ptd
100 W
A little about me... first build, mid 40's, 2 yrs college, math nut (finished calc), mad spreadsheet enthusiast, some pretty decent CAD skills, decent shape, way too much time on my hands, and really impressed with some of the creative minds walking the halls here. I'm looking to build an electric mountain bike, that really can do grades (I live by Mt Shasta, short legs with 15-20% grade, mostly 5-10%). Have the bike (ancient Marin hardtail with front suspension, or a full suspension Specialized), need the electric. Looking to spend less than $1000, so i'd like to get it pretty close to right the first time (might be hopelessly optimistic though, lol).
legality, performance, drive train, controllers, power supply, and mounting.
Legality... I'm in California, so, less than 1000 watts, and less than 20 mph motor powered. I'd like to do lots of forest service roads, which don't allow motor vehicles.
Performance... I'd like efficiency (if I have to go with batteries) and hill climbing. I love the idea of extreme efficiency (for both extended range or power, and lower heating effect), like 92-94% range (which seems to be the rc's or the coreless axial flux's domain), and staying pretty close to whatever range for a variety of speeds and inclines. As for hill climbing, I'd just like to do serious grades, without a meltdown. I don't mind if the speed drops accordingly. Best guess as to typical requirement's...
15 mile round trip, 310 lbs (that includes 30 lbs of motor and battery), 960 sq in frontal area, .7 Cd, .016 road resistance, +/- 10% average grade, 10 mph uphill, 736watts, 108.8 Nm, 74 wh/mi for 7.5 mi uphill, ~550 watts?
35 mile round trip, 270 lbs (also inclusive), 960 fa, .7 Cd, .008 rr, +/- 5% average grade, 20 mph all the way, 820 watts, 60.5 Nm, 41 wh/mi for 17.5 miles uphill, ~720watts?
Maximum grade, 310 lbs, .016 rr, 20% grade, 6 mph, 790 watts, 194 Nm
I figure I should add in motor efficiency, and dod of 70-80%, so 950 watts of power and 1100 watts of batteries seems about right.
Do these numbers seem typical and/or accurate? I guess my question's are, what other “stats†should I be looking at, to get an idea of which motor would work for these figures? Should I be looking at wider gear inch ratios? Or is it just a matter of the motor drawing more current?
Drive train... My first thought at specs for a motor, would be 4-8 inch diameter (could go larger but I was hoping to put a power supply in the triangle as well), 2-4 inch length, 1000 watts rated power, 36-48 volts, 25-30 amps max. Am I wrong, or is this sounding a lot like a hub motor? Would the conversion be easy enough? Them being a kind of outrunner, and all? Do they make covers with reverse threads? Say for rh drive or something? Then get one with a single speed freewheel? Or get one with a disc brake mount, and adapt a lefty freewheeling sprocket to it somehow? Then fabricate a dropout that attaches between the seat tube and down tube?
It seems to me that little to no cogging would eliminate the need for a motor freewheel and adapter altogether, so coreless would be good (and since the diameter could go up, the tooth depth could go down, for the same amount of copper?), but I'm wondering how much coreless affects torque, sounds like a lot, unless you add a lot of magnet weight, although I'm not too opposed to that. I haven't even seen a lot of coreless motors, at least for a decent price, let alone pancake style.
It sounds like for grades gears are in, and hub drives (in a wheel) are out. I don't think I need 21, 7 should do just fine (although I'm looking at 11-32 instead of 12-28). I'm thinking ultra low kv (as in, how do those gearless hubs do it? Get 10 kv? I know part of it is the increased pole count, and more winds, but it seems to me something else must be going on there). Rc motors just seem to have way too much rpm, and the complexity and cost of that kind of reduction seems a bit much. I guess cyclone is pretty close to what I'm thinking, but cost is a little high, quality of that planetary seems weak (maybe not at my power levels, although I am looking at huge amounts of torque), and I can't help thinking that powering through the chainwheel isn't great on a cheap freewheel.
