Electric Bikes
Jarrod Tuma
George Patterson
On 11 April 2012 21:19, Jarrod Tuma <jtum...@gmail.com> wrote:
> Hi Guys
> I'm sure there are others around who are either already working on EV's, or
> have in the past, or interested for the future! I haven't been to a physical
> meeting in a while so thought I'd ask the mailing list, wider audience
> anyway.
Yep, there has been some discussion on EVs in passing in the recent past.
> I am starting on an "electric assisted" bicycle project, I'm aiming for
> 2KW peak power and
> half-hour runtime at reasonable speeds, with pedal
> assist, Basically a useful urban travelling bike. it also has to be light
> enough not to be a hard slog when the batteries go flat.
>
> I started this all off by playing with a fisher&paykel washing machine motor
> and a $30 ESC from hobbyking, which did not work, I believe because the ESC
> expected the motor to have far less inertia than it presented. So I modified
> the esc to use sensors for controlling the motor, reprogramming the micro,
> which did work, and was a bit of fun but was never going to make a practical
> bike, so I thought it was time to get serious.
>
> I have bought a 2.7KW brushless outrunner and 300WH (16 x 5000mAh lipo
> cells) of batteries from hobbyking to experiment with, all weighing 2.7KG
> and cost $180 inc shipping from aus warehouse.
I hope that you did the research. The maximum motor size allowed on a
electrically-assisted bicycle is 200w when ungoverned (whatever that
means).
from http://www.vicroads.vic.gov.au/Home/SafetyAndRules/RoadRules/Bicycles.htm
> A motorised bicycle is not classed as a bicycle if:
> * the motor is the primary source of power
> * the motor's power output exceeds 200 watts (whether or not the motor is operating).
>
> These are considered to be motorcycles.
This means that if you remove the pedals from an electric bike, you
require a motor cycle helmet and register the bike as a motorcycle.
The 200watt motor rating is annoying as Europe has a maximum of 250w
which is more reasonable and more readily available.
I'm don't really have the background to be able to comment fully but a
quick scan
> - ATMEGA or ARM controller for it all
> - Modular/adaptable battery monitoring and balancing, allowing the use of
> cheap hobby lithium cells which come without safety circuitry
>
> These ones are desirable, but secondary
> - Regenerative breaking, this should be mostly implementable in software
> without much extra hardware.
I hope it will be doing more braking than breaking. :-)
Regen. braking becomes less efficent at lower speeds and should not
replace the friction brakes. Perhaps use the motor on the back wheel
only. Perhaps supercaps here.
With any project, you are selling it to someone who will need to put
aside their current project for a bit to work on a group project so it
needs to capture their interest. Making yourself available to answer
questions and discuss it does help. Feel free to attend a meeting and
pitch it to the group.
I'd suggest writing some documentation on-line as a brain dump of
ideas as people will read it fi they know about it and are even
remotely interested. It also helps ensures that ideas are not
forgotten, such as the vicroads link that I provided above.
Regards
George
Clifford Heath
Though I don't currently have spare time for a motor control project,
it's an area I studied intensely for a few years, so I'd like to join the
discussion.
On 11/04/2012, at 9:19 PM, Jarrod Tuma wrote:
> I started this all off by playing with a fisher&paykel washing machine motor
I thought those were basically stepper motors (i.e. electrically commutated)?
They have huge torque and I heard of a UK project that produced blindingly
fast go-carts with that type of motor.
> ... the intersil HIP4086 N-channel fet bridge driver
Not familiar with the 4086, but if it's anything like the 4081, it's excellent.
Beware though, these kinds of currents and the *very* fast switching
require *very* careful board design. You'll get massive EMI really easily
and if you don't kill your own logic, you'll at least kill radio and TV for
blocks around and have ACMA down on you like a ton of bricks.
Assuming you get past making smoke, that is :). Budget for at least
a half-dozen sets of FETs… There are all kinds of tiny transients that
become deadly at these currents.
Self-oscillating programmable current-limit is pretty easy to implement and
a good idea. Just a comparator comparing the current waveform from a
sense resistor to a programmable voltage. Look at the reference circuits
around the National LMD14245 for ideas.
> - Regenerative breaking, this should be mostly implementable in software without much extra hardware.
I doubt software is fast enough (assuming you meant braking, not breaking :)).
Generally you turn on both low-side FETs to create a dead short - one
FET conducts in the reverse direction, but that's not a problem b/c the
resistance is so low that the reverse voltage never turns on the body diode.
The short must be left there until *but not longer than* the current rises to
just below saturation of the motor inductance. Then you rapidly switch so
the inductor current is steered back into the battery (i.e. the motor inductance
spans a higher voltage than the supply). Leave it there until the current
reaches zero (but doesn't reverse), rinse and repeat. I.e. it's a very adaptive
and real-time thing, hard to get right in software.
There's (used to be?) a lot of good clues around this at 4qd.co.uk - that
business specialises in these so he doesn't share all his secrets, at least not
all in the same circuit… but most of the tricks are there if you look hard
enough.
