adrian_sm
1 MW
Thanks Richard. Did you mean to link to a specific product?
Commuter Booster - <1kg Friction Drive
- Thread starter adrian_sm
- Start date
Jehu
10 mW
adrian_sm said:
Like i said, not real familiar with electronics. this is programmable, not sure if it can do all u need :?
http://www.australianrobotics.com.au/products/mini-maestro-18-channel-usb-servo-controller-assembled
Adrian,
have you considered using a Castle Creations Phoenix ICE 50A controller? It says it can support 50A continuos which would be around 1000W and if it is as good as the HV160 (which I've just read is good..) it maybe could be something to recommend when using the CB? If you get enough quantity you maybe could talk to castle about them making another firmware that supports potentiometer input and current control?
Moreover, I came across some site stating that with some RC Car ESC's you could make the throttle input map against current instead of voltage but i don't have any resources to reference this to. Does anyone here on ES know if this is the case or not?
have you considered using a Castle Creations Phoenix ICE 50A controller? It says it can support 50A continuos which would be around 1000W and if it is as good as the HV160 (which I've just read is good..) it maybe could be something to recommend when using the CB? If you get enough quantity you maybe could talk to castle about them making another firmware that supports potentiometer input and current control?
Moreover, I came across some site stating that with some RC Car ESC's you could make the throttle input map against current instead of voltage but i don't have any resources to reference this to. Does anyone here on ES know if this is the case or not?
recumpence
1 GW
We have been looking into this for some time now. From what all of us here have found, that is not true.
It would be nice, though!
Matt
It would be nice, though!
Matt
adrian_sm
1 MW
It would be very nice if true. 
I have been searching around for people extracting original firmware from RC ESCSs , or modifying them. The most interesting one I found was the multirotor microcopter guys that took a cheap tower pro ESC and have been using custom firmware to give them a much faster throttle response required for The stability of these aircraft. It looks like they recreated the schematics, then wrote brand new code.
http://www.rcgroups.com/forums/showthread.php?t=766589
Then they did the same for the 50Amp model
http://www.rcgroups.com/forums/showthread.php?t=774524
Shows that it is all possible at least. But probably not worth the effort unless i am just modifying someone else's code.
I have been searching around for people extracting original firmware from RC ESCSs , or modifying them. The most interesting one I found was the multirotor microcopter guys that took a cheap tower pro ESC and have been using custom firmware to give them a much faster throttle response required for The stability of these aircraft. It looks like they recreated the schematics, then wrote brand new code.
http://www.rcgroups.com/forums/showthread.php?t=766589
Then they did the same for the 50Amp model
http://www.rcgroups.com/forums/showthread.php?t=774524
Shows that it is all possible at least. But probably not worth the effort unless i am just modifying someone else's code.
You are probably right. Developing the code from scratch will take big resources.
Those TowerPro mods have I seen before. The 50A limit is a little too low but maybe the firmware could be usable for an other controller?
I'm going to try programming my own throttle interface on an ATMEGA 168 and use an allegro ACS756 halleffect current sensor. I'm currently waiting for the sensor and then I will need some spare time, but it should not take too long time to make a "prototype" program to test with. The nice thing about using the allegro current sensor is that it will be possible to measure the phase current instead of just battery current if one wants torque limiting.
Recipe for a simple throttle interface:
*Pot meter input
*Current measurement input
*PI regulator with pot meter as setpoint and 1-2ms pulse as output to cope with standard ESC
*Some sort of voltage reference to make the current measurements more reliable than just using the BEC output.
*Battery voltage monitor to cutoff at a preset level
Something more that is needed for the basic throttle interface?
/bose
Those TowerPro mods have I seen before. The 50A limit is a little too low but maybe the firmware could be usable for an other controller?
I'm going to try programming my own throttle interface on an ATMEGA 168 and use an allegro ACS756 halleffect current sensor. I'm currently waiting for the sensor and then I will need some spare time, but it should not take too long time to make a "prototype" program to test with. The nice thing about using the allegro current sensor is that it will be possible to measure the phase current instead of just battery current if one wants torque limiting.
Recipe for a simple throttle interface:
*Pot meter input
*Current measurement input
*PI regulator with pot meter as setpoint and 1-2ms pulse as output to cope with standard ESC
*Some sort of voltage reference to make the current measurements more reliable than just using the BEC output.
