustained but not the claimed peak. I think it's just timing confusion because of the extremely rapid loading and unloading at high throttle position. Probably CPU can't respond fast enough or just software way to crummy.

 

[rf]

...
The HK150A is rated at 1080A peak (hahaha - but that's the claim).

That's the typo. Just looked at that web page.

-- Cont./Burst Current: 150A/1080A
most likely meant to be:
-- Cont./Burst Current: 150A/180A

The problem is I can't think of another use where you would get the sudden break of load and then reapplication of it in such short spurts so its no wonder I get an exception in the behaviour - certainly not on an RC car (much lower load) nor ebike (impossible to load and unload the wheel so rapidly). The motor and ESC stay reasonably cool and the current does get about the max sustained but not the claimed peak. I think it's just timing confusion because of the extremely rapid loading and unloading at high throttle position. Probably CPU can't respond fast enough or just software way to crummy.

I think that CPU is plenty fast. That ESC can drive MUCH faster motors. As you suggest, your circumstance is so far away from what the programmer expected that he assumed the proper thing to do (for rapid RPM change) was to do a reset. ESCs measure the timing of every rotation. It didn't lose track of where it was, but it wasn't taking any chances.

Cheap ESCs are constantly at risk of entering conditions that will destroy their MOSFETs. So anything beyond 'reasonable' is considered grounds for a reset. Just to be safe.

If we had the source code to that ESC we could probably fix the problem without too much trouble. :-/ Perhaps the SimonK code already responds more reasonably. Wouldn't be surprised. Don't know if the HK150A has an Atmel processor or not.

 

[simonjook]

Unfortunately its certainly not listed as SimonK compatible. I will try advancing the timing one more step this wek to see what it does. I'm pretty sure I've seen peaks in excess of 180A and I'm not sure the 1080A is a typo. A number of the competitor designs (Hobbywing etc) have similar burst specs. Whether true or not 1080A is feasible though the bursts would need to be extremely short.

 

[simonjook]

Another thought - because I have issues on high load unload cycles I might just need to add some more caps to combat a ripple that might be forming. Now the pain I can't get good caps locally :|

 

[yelnatsch517]

I agree, the ESC processors can handle much higher frequencies, 10+ times higher.
Can you get a vid of the issues you are having? What brake setting do you have programmed on your ESC?

 

[simonjook]

I agree, the ESC processors can handle much higher frequencies, 10+ times higher.
Can you get a vid of the issues you are having? What brake setting do you have programmed on your ESC?

It happens maybe once or twice a week (I ride twice a day) so it's not that frequent and reproducible. It also nearly ends with my face on the pavement even when I suspect it will happen so not super keen on trying to reproduce it.

 

[Damgaard]

The new alien board will be 2.5kw power and about 4kg with range of 8\10 miles, depending on the road structure.

@beetbocks, @brunotolli
Wow, a 4 kg eboard sounds cool! Do you guys have an idea of how cheap you can make it, and when it'll be available?

 

[beetbocks]

Hi Damgaard
It will hopefully be in the next couple of months- i suspect..

 

[rf]

The new alien board will be 2.5kw power and about 4kg with range of 8\10 miles, depending on the road structure.

@beetbocks, @brunotolli
Wow, a 4 kg eboard sounds cool! Do you guys have an idea of how cheap you can make it, and when it'll be available?

Very cool.

It might be instructive to see an accounting of the weight of each subsystem. Deck, motor(s), motor mount(s), trucks & bearings, wheels, batteries, electronics. For a better understanding of the target areas.

How light can we go? Pocket-boards, here we come!

 

[simonjook]

You so dont want to ride an e-pocketboard unless you have a lot of padding on your rear end.

4kg is pretty damn light - my Arbor longboard (a pretty average board) weighs a little over 3kg complete so 4kg would feel like carrying around a normal longboard. That's a good weight and it's going to be a tough goal. Mine is currently under 8kg (I forget what it was - I think I stated earlier in the forum when I weighed it) and I've been really happy with that - I pick it up and carry it barely noticing or thinking about the extra weight so getting lighter would just make it more awesome.

