Commuter Booster - <1kg Friction Drive

Great. Where are you from. I am assuming english isn't your first language.

BTW it doesn't look like you have a counterbalance spring to help lift the motor, or dead stops to limit the travel of the swing arm. These are very important to how the drive works. Hope you have something in mind to perform these functions.

And good work on making your own. I love seeing people build them.

- Adrian
 
Sorry for my english. I´m spanish.

Respect of your questions, before I make the fan, the motor is been hot when I climb the mountain, after the add the fan the bike don´t hot. It´s cold
 

Attachments

  • 103_0010aa.gif
    103_0010aa.gif
    170.9 KB · Views: 3,673
Fantastic. Thanks for the data. One I am done releasing the Beta Commuter Boosters I will have to look into fans a bit more.

I have some of the new SK3 motors from hobbyking some of them have built in fans. But I have not had a chance to gather real temp data yet (been too busy programming) The internal fans also make it a bit harder to mount the temperature sensor in the motor windings as you have to get past the fan. :?
 
you have to adapt the fan into the motor.

I buy the fan in Hobbyking.

http://www.hobbyking.com/hobbyking/store/uh_viewItem.asp?idProduct=4846

the absorber to buy in ebay:

http://www.ebay.es/itm/400249824126?ssPageName=STRK:MEWNX:IT&_trksid=p3984.m1497.l2649

the absorber has two functions. First the motor to near the wheel and when the motor to stops it is brake
 

Attachments

  • OR001-01502.jpg
    OR001-01502.jpg
    42.1 KB · Views: 3,665
  • 9394001225008.jpg
    9394001225008.jpg
    42.9 KB · Views: 3,661
Nice. Looks and sounds like it is running well.

How does the little shock absorber work? Is this what you use to limit the amount of engagement with the tire as well?

I assume you adjust the mounting points to give limit the amount the tire engagement, then adjust the spring pre-load to take the weight of the motor. Then you must have a different dead-stop to stop the motor from moving too far away from the tire. Nice.

- Adrian
 
Hi, this is the promise.

For me, I have two form of put the fan.

I prefer the second option, it have less danger

1º option
Fill the centre of the fan with mix “epoxi” and fibre carbon and drill the centre.
103_0032 copia.jpg
Put the fan outside of the support and glue it. See the photo down
IMG_0083 copia.jpg
But the problem is if you touch the fan. It’s very danger.

That’s my solution; put the fan inside the support. It is impossible to contact with your leg.
2 º option
Cut the blade of the fan.

103_0031 copia.jpgView attachment 3
Cut the drawing (paper). I sent the file of autocad "pletina motor bicicleta"

103_0042 copia.jpg103_0034 copia.jpg
View attachment 7

And glue the paper on the cylinder

103_0045 copia.jpg

Cut the cylinder.

103_0046 copia.jpg

Now, sand the blade and the hammer is inserted inside the cylinder previously cut and paste it with epoxy glue

This is the end result.
103_0024 copia.jpg
 
Hi, I think that my invent for the system for surface irregular . This system is impossible to contact the motor with wheel until you the motor control to have got about 20%.

I don´t make, because i am very busy in this moment.
 

Attachments

  • impulsor motor bici.doc
    289.5 KB · Views: 125
impulsor motor in Spanish and English .

Conexión del servo según esquema inferior. (Connect to the servo follow the drawing of down).View attachment 5
Wire Y for servo:
http://www.hobbyking.com/hobbyking/store/__9908__Futaba_Y_Servo_Lead_30cm_Length_5pcs_bag_.htmlView attachment 4

El sistema se opone al contacto accidental (irregular terreno) del motor con la rueda, ya que el servo limita el movimiento del motor y solo toca a la rueda con el 20%. Una vez, a tocado el motor se queda cogido por fricción y el servo se mueve libre.
El amortiguador vale para evitar que se rompa el servo.