*Battery voltage monitor to cutoff at a preset level
Something more that is needed for the basic throttle interface?
/bose
adrian_sm
1 MW
Slap on a wheel sensor for minimum speed throttle cut-off, and a pot for user selectable current limit, and you have me sold.
I'll be using a button throttle, but that is easy to make work using the pot input.
I would bring out a few button digital inputs, as I am sure they will end up being useful.
If your prototype works out, and you want to make a batch of real pcbs, let me know and I will see how I can help out. But I'll probably be doing something very similar myself if I can make some more time magically appear in my life.
If we do want to make some PCB's in the future, it may be wise to think about future expansion possibilities. Since this is such a capable uC, may as well make the extra IO's available for extra buttons, pots, sensors, display etc. But something with the above list is all the commuter booster will need.
- Adrian
I'll be using a button throttle, but that is easy to make work using the pot input.
I would bring out a few button digital inputs, as I am sure they will end up being useful.
If your prototype works out, and you want to make a batch of real pcbs, let me know and I will see how I can help out. But I'll probably be doing something very similar myself if I can make some more time magically appear in my life.
If we do want to make some PCB's in the future, it may be wise to think about future expansion possibilities. Since this is such a capable uC, may as well make the extra IO's available for extra buttons, pots, sensors, display etc. But something with the above list is all the commuter booster will need.
- Adrian
Sure that sound alright. I'll think about those things when I start and I will keep you posted if I get any progress.
One thing that comes to my mind is a micro SD card slot. It would make it possible to save ride data and to make a customizations file that can be edited on PC. Maybe an RS232 connection would be nice too.
Hmm sorry for filling your thread with my thoughts
/bose
One thing that comes to my mind is a micro SD card slot. It would make it possible to save ride data and to make a customizations file that can be edited on PC. Maybe an RS232 connection would be nice too.
Hmm sorry for filling your thread with my thoughts
/bose
adrian_sm
1 MW
No worries. But for a first version of my interface, I'll be doing everything to keep it simple, barebones stuff.
Otherwise I'll take forever trying to get all the extras functional, when all I want is for the basics to be rock solid.
Otherwise I'll take forever trying to get all the extras functional, when all I want is for the basics to be rock solid.
nieles
10 kW
hi,
what about the arduino platform? its atmega 168/328 based.
making a program for this should not be an very big task if you know the program language.
and there is a special forum for the arduino with lots of people to help out.
you can load a new program very simple with an usb cable.
for the first version you can simply make a "shield" for the standard arduino uno.
http://www.watterott.com/en/Freeduino-Proto-Shield-Kit (this is an add-on for the arduino)
something like this..
what about the arduino platform? its atmega 168/328 based.
making a program for this should not be an very big task if you know the program language.
and there is a special forum for the arduino with lots of people to help out.
you can load a new program very simple with an usb cable.
for the first version you can simply make a "shield" for the standard arduino uno.
http://www.watterott.com/en/Freeduino-Proto-Shield-Kit (this is an add-on for the arduino)
something like this..
adrian_sm
1 MW
Yep. That would do. I never really understood what the arduino platform are. But they are basically a uC with the IO's exposed on a PCB. Then they have various extras to make life easy, like oscillators, power supplies, USB, bluetooth etc.
Should be a simple hardware platform to use, I just need to finalise what I want from the device then pick one to play with.
Even the basic ones like this would probably do.
http://www.sparkfun.com/products/9218
Wheel sensor should be easy.
Shunt: still undecided on best options
- shunt resistor - ala watt meter
- shunt sensor - allergo
- voltage drop on wiring
- something else
Buttons: easy
Pots: easy
PWM out: easy
Of course I can do most of this with my modified watt meter, it has a ATTiny in it, but just limited code space, so I would have to write my code from scratch, and just focus on the core functions to fit it all in. Then the code should be easily transferable to other ATMega chips.
- Adrian
Should be a simple hardware platform to use, I just need to finalise what I want from the device then pick one to play with.
Even the basic ones like this would probably do.
http://www.sparkfun.com/products/9218
Wheel sensor should be easy.
