update on > 45mph ebike after 4500 miles

[mvly]

Here is my update on > 45mph ebike after 4700 miles. ~4500 on the image with CA below


This will be a long post, but I think it is warranted for all those who want to get into this hobby for the long run.

Here is the original build link and It has since then change quite a bit.
http://endless-sphere.com/forums/viewtopic.php?f=4&t=33528&hilit=temperature+K2

Where to start?...


------------------
BATTERY:

I bought 8 sticks of 10s 25-50C Nanotech back in late 2011. This gave me 74V20Ah battery. I religious charge this pack to 82V (4.1V per cell) and never discharge more than 72V (3.6V per cell) at idle. Moreover, I balance charge EVERY TIME. Balance charging gives me two benefits: 1) The cells are balanced everytime. 2) I get to see which cells are acting up and which cells are fine.
After 1.5 years or so, I can say Lipo is NOT a good battery for the long run. As of now I am at 255 cycles. It is spring now and the batteries are doing a bit better. But in the winter, it sagged from 82V to 76V if I was drawing 50A. Keep in mind, 50A < 3C. Yet this thing was sagging > 0.2V per cell. Totally UNACCEPTABLE! When it was new, it would go from 82V to 79V at 50A and in the cold. Moreover, even with balancing everytime, I now have one weak cell which discharges faster than others. But for some reason, it also come back up faster than the others. In any case, Lipo is NOT the way to go. Look for other chemistry which are still high C but last longer in terms of cycles and usages.

I was hoping to get 10K miles out of this battery before crazy sagging problem. As of now, I don't know how much longer it will last, but capacity is definitely reduced. I use to be able to draw 16Ah out of the pack when new, but now it is more around 13Ah. Summer is coming around and maybe it will help a bit, but who knows. At this rate, I doubt it will help much.

Just for comparison, the specs of these batteries from hobbyking are: 250 cycles or more. Assuming 80% DOD on my 20Ah, this is 16Ah. 16Ah*250 = 4000Ah passing through the battery before it is 80% capacity. Even with the care and maintenance, I am at 2543Ah or a bit more than 1/2 of the batteries specs and it is already giving problems. Meaning their 250 cycles to 80% at 80% DOD is total bull-crap! I suspect similar with other lithium cells. I mean yes they have data on those batteries to prove it, but they do NOT have real-world day to day uses data. Those data can only come when you let user use it over a long period of time.

Here is how I mounted them batteries:
http://endless-sphere.com/forums/viewtopic.php?f=3&t=37597&hilit=battery+mount
Yes yes, I know I should wrap it up, but at this rate, by the time I do, the battery will be shot!

BATTERY conclusion:
If you want something for the long run, Lipo is NOT the way. Some members here might say just keep it between 3.6V and 4.1V per cell and you will be in for the long run. NOPE! You will might get a few more cycles, but not much. And by I mean few, I mean 5-10 cycles! For high power ebikes, voltage sag is everything and once it sags like my battery, it usually means it's time to let it go.

------------------
BIKE FRAME:
This was my 2nd ebike build. The first was a Hybrid bike and at > 30mph speed on USA streets, things started to break down. I wont bore you with the details, but it suffice to know I needed to go to a FS frame.

If you are going to build anything > 20mph and as heavy as an ebike, you definitely need to go FS. Some members might say, it is sufficient to use THICK tires or seat shocks. I say NOPE! You need both and more! Show me a scooter, motorcycles that doesn't have Full Suspension!

Not only do you need Full suspension, you need good torque arms. I wasn't going to risk my body at > 45mph to have my dropout break. So something super thick and mounted properly is required. And if you are running regen you definitely need more insurance. I ended up using what DoctorBass recommended (DP420 black) AND screw holes because without the screw holes, the Epoxy would break during hard regen. Now I have both, the screw holes and the DP420 black to support each other. The DP420 helps the secure the screws which in turns helps the DP420. It is a pretty solid design. So far, >3K miles without incident.

While you are at it, you would need Nord-lock washer. I was using regular washer before and it would keep getting loose after a few rides no matter how hard I twisted it. Now it's bullet proof. I have not since unscrewed my nord-lock 4 months now since the last tire change.



BIKE FRAME Conclusion:
1) GET A FULL SUSPENSION FRAME. Ideally a MTB or Downhill.
2) Get torque arm. A strong one like and bi-directional if you want to run regen and fast acceleration. Some torque arm are only good one direction.
3) Secure the torque very well. Ideally it should never be removed.
4) Get Nord-Lock Washer.

------------------
TIRES:
View attachment 2
One of the worst things that can happen to any ebike is a flat. Though there are solutions out there, they are expensive and might not fit the bill of a high power ebike. Some examples are tubeless tires, carbon struts tires, etc. All of them are expensive and cannot be easily mounted on a hub motor.

So I went ahead and took some advice from few fellow members to put old used tires inside of new tires. But I went a step further.

Currently here are the layers in my bike tires: (Layer 1 is outside, 5 is inside)
1: MAXXIS hookworm 26x2.5
2: Swahable Big Apple Ballon 26x2.15
3: Tuffy 26x2-2.5 Lining
4: SunLite Thorn resistance 26x2.15 tube WITH sealant
5: Double thick rim tape. (essentially using one rim tape or another). <--- This is critical. For some reason Velocity Pyshco rims have this extra elevated center rib which cuts/wear into the tubes. With single rim tape, this caused a flat after a few hundred miles ride. Hopefully the double rim tape and thorn resistance tube helps deals with this problem.