So I'm thinking, put a hub motor into the triangle, ditch the front derailleur altogether, and run the chain right over the top of a high quality (white ENO?) freewheeling hub motor sprocket, (back through an idler, maybe), then around a cheap freewheeling, single, chainwheel. With a 13T motor sprocket, and an 11-32 rear sprocket, and say a 6kv 36v motor, that gives me 260 rpm and 20 mph in high gear, and 88 rpm and 7 mph in low gear. I know i'm leaving out load and no load speed, but couldn't that just be adjusted by selecting a higher kv motor (say 8, 9 or 10 kv)? Then there's a slight difference in power requirement's at those speeds too...
After spending a little while here, going through dozens if not hundreds of threads, I've now become obsessed with the parameters surrounding motor design. (which has resulted in quite a bit of the paralysis by analysis syndrome, but I'm up for a good challenge)
Controllers... Mrbill's dyno of the mars axial flux had some interesting data. So I had to make a spreadsheet. I plotted the lines of peak efficiency at different voltages, and noticed something, that at peak points of efficiency, the amperage was always very nearly 2/3 the voltage. Is there a formula there? or is it just an anomaly of that motor? Do all motors exhibit this behavior, to some degree or another? Is there a way to do this, maintain a maximum efficiency by keeping the ratio of volts to amps in line? Would it be possible to use PWM, and just say, halve the duty cycle, double the amperage, and end up with basically the same amount of power, just at a different voltage? If not, and the voltage is bound to rpm, can't we just drop a gear? Raise the motor rpm, raise the voltage and amperage proportionally, get more power to accelerate? And what are the potential losses involved? (it's hard to imagine the free lunch transformer). If this is the case, you end up with a nice power curve, that increases exponentially with speed, like power required does. Want to slow down or cruise?, gear up, like overdrive. Want to accelerate? gear down, want to accelerate faster?, gear down again, kinda like a car, except better. If I weren't interested in efficiency, I'd mate it to nuvinci somehow, like link the shifter to the throttle, seems like the perfect compliment.
I like the idea of a speedometer linked to the controller, and having the controller only feed motor power if the value is 20 mph or under. Or maybe something along the lines of a “soft STOP†(or reduction), when approaching 20. A governor, of sorts, that would make it easy to comply with legislated speed limits.
I was also thinking of having the controller be able to take power from a generator (and even control the motor, throttle wise), and feed it primarily to the motor, or back to the battery (in the case of deceleration). Lastly, I was thinking paddle shifter. The derailleur is just being activated by a cable. There's gotta be a simple way to simulate this. I'm thinking servo, but it could be worm gear if that didn't work. First advantage, you could put a throttle sensor on it, and control the motor a little while changing gears (like “un torqueâ€, or no load, the motor), seems like it would be great for the freewheels and the rear sprockets. Second advantage, you could hook up something similar to an automatics load sensor, and have it change gears FOR you, aka automatic transmission, for those out there in 406a land (Californias motorized gas bike law requires an automatic transmission).
