Laptop Powered E-Bike

[rebelrider.mike]

I was able to swap the wheels and tires, so the motor is on my original bike now, but still with matching tires. I also got the center stand and front fender moved over. Fitting the axle was tricky. I needed several spacers including a couple of "C" shaped ones that I made myself from washers.

The old motor and drive chain has been removed from my bike, as it is sadly only useless weight now. It was a bit painful to remove it, but I'll save it for a future project. It's still a variable speed permanent magnet motor, so lots of potential uses.


Thanks to You Tube, I taught myself how to adjust derailleurs, and I got the rear axle all cleaned up on the extra bike that I'm fixing up for my brother in law. So that's all done and ready to go. All three bikes are now back on their feet.


I've been wandering what to do with the blue bike once I have it fixed up as a regular bike. It's too tall for anyone in my family. This afternoon, I was talking to the neighbor's daughter, and it occurred to me that she might be able to use it. She's tall and doesn't have a bike. So this might work out perfect.

I'm concerned about the easy-off (or whatever you call it) front wheel axle not fitting the blue bike's fork just right. I have a wheel that is not easy-off that may fit. I can't get to it right now. It's backed behind a bunch of stuff I need to move first.

Anyway, I'm happy with my progress so far.

 

[rebelrider.mike]

The spare green bike has now gone to my brother in law.

Took me a whole day to get a digital speedometer installed. First, the magnet clip didn't fit the thicker spokes of the motorized wheel. So I had to widen it with a file. Then the magnet was too far away from the senser, and I didn't want the senser to stick way into the wheel area. That would make the wheel potentially harder to get off if I have to change the inner tube. So I made an extension with some round magnets I had.

I've been thinking about the best way to manage the battery Voltage vs. the controller and motor's. The battery is 8s16p lithium ion (Samsung ICR18650-22F cells if you want to know) and has a Voltage range of 33.6-22V. The motor and controller are rated for 48V, 1000W, but I don't know what range the controller will accept since it's a no-name with no documentation. I can assume that it's ok between 52.8-42V since it was originally running on a 48V lead acid battery.

I could reconfigure the battery to be 13s, but that would be a lot of work, and I'd need a new BMS, new connectors, and it would still be outside the assumed Voltage range. I think it would be much easier to use a boost regulator. The larges one I've been able to find accepts a maximum current input of 40A, and a maximum load of 1800W.

I made a little chart in order to show some estimated values:


If I limit the battery Voltage to 25.3V minimum, under load, that keeps the boost converter at 40A or less. It also happens to keep each cell at an average of 2.1A discharge. Assuming the motor runs at full power all the time (an extreme that would not happen in real life) the battery should last a little less than an hour. I estimate the capacity of the battery to be roughly 32Ah. If a 1000W motor can push a bike at 30mph like the You Tube videos say, that's a range of about 30 miles. Way more if I'm pedaling most of the time. Less if I'm going uphill all the time.

I won't be going 30mph. If I need that much speed, I have a motorcycle. I'm more interested in the torque the motor has, and if it will help get me up hills. I wonder if the motor has regenerative braking? I'm guessing not. It has three power wires, so I think that means it's a BLDC. If I understand correctly, those don't do regenerative braking. I wonder if I could put a generator on the back wheel and engage it when I'm coasting... Another project for another time.

 

[floydR]

it's a BLDC. If I understand correctly, those don't do regenerative braking.

BLDC can do regenerative braking, it isn't the motor that controls regen. It is the controller and type of motor, series wounddc motor difficult to get regen, seperately excited dcmotors fairly easy to get regen, three phase motors (BLDC, Induction, PMSM, etc) all can do regen. depending on the controller.
later floyd

 

[rebelrider.mike]

Interesting. I don't think mine does though. With the brake lever pulled enough to cut power to the motor but not enough to actually brake, the wheel seems to spin fairly freely. Regen would convert the wheel's motion and slow it down fast right? Also I suspect I have the cheapest controller that money can buy seeing as there's not make or model on it. But that's ok. Just a passing thought.

I spent four hours yesterday tinkering and completely baffled as to why I can't get more that 20A out of my 40A boost controller. I was trying to set the max. current to 20A. Later I went back and re-read the add carefully, and saw in several places that while 40A is the max. input, 20A is the max. output. I got lucky there. I'm going to try 52V output later and see what happens. I even made a new chart:


Today I got the bike together enough to take it around the block! It did pretty good although, I couldn't get up the really steep hill in my neighborhood. The smaller hills were fine though. I suspect the booster is a power bottleneck, so I'll do a little more experimenting and data collecting. But the booster, controller, and motor were all cool after a 5-10 minute ride.