Shunt: still undecided on best options
- shunt resistor - ala watt meter
- shunt sensor - allergo
- voltage drop on wiring
- something else
Buttons: easy
Pots: easy
PWM out: easy
Of course I can do most of this with my modified watt meter, it has a ATTiny in it, but just limited code space, so I would have to write my code from scratch, and just focus on the core functions to fit it all in. Then the code should be easily transferable to other ATMega chips.
- Adrian
41south
10 mW
These are great devices Adrian, I've been doing some work with them lately on a couple of ham radio projects. I wouldn't call myself a skilled programmer by any means, but happy to help out where I can if you go this way. I personally develop on the Arduino UNO as it already has the usb etc on board. As you say the code can be ported to the other devices once you have finished mucking about with it.
Wheazel
10 kW
I really like the simplicity of this unit.
Do you know how much power it requires to maintain 30mph on flat ground? Talking about the larger motor version.
Reason I ask is because I recently did some testing with a pocketbike slipperclutch gearbox and final chaindrive. It needs 1100-1200W to maintain 30mph.
I suspect the gearbox is eating efficiency for breakfast. Want a comparision to a frictiondrive setup. Bikes are similar.
Do you know how much power it requires to maintain 30mph on flat ground? Talking about the larger motor version.
Reason I ask is because I recently did some testing with a pocketbike slipperclutch gearbox and final chaindrive. It needs 1100-1200W to maintain 30mph.
I suspect the gearbox is eating efficiency for breakfast. Want a comparision to a frictiondrive setup. Bikes are similar.
adrian_sm
1 MW
http://sites.google.com/site/commuterbooster/performance
Wheazel
10 kW
Wow thats great numbers. My setup surely needs some rethinking. Thanks!
adrian_sm
1 MW
I have had a few questions from the alpha testers just clarifying how to install the shaft in the kit so I thought I would write a post that everyone can see.
1) circlips
- No, both the circlip that comes with the shafts from hobbyking and the circlip grooves on the shaft are not used
2) shaft flats
- the shaft flats on the shaft provided by HobbyKing are not used.
- best to press the end of the shaft with out the flats into the swing arm
3) installing the shaft in the swing arm
- You will need to press this in to position. I just do this on my drill press, to make it a controlled process. But you could just place the motor plate on a block of wood, put the shaft in the pivot block, rest the shaft/pivot block on the motor plate and line up the shaft with the hole (using the pivot block as a guide to keep the shaft perpendicular), then tap the shaft in to position with a hammer. so that the end of the shaft is flush with the outside face of the motor plate.
4) swing arm shaft securing grub screws
- these will need to be tightened a few times over the first 100kms or so. As the grub screws seat into the softer shaft material.
5) grip tape
- why is it in two pieces? It was just easier to ship that way : )
6) throttle interface
- yes I am developing my own Commuter Booster throttle interface. But it is not ready for release yet. I still recommend the modified servo tester and ramp control for the DIYer. But once I have the current limitting in place I will be releasing a smaller neater commuter booster model targeted at being a legally limitted product. So 200/250/750 w, pedal first, speed limited assist etc.
That is all for now.
- Adrian
1) circlips
- No, both the circlip that comes with the shafts from hobbyking and the circlip grooves on the shaft are not used
2) shaft flats
- the shaft flats on the shaft provided by HobbyKing are not used.
- best to press the end of the shaft with out the flats into the swing arm
3) installing the shaft in the swing arm
- You will need to press this in to position. I just do this on my drill press, to make it a controlled process. But you could just place the motor plate on a block of wood, put the shaft in the pivot block, rest the shaft/pivot block on the motor plate and line up the shaft with the hole (using the pivot block as a guide to keep the shaft perpendicular), then tap the shaft in to position with a hammer. so that the end of the shaft is flush with the outside face of the motor plate.
4) swing arm shaft securing grub screws
- these will need to be tightened a few times over the first 100kms or so. As the grub screws seat into the softer shaft material.
5) grip tape
- why is it in two pieces? It was just easier to ship that way : )
6) throttle interface
- yes I am developing my own Commuter Booster throttle interface. But it is not ready for release yet. I still recommend the modified servo tester and ramp control for the DIYer. But once I have the current limitting in place I will be releasing a smaller neater commuter booster model targeted at being a legally limitted product. So 200/250/750 w, pedal first, speed limited assist etc.
That is all for now.