Power supply... lipo vs lifepo4, vs generator. batteries, fixed power, heavy, and expensive up front, but cheap per cycle. thinking conhis ebay battery, 36v, 30ah, $625 delivered, but leaning towards lipo now. Zippy 10x 6s 5Ah, about $400, +bms + charger. what's the shipping like from HK? I've seen a couple of threads now, with generators, and wonder if a little gx35, ($250 delivered, or the gx50 at $325) directly connected to an rc motor and controller (say $125 delivered), might be the ticket. Just saw one exactly like that on a German website, but there was no data on it (I'm wondering about efficiency though, as in, why the motors can be 85% efficient as motors, but only 65% efficient as generators? Is there some principle here?) Made me start thinking about rc gas too. There's an RCGF 15cc motor that's $215, delivered, puts out 2.1hp, runs at 1500 to 15000rpm, and weighs 2 lbs. 2 stroke type fuel though, and probably screams like a banshee, but I figured I could use some of the extra hp in silencer/spark arrestor type exhaust. But it sounds like the perfect match to an RC motor, rpm wise, as a generator. I worked out the battery at $.35 per kw cycle, and the generator at about $1 per kw cycle. But the battery was around 30 lbs, the generator setup, 5 for the RCGF, 10 for the gx35, and 15 for the gx50, and the gen has the advantage of just carrying a little more gas, like 1 lb, for each extra cycle (not to mention the “refuel†time is nil and available on the road). The gx** is probably the best choice, as it's also carb compliant, which here in California is a something of a requirement. A lot quieter, regular gas, same price, just 5 or 10 lbs heavier, and a fair bit bulkier. The gen setup should also have a battery, powerful enough to start the motor, and able to act like a shunt during deceleration. So, the gx35 fuel consumption, as close as I can figure, is about 15% efficient, and generators are about 65% efficient, which gives us about 3600w of power per gallon. At around 15w/mi avg (I'm thinking with grades and return trips I'm looking at around 37 wh/mi), that's 240 mpge (or 100 for me, which is ok). What if we could get 20% on the motor, and 85% on the gen, this gives us 415 mpge (or 170 for me). Could we get this? What are the factors of efficiency here? Fuel injection? Electronic ignition? Switch to diesel? Better bearings? Balancing? Porting and polishing? Wouldn't a lost of these also get us to an rpm more like the rc motor, say 10k? And what about longevity? Slick50 type of additive to reduce friction? And how do I get it really quiet without losing TOO much power? Any idea of cost on these mods? Or is there some little super efficient mini diesel generator out there that I just haven't found yet?
Mounting... seems pretty straight forward, more of an afterthought really. Although, adjustability is a key factor. Being able to accept different motors, and being able to slide to take up slack in the chain might be a requirement (or idler setup). Also, being able to fit different bikes might get a wider range of people to jump in on development.
A few credits so far... Miles, you are one awesome researcher, good luck with anything you think up. LFP, you're a madman, pushing the limits of machine and man, stay safe, and when are you gunna take those pedals off your motorcycle? Jeremy, you have the most diverse and interesting collection of vehicles i've seen in a long time, do you even OWN a car? May you be forever young. There's a lot of people that really appreciate your patient and kind advice. That goes for all you guys, and many many more.
Well, looking forward to getting to know a few of you guys here, and seriously, thanks for all the inspiration, and help, here at the deep end of the pool.
Paul
legality, performance, drive train, controllers, power supply, and mounting.
Legality... I'm in California, so, less than 1000 watts, and less than 20 mph motor powered. I'd like to do lots of forest service roads, which don't allow motor vehicles.
Performance... I'd like efficiency (if I have to go with batteries) and hill climbing. I love the idea of extreme efficiency (for both extended range or power, and lower heating effect), like 92-94% range (which seems to be the rc's or the coreless axial flux's domain), and staying pretty close to whatever range for a variety of speeds and inclines. As for hill climbing, I'd just like to do serious grades, without a meltdown. I don't mind if the speed drops accordingly. Best guess as to typical requirement's...
15 mile round trip, 310 lbs (that includes 30 lbs of motor and battery), 960 sq in frontal area, .7 Cd, .016 road resistance, +/- 10% average grade, 10 mph uphill, 736watts, 108.8 Nm, 74 wh/mi for 7.5 mi uphill, ~550 watts?
35 mile round trip, 270 lbs (also inclusive), 960 fa, .7 Cd, .008 rr, +/- 5% average grade, 20 mph all the way, 820 watts, 60.5 Nm, 41 wh/mi for 17.5 miles uphill, ~720watts?
Maximum grade, 310 lbs, .016 rr, 20% grade, 6 mph, 790 watts, 194 Nm
I figure I should add in motor efficiency, and dod of 70-80%, so 950 watts of power and 1100 watts of batteries seems about right.
Do these numbers seem typical and/or accurate? I guess my question's are, what other “stats†should I be looking at, to get an idea of which motor would work for these figures? Should I be looking at wider gear inch ratios? Or is it just a matter of the motor drawing more current?