Oh yeah, one final thing. If your motor has an output greater than 200W,
you're subject to all the motorcycle law - helmets, licenses, engineering
certification, roadworthiness, registration. There is hope that we might follow
the direction Europe looks to be taking and increase that limit to 1000W, but
for now… please consider.
Clifford Heath.
Jarrod Tuma
I hope that you did the research. The maximum motor size allowed on a
electrically-assisted bicycle is 200w when ungoverned (whatever that
means).
from http://www.vicroads.vic.gov.au/Home/SafetyAndRules/RoadRules/Bicycles.htm
> A motorised bicycle is not classed as a bicycle if:
> * the motor is the primary source of power
> * the motor's power output exceeds 200 watts (whether or not the motor is operating).
>
> These are considered to be motorcycles.
This means that if you remove the pedals from an electric bike, you
require a motor cycle helmet and register the bike as a motorcycle.
The 200watt motor rating is annoying as Europe has a maximum of 250w
which is more reasonable and more readily available.
I'm don't really have the background to be able to comment fully but a
quick scan
I hope it will be doing more braking than breaking. :-)
Regen. braking becomes less efficent at lower speeds and should not
replace the friction brakes. Perhaps use the motor on the back wheel
only. Perhaps supercaps here.
With any project, you are selling it to someone who will need to put
aside their current project for a bit to work on a group project so it
needs to capture their interest. Making yourself available to answer
questions and discuss it does help. Feel free to attend a meeting and
pitch it to the group.
I'd suggest writing some documentation on-line as a brain dump of
ideas as people will read it fi they know about it and are even
remotely interested. It also helps ensures that ideas are not
forgotten, such as the vicroads link that I provided above.
I thought those were basically stepper motors (i.e. electrically commutated)?
They have huge torque and I heard of a UK project that produced blindingly
fast go-carts with that type of motor.
Not familiar with the 4086, but if it's anything like the 4081, it's excellent.
Beware though, these kinds of currents and the *very* fast switching
require *very* careful board design.
Look at the reference circuits
around the National LMD14245 for ideas.
I doubt software is fast enough (assuming you meant braking, not breaking :)).
George
--
You received this message because you are subscribed to the Google Groups "Connected Community HackerSpace" group.
To post to this group, send email to connected-commu...@googlegroups.com.
To unsubscribe from this group, send email to connected-community-h...@googlegroups.com.
For more options, visit this group at http://groups.google.com/group/connected-community-hackerspace?hl=en.
Clifford Heath
> ... the transients are not so bad with most motors due to winding inductance.
Large inductance means you get fast voltage transients. Plan to mount
a small high-current inductor right at the switches so those steps don't
couple in elsewhere. Keep the return loops super-close, preferably on
the exact opposite side of the board. Twist all high-current wiring, etc…
Don't forget about ground currents; ground is just the trampoline that
everything else bounces off.
> I doubt software is fast enough (assuming you meant braking, not breaking :)).
> software can be pretty quick ;)
> This is not a high-rpm motor, but yeah it could be a challenge.
The RPM isn't the issue. The braking effect is achieved in response to
the L/R time constants, not the RPM.
Clifford Heath
Ben Somerville
There is very active Electrical Vehicle branch of the ATA (ata.org.au)
where members have already built EV bikes, cars etc.
Meeting Dates:
May 16 2012 – This is a joint meeting of the ATA Melbourne branch and
the ATA Melbourne EV branch – Topic e-bikes. Further details to
follow. This topic about electric bikes would be of interest to you.
April 30 2012 – A Volvo conversion – Greg Sievert and Wayne Bowers
Introductions – Mario Giannattilio
The ATA EV melbourne branch provides a forum for anyone interested in
Electric Vehicles such as bicycles, cars and other forms of transport.
Format
There is a monthly 30-45 minute meeting where matters pertaining to
electric vehicles are discussed followed by a guest speaker who speaks
to one of the many topics relating to electric vehicles. The formal
part of the night is followed by an informal gathering for tea, coffee
and biscuits and a chance to catch up with like minded people. There
is no need to book and everyone (ATA member or not) is most welcome.
Meeting details
Meetings are usually held on the 4th Wednesday of each month at 7:00pm
in the Engineering Faculty, Swinburne University, Hawthorn campus
(entry via the EN building). but meeting dates may vary, check
bulletins at ata.org.au
For a map of Swinburne see www.swinburne.edu.au/campuses/hawthorn/index.html
If you get lost on the campus, ring 0433 187 702.
Live web streaming and on-demand content.
Live web streaming of the next meeting and recordings of previous
meetings can be viewed at
portal.rivustv.com/listContentPPP.php?user=12109. No charge for live
streaming but small charge for after event viewing.
Ben Somerville
Member CCHS and ATA Melbourne Branch Committee member
Jarrod Tuma
Ah just re-read your previous email, missed the bit where you explained regen braking. so would this usually be done with some analog circuitry? I guess the micro could measure the battery current and adjust its PWM duty cycle to maximise this, or just guess, which would not be 100% effective but should still do something as long as the duty cycle remains short.