When I built the battery 6 years ago, I used cells that were at the storage voltage of 3.8V. That would have put the battery at around 30.4V I've since done some testing with the old motor and the new one, and even after all this time it's still at 29.3V. I never did charge it, as I still don't have anything that can do 33.6V. My bench power supply can go up to 32.1V, so today I charged it as best I could.


All my chargers do 6s or less. Someday I'll have to come up with an 8s charger.

Anyway, since it seems to be working, I need to come up with a safe, ventilated spot to put the booster. I'm currently tinkering with a bit of PVC pipe:


I'm pretty stoked about todays results though. The last version never made it past the driveway!

 

[rebelrider.mike]

I got my latest video published:

View: https://youtu.be/6zjqibdpiqI


It's a little behind where I am in real life, but that's pretty much how my projects go.

I was actually doing some cable management a couple days ago, getting ready for a test ride with Volt, Amp, and temperature gauges for the battery and boost controller. I made a mistake though, and put the Amp meter shunt on the positive side. These little meter gauges want the shunt on the negative side. So I melted a wire and some connectors. I've got about half repaired now, and it seems none of the actual electronics are damaged. Only the Amp meter sense wire is disconnected, and everything else still works.


Today would be perfect test riding weather, so hopefully I can get it going by this evening.

 

[rebelrider.mike]

Today was bike day! The family and I went on an 11 mile ride on pretty flat ground.


The ebike did great! I did a bunch of estimating, and figure I got roughly 0.64 of a mile per 1% of battery. So I might have a range of maybe 138 miles per charge. Based on a single flat ride, that's probably not accurate, but it's at least a ball-park figure to work with for now.

Nothing fell off. Nothing got hot. And for once I wasn't the slowest person in the group!

Lots more to work on since this is still just the prototype, but this was a fun day!

 

[DiggsUt]

I haven't stopped by in awhile and the thread title caught my attention. (Pardon detour.) I've made several ebike battery packs from laptop cells and it was one of my worst ideas. I repair computers so I had a bunch of cells to work with. I cycled and matched cell capacity but did not use the same brand of cells or cell original capacity when selecting for ebike packs. First season they worked great. Second season I couldn't ride more than a couple miles and it would shut down. I pulled the pack and opened it up. I found leaking cells everywhere. I'm thankful that's all it was. I still build packs but used new matched cells.

 

[OffGridInTheCity]

I found leaking cells everywhere. I'm thankful that's all it was. I still build packs but used new matched cells.

I've only done powerwalls / low stress on 18650 cells and not yet seen 'leaking cells'. What caused the leaking do you think? Too high current (cell abuse)? or harsh outdoor / temps? or?

 

[DiggsUt]

Not temps. Kept inside in the winter and used on nice days in the summer. Cell abuse most likely. It was used pretty harshly on a 1,000 watt mountain bike (30 amp max continuous) and ridden down to BMS cutoff quite a bit and charged to 100% with little or no balancing. Do that to unmatched mixed 18650 cells and some are bound to fail. The leaking surprised me. I'm just glad it wasn't a Lithium fire during charging or the like. That was my first pack ever made. Pack building has come a long way for me and my toys since then. (see below)

 

[rebelrider.mike]

My first battery pack was similar, but used as a lawn mower starter battery. I had similar issues with individual cells failing, but never had leaks. I'd guess though, that maybe the vent popped when the cell got hot and let the electrolyte out?

My ebike cells are all the same make/model, but came from many different sources. I've found (watching lots of YouTube) that folks tend not to think about current draw, and make their batteries way too small for what they're doing. A lot of laptop cells are only rated for 500mA of continuous draw. So if you're asking for 40A from the battery it'll need 80 cells in parallel. That was my first ebike attempt, and the batteries I made were just too big, and I had other issues too. Identifying old cells and finding their rated specs is pretty important I think.

The battery I have now are all Samsung ICR 18650 22F cells rated for 4.4A continuous. Many came from old laptop batteries, so I get to keep the thread title without lying, but a lot came from old modem backup batteries. I've got 16 in parallel for 70.4A continuous, but I'm only drawing about 40A with my current setup. That's at full throttle too, which I haven't been using all the time.