Drive train... My first thought at specs for a motor, would be 4-8 inch diameter (could go larger but I was hoping to put a power supply in the triangle as well), 2-4 inch length, 1000 watts rated power, 36-48 volts, 25-30 amps max. Am I wrong, or is this sounding a lot like a hub motor? Would the conversion be easy enough? Them being a kind of outrunner, and all? Do they make covers with reverse threads? Say for rh drive or something? Then get one with a single speed freewheel? Or get one with a disc brake mount, and adapt a lefty freewheeling sprocket to it somehow? Then fabricate a dropout that attaches between the seat tube and down tube?
It seems to me that little to no cogging would eliminate the need for a motor freewheel and adapter altogether, so coreless would be good (and since the diameter could go up, the tooth depth could go down, for the same amount of copper?), but I'm wondering how much coreless affects torque, sounds like a lot, unless you add a lot of magnet weight, although I'm not too opposed to that. I haven't even seen a lot of coreless motors, at least for a decent price, let alone pancake style.
It sounds like for grades gears are in, and hub drives (in a wheel) are out. I don't think I need 21, 7 should do just fine (although I'm looking at 11-32 instead of 12-28). I'm thinking ultra low kv (as in, how do those gearless hubs do it? Get 10 kv? I know part of it is the increased pole count, and more winds, but it seems to me something else must be going on there). Rc motors just seem to have way too much rpm, and the complexity and cost of that kind of reduction seems a bit much. I guess cyclone is pretty close to what I'm thinking, but cost is a little high, quality of that planetary seems weak (maybe not at my power levels, although I am looking at huge amounts of torque), and I can't help thinking that powering through the chainwheel isn't great on a cheap freewheel.
So I'm thinking, put a hub motor into the triangle, ditch the front derailleur altogether, and run the chain right over the top of a high quality (white ENO?) freewheeling hub motor sprocket, (back through an idler, maybe), then around a cheap freewheeling, single, chainwheel. With a 13T motor sprocket, and an 11-32 rear sprocket, and say a 6kv 36v motor, that gives me 260 rpm and 20 mph in high gear, and 88 rpm and 7 mph in low gear. I know i'm leaving out load and no load speed, but couldn't that just be adjusted by selecting a higher kv motor (say 8, 9 or 10 kv)? Then there's a slight difference in power requirement's at those speeds too...
After spending a little while here, going through dozens if not hundreds of threads, I've now become obsessed with the parameters surrounding motor design. (which has resulted in quite a bit of the paralysis by analysis syndrome, but I'm up for a good challenge)
Controllers... Mrbill's dyno of the mars axial flux had some interesting data. So I had to make a spreadsheet. I plotted the lines of peak efficiency at different voltages, and noticed something, that at peak points of efficiency, the amperage was always very nearly 2/3 the voltage. Is there a formula there? or is it just an anomaly of that motor? Do all motors exhibit this behavior, to some degree or another? Is there a way to do this, maintain a maximum efficiency by keeping the ratio of volts to amps in line? Would it be possible to use PWM, and just say, halve the duty cycle, double the amperage, and end up with basically the same amount of power, just at a different voltage? If not, and the voltage is bound to rpm, can't we just drop a gear? Raise the motor rpm, raise the voltage and amperage proportionally, get more power to accelerate? And what are the potential losses involved? (it's hard to imagine the free lunch transformer). If this is the case, you end up with a nice power curve, that increases exponentially with speed, like power required does. Want to slow down or cruise?, gear up, like overdrive. Want to accelerate? gear down, want to accelerate faster?, gear down again, kinda like a car, except better. If I weren't interested in efficiency, I'd mate it to nuvinci somehow, like link the shifter to the throttle, seems like the perfect compliment.
I like the idea of a speedometer linked to the controller, and having the controller only feed motor power if the value is 20 mph or under. Or maybe something along the lines of a “soft STOP†(or reduction), when approaching 20. A governor, of sorts, that would make it easy to comply with legislated speed limits.