 

[torqueboards]

Hi Damgaard
It will hopefully be in the next couple of months- i suspect..

That's awesome. The board will perfect for a morning commute to work at only 8.8 lbs that's amazing. If it pulls up hills and gets the 8-10 miles will be nice. I would carry extra battery pack or two to get an extra 10-20 miles when I need it. The next thing is to create a nice carrying case or holder to hold the board when walking up/down stairs and/or on the bus/train.

What size board and board would you be using to get the 4kg weight? Are you also using 1x motor or 2x motors? Board will be nice too with a small wii remote.

 

[rf]

Hi Damgaard
It will hopefully be in the next couple of months- i suspect..

That's awesome. The board will perfect for a morning commute to work at only 8.8 lbs that's amazing. If it pulls up hills and gets the 8-10 miles will be nice. I would carry extra battery pack or two to get an extra 10-20 miles when I need it. The next thing is to create a nice carrying case or holder to hold the board when walking up/down stairs and/or on the bus/train.

What size board and board would you be using to get the 4kg weight? Are you also using 1x motor or 2x motors? Board will be nice too with a small wii remote.

We need a nice sturdy, snap-in, quick-release battery pack.

 

[simonjook]

We need a nice sturdy, snap-in, quick-release battery pack.

I am working on that with an auto-antispark. Will report back with pictures when I get it setup. My ESC playing up means I'm not really committed to the design yet in case I need a new ESC and therefore enclosure.

 

[sk8norcal]

this one is pretty small,
http://www.scratcher.de/cityboard/street.htm

weight: 11,55 lbs (5,24 Kg)
topspeed: 25 mph (41 km/H)
range: min. 17 miles (28 Km) but over 19 miles (30 Km) by normal drive no fullspeed with some kick-pushing.

 

[beetbocks]

With Richard we are working to give a complete kit ( electronics and mechanical ) to give to our customers an easy plug and play system. the kit will be more compatible with different kinds of board. Also the new electronics kit will be available in a box enclosure, ready to be fit on your board, with switch, without having to worry about sparks, plug, connectors. Also will be available a direct plug for charge the batteries, without disconnecting them always. The only wire you will see, will be the motor\s connectors.
Also available will be New firmware for a better sync, and 15% more torque on startup, expecially designed for mountain board. Everything to give to our customer a very good and reliable system at a friendly price. I dont think its fair that those who love this hobby need to pay 1500$ for an E board so, a friendly kit is our goal. Then the customer can choose from the different levels of power available and consequently choose how much to spend. Also the nunchuck style controller is ready. I'm just waiting to come back and do some test with Richard. It is programmable, but this functionality will not be available for safety reasons. It is factory programmed with fail safe, brake delay (ms) and throttle delay(ms).

 

[tuurb]

this one is pretty small,
http://www.scratcher.de/cityboard/street.htm

weight: 11,55 lbs (5,24 Kg)
topspeed: 25 mph (41 km/H)
range: min. 17 miles (28 Km) but over 19 miles (30 Km) by normal drive no fullspeed with some kick-pushing.

Jes!

His boards are really nice.
Would love to read something about the items he used (what esc, what pulley's, what reduction, what battery etc, what charger etc.)

I tried to find more info on their German Forum but could'nt find any built threads...(also possible that my German is not good enough )

So Chrizz: please share some more info. Thanks in advance.

Tuur

 

[simonjook]

this one is pretty small,
http://www.scratcher.de/cityboard/street.htm

I saw that before I started mine - it's not exactly rider friendly - such a small board at 40km/h and even that is 5.25kg

 

[board3d]

this one is pretty small,
http://www.scratcher.de/cityboard/street.htm

weight: 11,55 lbs (5,24 Kg)
topspeed: 25 mph (41 km/H)
range: min. 17 miles (28 Km) but over 19 miles (30 Km) by normal drive no fullspeed with some kick-pushing.

Interesting battery case. I see what looks like an on/off switch and a charging port.