- Adrian
Fietsbel
10 mW
After contacting Adrian about my own built that has been kinda inspired by this one, I made a new thread for it: http://endless-sphere.com/forums/viewtopic.php?f=28&t=28895
A little teaser;
And to Adrian; thanks again! :wink:
A little teaser;
And to Adrian; thanks again! :wink:
adrian_sm
1 MW
No problem. Awesome effort BTW.
Check it out peoples, if you have the time, and some tools you too can make your own.
Check it out peoples, if you have the time, and some tools you too can make your own.
adrian_sm
1 MW
A short quote from a different thread that neatly explains the start-up algorithms for sensorless controllers. Thought others mind find it useful in understanding in more depth how these controllers work.
http://www.endless-sphere.com/forums/viewtopic.php?f=28&t=28868&start=15#p416574
Thanks Luke.
http://www.endless-sphere.com/forums/viewtopic.php?f=28&t=28868&start=15#p416574
liveforphysics said:
Thanks Luke.
adrian_sm
1 MW
Just finally finished making a compact dual swing arm 50mm motor Commuter Booster.
The last bit I just finished was the torsion spring adjustment.
On the larger 63mm commuter booster, and the single swing arm 50mm this is done by rotating the collar on the end of the main pivot. For the newer compact dual swing arm design I had to make this adjustment integral with the right hand swing arm. It was a bit more hassle, but makes for a very clean looking design, and better support for the motor.
Here is a side by side with the 63mm CB. I took the photo a while back, but the only difference to the 50mm CB that you could see from the outside would be a single M5 grub screw hole on the right swing arm.
I think this design, with the upcoming current limiting interface, will be a great light weight, moderate power design that with it's smaller size will suit more road bikes. For now I will use my modified watt meter to keep the current in check to about 30Amps.
Next one will be the other end of the spectrum. Using a 80-100 turnigy, terminate in star for a 83kV. I'll try running it at about 48V with a sensored controller. This thing should be evil, with silly amounts of torque. Should be amusing.
- Adrian
The last bit I just finished was the torsion spring adjustment.
On the larger 63mm commuter booster, and the single swing arm 50mm this is done by rotating the collar on the end of the main pivot. For the newer compact dual swing arm design I had to make this adjustment integral with the right hand swing arm. It was a bit more hassle, but makes for a very clean looking design, and better support for the motor.
Here is a side by side with the 63mm CB. I took the photo a while back, but the only difference to the 50mm CB that you could see from the outside would be a single M5 grub screw hole on the right swing arm.
I think this design, with the upcoming current limiting interface, will be a great light weight, moderate power design that with it's smaller size will suit more road bikes. For now I will use my modified watt meter to keep the current in check to about 30Amps.
Next one will be the other end of the spectrum. Using a 80-100 turnigy, terminate in star for a 83kV. I'll try running it at about 48V with a sensored controller. This thing should be evil, with silly amounts of torque. Should be amusing.
- Adrian
adrian_sm
1 MW
Shaft Install
I have been shipping the Alpha kits without the shaft.
The shaft has an interference fit into the swing arm, so requires it to either be pressed, or tapped into place.
Here are a few pics of the two methods I use to install them.
First is using my drill press as a press. Not exactly what it was designed for, but works in a pinch.
I use the pivot block as a guide to ensure the axle is perpendicular to the swing arm.
The other method is with a trust hammer. Just protect the end of the axle with a bit of scrap wood from the hammer, then tap in to place.
I have been shipping the Alpha kits without the shaft.
The shaft has an interference fit into the swing arm, so requires it to either be pressed, or tapped into place.
Here are a few pics of the two methods I use to install them.
First is using my drill press as a press. Not exactly what it was designed for, but works in a pinch.
I use the pivot block as a guide to ensure the axle is perpendicular to the swing arm.
The other method is with a trust hammer. Just protect the end of the axle with a bit of scrap wood from the hammer, then tap in to place.
adrian_sm
1 MW
Update: CB80
Here is a sneak peak of a new prototype. The CB80.
I have tweaked the CB63 geometry and clearences to allow a turnigy 80-100 to be mounted up.
Let the fun begin.
Here is a sneak peak of a new prototype. The CB80.
I have tweaked the CB63 geometry and clearences to allow a turnigy 80-100 to be mounted up.
Let the fun begin.
