It’s obvious this bike has some extra parts. But look closely and you’ll see the chainring has no chain connecting to it. Pedaling will get you nowhere since [PJ Allen] rerouted the chain in order to drive this bicycle using an electric motor.
He’s got beefy motor which pulls 350 Watts at 24 Volts. For speed control he opted to use an Arduino, pumping out PWM signals to some MOSFETs. This results in an incredibly noisy setup, as you can hear in the bench test video after the break. But once this is installed on the bike it doesn’t quiet down at all. You can hear the thing a block away.
The original road test fried the first set of 7A MOSFETs when trying to start the motor from a standstill. It sounds like the 40A replacements he chose did the trick through. We didn’t see any information on the battery life, but if he runs out of juice on the other side of town we bet he’ll be wishing he had left the chain connected to the crankset.
I like it, I used to have an electric bike that had a ratcheting mechanism in the chainset so that the chain could route via the motor, over the cranks and back to the rear wheel
WOW that is loud at first I thought I heard a steam engine before the PWM whine started.
The sound’s not that bad. Still, I’m sure he’ll try PWM’ing at 20khz once the MIC4427 is installed.
Looks like a fun ride.
Looks like he has good control over the speed and the mosfets hold up. That is most of be battle. Nice.
20khz should quite it down. It is possible to switch mosfets and IGBT at that frequency you tend to generate a lot of heat due to turn on and turn off times. I would actually try lower frequencies say around 500hz. If you like the higher frequencies just try different frequencies within what the power components can handle.
With industrial 3 phase drives I have found that I can find pwm frequencies that would cause some motors to painfully scream. Usually got a pretty quite motor around 8khz and maybe a little rumble at the lower end of the frequency range. I can’t remember what the dc pwm drives operated at.
Well that’s my two cents. Nicely done.
Actually the noise might not be a problem at all, right now electric cars are deemed “too quiet”, so they are required to add some noise making devices (absurd I agree). But this-lower-than-20kHz commutation is a great solution without any modifications… plus you can show off.
“It is possible to switch mosfets and IGBT at that frequency you tend to generate a lot of heat due to turn on and turn off times.”
If you’re having that problem, you need a stronger driver circuit to drive the gate capacitance. The noob error is to wire the MOSFETs directly to output pins on the MCU and think that a few milliamps is enough to open and close it fast.
That’s a good point. I am sure a lot of people don’t think about that. I use to repair AC drives the ones with the IGBTs had some strong driver circuits. Of course the old Bipolar transistor units had HUGE driver circuits. One thing I think people tend to miss is to ensure that the device is off negative bias the gate junction. Although unless he is going to run at high switching frequencies he can probably just ignore.
He could have used the standard PWM motor controller designed for that motor, quietly, simply, and more efficiently for all of $15.
But what fun is that? Here he gets the dual benefits of having an unnecessary Arduino for the internet-based attention whoring, and unnecessary noise for the requisite attention whoring in real life. What could be better than that?
Well, other than a second Arduino, driving a deafening klaxon and some flashing lights of course.
The real achievement unlocked here had to do with making a motor mimic a crying baby. I still like it, I have thought about this every time Ive seen a drill and a bike next to each other since I can remember. +1 for doing it even if the sound it makes makes you wish you didnt.
Just a point of reference – each time I pass by, the birds’ chirping/s can *still* be heard. How’s that for perspective? If you came over you’d be like me, wondering what the issue is. (“Noise? What noise?”)
what we, in the electric? bike world call a ‘freewheel’ would have solved this problem and made the bike a little more.. how you say? ‘street legal’ (when i get pulled over, i point out the fact that i have pedals, and they actually move the bike, and therefore I am riding a bicycle and although you clocked me going 47 on a bicycle, “Your argument is invalid”, because I do not need tags for this “motor vehicle”) lol
Where I live (British Columbia, Canada), riding an electric bike without functioning pedals will get you a ticket for an unregistered, unlicensed motorcycle.
As a consequence, a lot of the “ebike” class scooters all have these entirely useless (but functional!) pedals crammed somewhere into their body.
“your argument is invalid” doesn’t count much with all the lawyering and all that.
Many places have a power limit; round these parts 200W is the limit for electric.
For petrol-driven engines it’s 22cc (brush trimmer motors), but I’d prefer to see & hear those shoved up the riders rear-end.
Then there’s speed limits for non-registered powered vehicles, I’ve no idea what that is here, maybe we don’t have one.
around here and I guess most of EU, the rules are something like this for an electric bike: max 250W, motor may only help when pedaling and speed is less than 25km/h
In practice it means that you have to have the bike approved by some sort of authority, or use only approved standard parts.
But in theory, if you get caught driving a homebuilt electric bike, you’ll get it confiscated, because the police don’t have the expertise or the equipment to test whether or not it adheres to regulations, and you have no way of reliably proving it to them, and they can’t let you go on an unregistered vehicle that may or may not be an illegal motorcycle.
Nobody has really tested that though.
pretty amusing. very loud. people wouldn’t have to stop you and ask “does that thing have a motor on it?” and “is that electric?”