(Well, the system is base on the moving auxiliary of the one servo for contact the motor with wheel.
The system is opposed to accidental contact (iregular field) with the wheel motor and the servo motor movement limited and only turns the wheel to 20%. Once, touched the motor is caught by friction and the servo moves freely.
The shock absorber applies to avoid breaking the servo.)

Posición inicial del servo (initial position).MECANISMO MOTOR BICICLETA inicial.jpg
Servo al 20%. Contacto con la rueda (attach the wheel)MECANISMO MOTOR BICICLETA al 20%.jpg
Posición al 50%. El motor sigue cogido a la rueda por fricción (the engine or motor is caught up by friction)View attachment 1
Posición al 100%. El motor sigue cogido a la rueda por fricción (the engine is caughting up by friction)MECANISMO MOTOR BICICLETA al 100%.jpg
 
Interesting mechanism. I like how you just use the existing throttle signal to move the servo. But it looks like the servo will reduce the force it is holding the motor into the tire as it reaches 100%. I assume you are just relying on the motor/tire friction to keep them together. This might not be good at low speed, high load situations. As you want as much drive engagement with the tire as possible to transfer the high torque.

Thanks for posting.

To be honest I have never had an issue with the motor tire engagement on my drives (unless they are not installed correctly), but I use a button throttle interface that accurately controls the motor acceleration for the user, which in turn controls the way it lifts the motor and engages the tire. Maybe for a system with an analog twist type throttle, your concept could be useful for people that only slowly apply the throttle at start-up. But if it only engages at a certain percentage throttle then you may already have a mismatch in speed between motor and tire just prior to engagement, which is not good once they come in contact. But I my system is has no real control over the speed mismatch either... Hmmmmm. Interesting.

- Adrian
 
Update: Beta Release Commuter Booster & Brain Box

Okay. Announcement time. :D

After a lot of testing, refinement, building, testing, refining, building, etc. I am now happy with my new throttle interface/ebike computer/datalogger/brain box.

I have settled on a schematic, nailed down one hardware varient, and finally crammed all the features I want into the firmware. I'll post up pictures in the coming week, and contacting the list of people that have put them names on the waiting list, and put up a thread in the for sale section. But here is a quick heads up.

Commuter Booster
- Only minor changes have been made to the mechanical drive unit, based on alpha test units
- specifically the main pivot shaft will now be a press fit into the swing arm

Brain Box
- button throttle only
- 8x2 LCD display
--- minimal size, but feature packed
- temp sensor input
--- currently just displays temp to screen, and data logs.
--- uses simple, cheap and available LM35 temp sensors
- powered via ESC throttle 5V (if your ESC doesn't have a BEC, you will need an external one)
- SD dataloging
--- each trip creates two data files
--- (1) trip summary stats to one file, and
--- (2) settings & real time stats to a separate log file
- 10bit analog input resolution (for volts, amps, temp )
- unique single button throttle interface
--- performs power limiting to user settable limit
--- cruise control mode
--- turbo mode, quick switch to higher user settable power limit
- fully programmable via single button input
--- default settings are always available to reset to if you mess things up
- battery fuel gauge
--- visual bar graph style battery meter
--- based on pack watt-hours
--- automatically resets when battery is recharged (no need to remember to reset it)
- ebike stats
--- all the usual stuff, plus a few new cool stats like distance left, and average power consumption
- custom designed for my friction drive
--- enforces minimum & max assist speeds
--- PID power & speed control
--- controls motor engagement/disengagement
--- limits power to preserve tires, maximise efficiency, minimise noise
- upgradeable/expandable
--- firmware is update via a standard usb cable, so future firmware can be installed easily
--- spare IOs are still available for future features like, analog throttle, more temp sensors, PAS sensor etc.
--- 16x2 display variant may come if there is enough interest in a larger interface, and more screen real-estate.


Pics and videos to come. Still awaiting the final parts for manufacture but have 95% of them, and design is now finally finalised. Woo hoo.