 

[rebelrider.mike]

I finally sat down yesterday and did some estimates, approximations, and outright guesses.

When I started my bike ride the other month, the battery was at 32.1V. That's the highest my benchtop power supply will go. When the ride was over, I still had 30.9V, and we had ridden 11 miles. My Amp meter wasn't working, so I don't know that, but I do know that the terrane was almost flat (we could tell that heading back was a little harder than heading out.) We averaged about 10 mph, and I used the motor intermittently, at about 1/4 power.

I worked out that each cell in the battery would have gone from 4.01V to 3.86V. Since cells have seeeexy discharge curves instead of strait lines, I used the following chart from a research lab to estimate some percentages:


Since I averaged about 1/4 throttle, I figure each cell would be experiencing about 1A when in use. Not very scientific, but it's the best I could do. Anyway, I followed the blue curves, and figured I went from 93.9% to 80.3%. With that huge guess, I estimate that the motor uses (with me pedaling on a flat surface) about 1.2% of the battery per mile.

At the moment, I can't charge the battery past 32.1V, or about 93.9%. Also, I have a low voltage cutoff set on the boost converter at 25V, or 1.1%. This keeps the boost converter from drawing more than 40A from the battery. The maximum current input that the boost converter can handle. (The battery itself can handle up to 70A.) That still leaves me with 92.8% of the battery that I can use. So at 1.2% per mile, that gives me a range of about 114.8 miles.

I doubt I'll actually test that range as my butt can only handle about 10 miles.

Another bit of trivia that might interest you: My ebike currently weighs in at 77 lbs. My son totally cheated and bought a pre-made ebike that weighs 94 lbs. I have a 987 Wh battery, and his is only 840 Wh. Both are made of 18650 cells. He also has a pedal assisted range of only 85 miles. He's got me beat in one area though; his bike is 1,800W and my bike is only 1,000W.

 

[rebelrider.mike]

I noticed that a lot of my old pictures and diagrams are not showing up on the first few posts. No wonder, it's a really old thread now. So, since no one asked for it, here are some updated diagrams I've been working on. Now in wide screen!


Now that the bike works, I'm thinking of improvements I can do, to make the whole thing better/safer. I'd like to add a switch to isolate the battery from everything else. While I'm at it, I could get a thermal breaker that would double as a switch and a "fuse". Though most of the thermal breakers I've seen have bad reviews.

Since the boost converter doesn't get hot during actual use, I can eliminate one of the fans. This would allow me to run the other fad directly off the boost converter's 12V fan connector, and eliminate the 12V relay. It also means I could move the second temperature sensor to the battery box and have two sensors on either side watching for hot cells or whatever.

I'm toying with the idea of adding directional blinkers to the rear of the bike. Mostly for my own tinkering enjoyment. Though not having to use hand signals on the street would be pretty nice. I'd also like to add a tail light and brake light. I bet I could add a relay to the brake lever switches to activate the brake light.

I still haven't built a proper charger yet, though I do have all the parts I need stashed away somewhere.

Since the bike has a hard time getting up steep hills even with both me and the motor working, I'm thinking of adding a second motor for short boosts. The voltage booster limits the battery to 40A, but the battery is good to use at 60A continuously, so I'm thinking of adding a 36V, motor and a voltage booster that can do 20A input. I could use the old modified starter motor, but it turned out to be so inefficient that I'd rather save up for a proper brushed DC motor. I looked into a rear wheel hub motor, but most of my bookmarks are broken, and what I've found so far are still way outside my price range.

Also I want to remove the battery box and reinforce it, and do some wire management inside the box. Maybe add some support or cushioning to the battery so all it's weight isn't floating on the three threaded rods. Also better connectors, and other stuff. Paint it. Maybe some waterproofing.

 

[rebelrider.mike]

Videos 12 and 13 are done.

View: https://youtu.be/DTTps3RtKHA

View: https://youtu.be/YLfA7eBrvsY


Part 12 was a lot of work on the mechanical side of things. Not all of it was on the eBike itself, but on the other two regular bikes that I've since given away.

Part 13 I got a lot of wiring figured out. After lots of tinkering the motor started showing signs of life.

I still have more video clips to go through before I'm caught up with real life, but I've also been very busy with other projects. Including rebuilding my Ford Ranger's transmission. I've never done a transmission larger than a riding lawnmower, so this is quite the learning experience.