Still:
Freewheel..
???
Profit.
They say that one of the problems with electric bikes is people can’t hear them coming, I say problem solved!
This looks almost identical to the bike I made except his motor is slightly more powerful than mine… I can tell you it does not work very well in real life. I have a double gear reduction between the 250w motor mounted where the rear brake was then uses the central pedal bearing to pass through and then return to the wheel. I welded a bigger sprocket onto the one way bearing so that you can coast. PWM system is built using a ratiometric hall effect sensor twist throttle fed into a S08 microcontroller which in turn drives a 1khz PWM signal to a 280A low loss power mosfet. That’s silicon rating, the to-220 is limited to like 80A but that’s still a long ways off, virtually no heat is generated due to the high efficiency and when the throttle is turned up all the way it just pushes DC so no switching loss. Battery life is poor with 24v 7Ah, but that’s probably because I am still geared too high and the motor is too weak to effectively pull you up a 2 degree incline by the time the batteries are at half capacity, but full charge on flat ground and she’ll get me across campus in a hurry. Just don’t hit a bump, BMX bike and feet on the front pegs is a real kick in the tailbone lol. Then I got a real (700cc) motorcycle and now she just sits in my room and gets brought out for engineering faire lol.
if he didnt want to use a freewheel… get another rear wheel to that bike and make it 2 wheel drive.
Just wanted to say that I made this on a lark, as you can gather from the blog that’s linked. As *Nick* posted, it’s a bit “engineering fare”, and definitely not a solution for the world’s ills.
The motor is nowhere near so loud as figured. The camera mic’s AGC keyed on that and really played it up. I find it reassuring, an audible cue so I know that it’s making the effort (and without adding a klaxon.) In comparison, mid-way or so, there’s a truck that starts-up from about 120 feet away and that seems much louder than it actually is, too. Anyway, in reality, the gearbox’s grinding away is more noticeable than the PWM.
By the way, “Ride No. 1!” shows *the* first ride, as it happened. I had a good feeling that it’d go, but didn’t absolutely know that, and I nearly jumped off the saddle when it actually got going on its own (happy me.)
I’ve found that after about 5 miles of cruising, the batteries start crapping out a bit. They’re two 12V/7Ah SLA, from bgmicro.
I used the Arduino (an inexpensive clone, actually) because.. that’s what they’re for. Who should whump up some one-shots, up-down counters, and oscillators and such when that can be accomplished with a little dev board and 30, 40, 50 (or whatever it is) lines of instruction? And, I can use it again for something else.
What about braking? Does the motor have a switch to disengage if he hits the brakes? Will it just crank the chain and power against the brakes until the motor stops?
The thing sounds like my car when I go fast in reverse.
There’s no disengage via the brake.
I just have the handheld, there’s a button for step up, step down, and one for Off – if any button gets pushed with Off, it goes Off.
Here’s mine if anyone is curious, another blog post details an older 555 based PWM driver with 3x IRF540s. The new system works much better but I haven’t updated the blog in a long time to reflect it’s new features.
http://bac-electronics.blogspot.com/2011/04/e-bike.html
While kinda neat as a DIY, 350 watts isn’t exactly beefy in a bicycle — that’s the BOTTOM END of hub motors, and those can be added to the front wheel without even touching the normal drivetrain.
… and those usually come with all the nice bells and whistles like disengaging the motor when you hit the brake and so-forth.
Doesn’t even look like it hauls; what is that, 12mph or so? a good hub motor can push up towards 35mph.
(having pedaled my internal gear hub bike down a hill to that speed, NOT sure I’d really want one that fast, but still…)
Was thinking on converting mine to electric, but if I do I’m going brushed hub motor (we have hills, brushless is a BAD idea when you need torque), in the 500W/48v range…
Though good luck finding anyone willing to lace a non-standard sized hub in this day and age. Half the bike shops locally in my neck of the woods look at you like you have two heads when you want a normal hub and rim relaced.
this is why jesus invented the hub motor..
Since when is a 350 watt 24v motor “beefy”? Also, working pedals is nice for legality and usability.
The noise is a feature…
I visited several times China recently. In bigger cities almost all scooters are now electrical models.
Take an almost silent 150 kg scooter + driver and the crazy way of Chinese driving it turns out that those scooters are the next best thing to create horrible accidents with pedestrians… not sure but from what I had seen, the statistics really should be peeked by now.
So having something as noisy as this gives again this valuable “Take care, someone with 50km/h is going to approach you” feedback
I have been nearly hit several times. Having said that, they’re freaking cool. It’s a shame that cars are rapidly over taking scooters and bicycles.
And I hope they get some emissions controls soon, as the air in china is… 1971 LA smog inferno bad. All the freaking time. I really wanted a space suit by my third visit.
Most camcorders use sound compression where loud sounds are softened and faint sounds are increased in volume.
The audio from a camcorder is not a good indication of how loud something is.
PJ says, it doesn’t sound loud in person so I don’t think we need to keep worrying about how loud it is.
BTW, Very cool project.