Sounds like an interesting challenge. Maybe for a later prototype
Clifford Heath
> You mean have inductors on the power rails?
No, a series inductor. Keeps the voltage slew rates lower on the power wiring
to the motor inductance. They will capacitively couple into everything otherwise,
though shielding can help.
> yeah, or keep the power rails away from any logic/signal and filter the supplies for the logic. or both.
You'll need major filtering on the power rails, again starting with a capacitor
directly across the supply at each H-bridge. Transient currents flow through
these caps, and those currents radiate more the bigger the enclosed loop area.
Loop area, loop area, loop area. Just keep thinking about that, and think about
every part of your current waveform, including the switching glitches.
> Fast voltage transients are not so bad, its transient current that causes EMI right?
No, you can radiate EM either capacitively (voltage) or inductively (magnetic).
Transient current radiates proportional to loop area, which makes it a biggie.
Magnetic coupling is much harder to shield... but both can be problems.
> Ah just re-read your previous email, missed the bit where you explained regen braking. so would this usually be done with some analog circuitry? I guess the micro could measure the battery current and adjust its PWM duty cycle to maximise this, or just guess, which would not be 100% effective but should still do something as long as the duty cycle remains short.
Right. You have a high inductance low-resistance motor, and you want to
put a dead-short across it, until the current reaches a level just shy of
blowing up your MOSFETs, less than your maximum charge current, and
short of inductor saturation (I think you implied it's a coreless motor, so
saturation is less likely to be an issue). If you leave the short there too
long, you'll need new MOSFETs :). If you let the current rise too high then
steer it into the batteries, you'll need new batteries. A lot of traps!
Definitely a case for doing it with hard-wired comparators...
Clifford Heath.
Andy Gelme
On 2012-04-11 21:19 , Jarrod Tuma wrote:
> I'm sure there are others around who are either already working on
> EV's, or have in the past, or interested for the future !
Whilst not specifically about electric bikes ... the New Zealand
Tumanako project provides open-source hardware and software to recharge
and drive electric vehicles ...
http://sourceforge.net/apps/mediawiki/tumanako/index.php?title=Main_Page
http://wiki.greenstage.co.nz/saker/Tumanako
http://blog.greenstage.co.nz/2009/03/let-fun-begin.html
Some of us may recognize Vik Oliver on the list of people ... who is
always into interesting projects.
> I'm still not entirely sure how CCHS shared projects work but if
> anyone is interested (and wants to enlighten me on the dynamics of a
> shared project) let me know
As George pointed out, the key is to present an engaging project that
will interest others enough to join in. There needs to be a project
champion who leads the effort and keeps the momentum going. Good
communication skills and some documentation helps. Using on-line
repositories for the hardware designs and software means that the
project artifacts are available for all.
I'd suggest that you start by having a project brief (goals, approach)
and some reference information, e.g gathering together the URLs
mentioned in your email ... in one place, on-line. It's good to have
somewhere to direct people who want to know more about your project and
the progress. That place may be the CCHS Wiki ... it's up to you.
The CCHS space provides a physical location to meet, get things built
and interact with people with different domain skills. In all other
respects, a CCHS group project is just like any other open-source
hardware / software project ... and the same approaches apply.
--
-O- cheers = /\ /\/ /) `/ =
--O -- http://www.geekscape.org --
OOO -- an...@geekscape.org -- http://twitter.com/geekscape --
Russel
Jarrod Tuma
Hi Jarrod
There is very active Electrical Vehicle branch of the ATA (ata.org.au)
where members have already built EV bikes, cars etc.
David Payne
On Apr 16, 11:25 am, Jarrod Tuma <jtuma...@gmail.com> wrote:
>
> > Hi Jarrod
> > There is very active Electrical Vehicle branch of the ATA (ata.org.au)
> > where members have already built EV bikes, cars etc.
>
> Thanks for the heads up Ben, might tag along tonight.
Association Inc. (AEVA).,
http://www.aeva.asn.au/melbourne , as you may have already learned
from someone at ATA.
(I'd be VERY surprised & disappointed if ATA was now trying to keep
people ignorant of "rival" groups as has long been the practice in eg.
most computer groups).
Jarrod Tuma
Andy Gelme
On 2012-04-17 22:40 , David Payne wrote:
> (I'd be VERY surprised & disappointed if ATA was now trying to keep
> people ignorant of "rival" groups as has long been the practice in eg.
> most computer groups).
Let's not create and publicly spread what appears to be untrue
speculation about another community group. If you must, please do that
on your own web-site or email list ... somewhere not associated with the
CCHS.
All of the people we've met from the ATA EV group have been helpful and
passionate about EV ... and don't seem at all like the sort of people
who'd engage in that sort of politics.
Let's keep the discussion technical ... thanks.
Ben Somerville
Ben