I was also thinking of having the controller be able to take power from a generator (and even control the motor, throttle wise), and feed it primarily to the motor, or back to the battery (in the case of deceleration). Lastly, I was thinking paddle shifter. The derailleur is just being activated by a cable. There's gotta be a simple way to simulate this. I'm thinking servo, but it could be worm gear if that didn't work. First advantage, you could put a throttle sensor on it, and control the motor a little while changing gears (like “un torqueâ€, or no load, the motor), seems like it would be great for the freewheels and the rear sprockets. Second advantage, you could hook up something similar to an automatics load sensor, and have it change gears FOR you, aka automatic transmission, for those out there in 406a land (Californias motorized gas bike law requires an automatic transmission).
Power supply... lipo vs lifepo4, vs generator. batteries, fixed power, heavy, and expensive up front, but cheap per cycle. thinking conhis ebay battery, 36v, 30ah, $625 delivered, but leaning towards lipo now. Zippy 10x 6s 5Ah, about $400, +bms + charger. what's the shipping like from HK? I've seen a couple of threads now, with generators, and wonder if a little gx35, ($250 delivered, or the gx50 at $325) directly connected to an rc motor and controller (say $125 delivered), might be the ticket. Just saw one exactly like that on a German website, but there was no data on it (I'm wondering about efficiency though, as in, why the motors can be 85% efficient as motors, but only 65% efficient as generators? Is there some principle here?) Made me start thinking about rc gas too. There's an RCGF 15cc motor that's $215, delivered, puts out 2.1hp, runs at 1500 to 15000rpm, and weighs 2 lbs. 2 stroke type fuel though, and probably screams like a banshee, but I figured I could use some of the extra hp in silencer/spark arrestor type exhaust. But it sounds like the perfect match to an RC motor, rpm wise, as a generator. I worked out the battery at $.35 per kw cycle, and the generator at about $1 per kw cycle. But the battery was around 30 lbs, the generator setup, 5 for the RCGF, 10 for the gx35, and 15 for the gx50, and the gen has the advantage of just carrying a little more gas, like 1 lb, for each extra cycle (not to mention the “refuel†time is nil and available on the road). The gx** is probably the best choice, as it's also carb compliant, which here in California is a something of a requirement. A lot quieter, regular gas, same price, just 5 or 10 lbs heavier, and a fair bit bulkier. The gen setup should also have a battery, powerful enough to start the motor, and able to act like a shunt during deceleration. So, the gx35 fuel consumption, as close as I can figure, is about 15% efficient, and generators are about 65% efficient, which gives us about 3600w of power per gallon. At around 15w/mi avg (I'm thinking with grades and return trips I'm looking at around 37 wh/mi), that's 240 mpge (or 100 for me, which is ok). What if we could get 20% on the motor, and 85% on the gen, this gives us 415 mpge (or 170 for me). Could we get this? What are the factors of efficiency here? Fuel injection? Electronic ignition? Switch to diesel? Better bearings? Balancing? Porting and polishing? Wouldn't a lost of these also get us to an rpm more like the rc motor, say 10k? And what about longevity? Slick50 type of additive to reduce friction? And how do I get it really quiet without losing TOO much power? Any idea of cost on these mods? Or is there some little super efficient mini diesel generator out there that I just haven't found yet?
Mounting... seems pretty straight forward, more of an afterthought really. Although, adjustability is a key factor. Being able to accept different motors, and being able to slide to take up slack in the chain might be a requirement (or idler setup). Also, being able to fit different bikes might get a wider range of people to jump in on development.
A few credits so far... Miles, you are one awesome researcher, good luck with anything you think up. LFP, you're a madman, pushing the limits of machine and man, stay safe, and when are you gunna take those pedals off your motorcycle? Jeremy, you have the most diverse and interesting collection of vehicles i've seen in a long time, do you even OWN a car? May you be forever young. There's a lot of people that really appreciate your patient and kind advice. That goes for all you guys, and many many more.
Well, looking forward to getting to know a few of you guys here, and seriously, thanks for all the inspiration, and help, here at the deep end of the pool.
Paul