I am really excited about this, because it makes my drive a much more refined and complete package. I will likely offer fully assembled kits with ESC and motor if people are interested. Just leaving you to supply and plug in your own battery. Doesn't get much easier than that.

Stay tuned. :D

- Adrian
 
Thanks John.

Scope creep got to me a bit. If I had not implement all the data logging stuff, this would have been so much quicker. And I suspect most people won't use the feature, except the ebike stats nerds. But since a normal microSD card will be big enough to capture 1sec data for thousands of rides, it could just be a nice black box for life analysis of bikes. Developing it all myself also allowed me to quickly iterate features, debug, etc. But now it is time for everyone else to join the fun and see what features/tweaks/changes others would like.

How's yours going? I noticed you are going with the CA-LRC on your beautiful new commuter, rather than your throttle interface. I expected to see it available by now.

- Adrian
 
I am getting a few through now but every time I get one or two, I end up selling them as part of a drive package so they are gone before i get to use one. Problem is that they are all hand made at the moment and we are still adding features so we don't want to tie our selves down to a run just yet. Also funds are a little tight to do a manufacturing run at the moment. The version we are working on at the moment has wheel speed and motor speed sensing so we can accurately match engagement and keep slip down to an absolute minimum. In other words the dreaded "design creap" is kick our butts again. :lol: Hopefully this will be the last major addition for a while however Bluetooth comms is also on the development list so who knows.

The CA is a great pre packaged solution for us though and really suits installations where a traditional throttle is the preferred method of speed control. Also its such a neat professional package. Packaging the interface into a good quality water resistant box is still one of our biggest issues. I dare say you would be in same boat on this front.
 
Kepler said:
In other words the dreaded "design creap" is kick our butts again. Hopefully this will be the last major addition for a while however Bluetooth comms is also on the development list so who knows.
:lol: :lol: :lol: Heard that before. As for bluetooth, skip that and go to ANT+. :lol:

Totally agree on the CA, and it looks like the new version Justin has in the works is going to narrow the margin between his and mine in terms of feature set. But for a button throttle mine will most definitely win out. I even prefer it for flat bar bikes.

Yep water resistance packaging has been fun. Hence the 10+ physical variants I have built. All identical schematic, but just trying out moving components around, different enclosures, different mounting points, etc etc. Finally happy with my solution now, adaptable in mount options, suit front or back wheel sensor, easy to remove battery, or stealth builds. Ended up using off-the-shelf enclosures with some minor modes. They won't take a high pressure hose (neither does a CA), but will be fine in rain. So it all looks pretty, and protects things, while still giving relative easy access to microSD, USB, or for adding sensors etc.

Mine is still hand built for all the termination and enclosure modes, but uses others mirocontroller/SD/sensor boards. I won't try and roll my own unless I get volumes to justify a custom board. So I won't be making any money back to on these, as they take a bit of time to make still.

- Adrian
 
Adriam, the servo don´t reduce the force it is holding the motor into the tire as it reaches 100%. I shall think use for road bad or for mountabikes (pothole), because in my case it exist contact the motor with tire on surfice irregurar
 
I needed to look at ermoustruo's drawings a few times to work out how it operated. Basically the servo will help engagement with a small amount of throttle applied. From there on the motor can climb tire and the servo position wont have any effect on how the swing arm moves.

I experimented with a similar strategy but used programming logic to jog the servo. It worked but I came to the realization that the extra complication just wasn't necessary. Smart design by ermoustruo though.
 
Update: Videos

Okay spent a chunk of today getting a few videos together to help those not familiar with the Commuter Booster.

They run through from building a battery, assembling the Commuter Booster, all the various parts of a system, installation, tuning, the new Brain Box ebike computer/throttle interface/datalogger, and a general walk around.

Here you go. I'll put these in a new 4 sale thread soon. I promise. :)

- Adrian

P.S. Apologies for the quality, these were just "first take" videos using my phone.

Building up battery packs.
[youtube]SFVFS0IBX5Y[/youtube]
[youtube]_F47oq8VubI[/youtube]

Assembling the motor to the Commuter Booster
[youtube]7l9Ql2Yi9WY[/youtube]

Walk through of all the main parts of the system
[youtube]k9umDHUhL9g[/youtube]

A complete installation on a road bike
[youtube]GPiVhZunaAE[/youtube]

The all new, all singing, all dancing Brain Box !
[youtube]9QoxyW_7RwM[/youtube]

Tuning/tweaking/optimizing the commuter booster engagement
[youtube]TeDcXKZfDpM[/youtube]

A quick walk around the bike to show it all off.
[youtube]lqzBMwYYgrA[/youtube]
 
Thanks for the kind words Matt. I appreciate it, I have always been a huge fan of the quality and performance of your creations. They are truely an inspiration.

Yeah. The friction drives are definitely not for everyone. But I am still amazed at the performance you can get out of such a small package, while still keeping your bike a bike. They are a totally different beast to a typical hub motor bike. My big old full suspension hub motor bikes is starting to feel pretty neglected. :lol:

Now that the development side of the Commuter Booster has settled down to a point that I am happy with the mechanical and electronics sides, I am starting to think what my dual suspension bike would be like with a big outrunner, and a chain/belt reduction..... Shaving some weight off my 30kg beast, getting a performance hit, and being able to pedal without feeling like I am on a beach would be awesome. But I think this one will just be a one off for me. It takes way too much time to make something suit a bigger audience.

- Adrian
 
Hello Adrian,

I've already registered my interest with you regarding your next release, however I do have 2 questions.

How does the commuter booster handle when set to 200w - I noticed on the video it was set to 500w default?

If set at 200w, would the minimum speed need to be adjusted upwards?

Brad
 
Hi Brad,

To be honest I have not spent a lot of time tuning it for 200w, and from initial testing I was not happy with how it performed.

The Long Answer:
The problem was that at this low a power level when the rider pedals, they are providing a similar amount of power as the motor. So the control loop needs to react quickly as the RC ESCs are actually more speed controllers, rather than power controllers, as they just set the PWM of the phase current.I may need to change the way the PID loops works to make it to my satisfaction. I have all the code implemented to be able to adjust PID parameters on the fly if necessary, the loop is evaluate in the 20-30ms range, but I just haven't got around to tuning it yet. This may take the form of on set of PID parameters during initial engagement, then switching to a different set.

One last things against the modest power limits with this drive, is that at low power levels the drive is more likely to disengage from the tire when you hit bumps on the road. I have never really noticed this in the past as I was typically riding at 500-1000w, and had a no load speed well above my cruising speeds. But today I went for a long test ride with a new motor that is a lower speed and smaller diameter than I have used in the past. At the end of the ride with battery voltage lower, then no load speed was now closer to my cruising speeds on the flat. Meaning that the motors BEMF was reducing the power to a level that was not keeping the drive engaged. Hence my concern.

This taught me that motor selection needs to be based on ensuring that the motors no load speed at the pack minimum voltage needs to be greater than the maximum assist speed for the bike. The simple solution for me was to just use a 6s pack for this motor, rather than the 5s pack.

The Short Answer:
So, will it work with 200w? Yes.
Is it optimsied for this yet? No.
Am I very interested in optimising it for 200w? Yes.
Will I updated the brainbox to suit 200w limits in the future? Yes.
Would I appreciate some help in optimising it? Sure.
Do you need to raise the minimum speed? No.

Cheers,
- Adrian
 
Thanks for your responce Adrian. I'd be keen to help you out - providing my limited knowledge of PWM control of RC motors is adequate.

To overcome the engagement concern, could installing an "engagement" switch (settable along with the spring tension and grub screw stops) help in regards to switching the PID loops?

Is your brain box measuring current flow values, or are you using the winding resistance to govern to a set power rating?

Brad
 
Back
Top