As you can tell, I really hate flats. My ebike is 90lbs with everything on it. Moreover, the battery is exposed so transporting it safely is difficult. Essentially my only hope is to use push it somewhere to repair the tires.

After about 1000 miles with the setup above, I have not had a flat since. I have found many staples, glass shard, etc. punctured in my tires layer 1. But I have not gotten a flat!

TIRES Conclusion:
Make your tires as flat proof as possible. Try what I have above and do more!

------------------
WHEEL:
I got my wheel build at Ilia at ebikessf.com. I don't know if he still is building wheels because he removed that advertisement from his website.
My first motor was built with 13 gauge spoke broke after 2-3K miles due to the way Ilia put them on the rims. He since perfected his method on the physco rims. I then got him to put on 12 gauge spokes on a new motor and I have not a spoke breakage since. Only time will tell if I will get more spoke breakage.
Another thing about rims is make sure it's solid. I have seen other rims break "cough cough Bionx" under stress. Especially if not on a FS frame.

WHEEL Conclusion:
GET 12 Gauge spokes and a downhill rims. Make sure it is build properly. I have heard of stories where people got 12 Gauge spokes but still broke because they had some noob "cough cough crystalyte" put on the spokes. Spend the extra money for less headache down the road.

------------------
MOTOR:
View attachment 1
I originally had the HS3548. It was fast. But the sensors blew and I didn't like running it on a sensorless controller. More about sensorless controller below. Moreover, there were other problems with the wheels. I mean it was the 1st Gen HS3548 so it was bound to have problems. I upgraded to a HS3545 motor also from Ilia. Supposively this one had NSK bearings and Honeywell SS41 sensors. So it should be better. So far, after 2-2.5K miles, I have not had problems since. It is also fast, though not as fast as the HS3548. It also comes with a Temperature probe. I hacked my arduino to get it to work with this temperature sensor. See the following post for the the temperature hack.


http://endless-sphere.com/forums/viewtopic.php?f=3&t=43442&hilit=+temperature

I guess I could upgrade to Cromotor or the Crystalyte H40xx or the 54xx series. but money is tight these days and the HS3545 has and hopefully will be serving me well for many more miles.

With the temperature feedback, I found the sweet spot for power on flats on my HS3545 is 2-2.5KW power input at 74V. On hills, it is around 1-1.5KW WITHOUT pedal assist and 1.5-2KW WITH pedal assist. Normally the temperature gets to 120 degree C and pretty much settles until I run out of battery. Anymore above those specs listed before will result in burnt motor.

MOTOR Conclusion:
Get the best motor you can afford. Going 45mph or above requires some serious motors. The Cystalyte HS35xx should be the minimum you need to run > 45mph speeds

------------------
CONTROLLER:
I started with the 1st Gen 18FET Lyen controller. I asked Lyen to modify my controller for 74V (20s) or 88.8V (24s) Lipo. Since then, I have only used it on 74V mode. This controller has been flawless since mile 0! I can truly say this controller is one damn good controller. I had a BMC 50V controller and it crapped out on me around 3K miles.

I modified the parameters a bit to make it work at higher settings and got rid of the low-voltage cut off because the CA does this for me. I also enable regen to save my brakes. Since it was modded to handle up to 88V, I was able to regen even when I am at 82V. Normally, if not modded, the regen only kicks in after your battery voltage drops below 79V.

The only problem I ever had with this controller was some capacitor/wire coming loose. Around 4.5K miles, the controller refused to turn on (5V regulator wire was disconnected from joint). I just had to open it up and resolder and glue some parts down. And it works perfectly afterwords. Perhaps even better because before some caps where gone and thus the throttle would be twitchy. Now the startup is much smoother like when it was new.

As reliable as these controllers are, I highly recommend carrying a sensorless controller as a backup. I carry a Crystalyte 12FET 74v-88V sensorless controller as a backup. They are cheap, but can get you out of tough situations. For instance, There was this one time when the sensor blown. I just popped in the sensorless controller and it got me home, albiet slower.

The problem with sensorless controller and high power is it gets hot real fast. At same power input (~1.5KW) the sensorless controller would get much much hotter than the sensored one. Maybe it has to do with my High speed wind motor, but overall sensorless is a pain. Startup is jerky and you need to be moving. Power seems weaker at the same power input... And YES I have calibrated both controller's shunt in the CA. So I know it's similar power inputs.

CONTROLLER CONCLUSION:
Get a decent SENSORED controller. And always invest in a sensorless controller as a backup.

 

[bowlofsalad]

Great write up.

I hope this doesn't offend, this is merely my inexperienced opinion. But I've seen a lot of people trying to turn a bicycle into a motorcycle. I like the idea in some ways, reliability is hard to beat in terms of transportation when when all may fail you can still pedal.

However, I am not completely certain it is wise to turn a bicycle into a motorcycle.

Suspension is something I am very interested in so it's nice to hear a strong opinion on the matter.

That level of anti-flat is about as strong as you can go. What effect on watt hour consumption would you say that has? I am sure it increases it, but to what degree.

Thanks for sharing.

 

[MadRhino]

You had poor performance with your Lipo, the reason should be investigated.

I use Zippy Flghtmax 5800 mah cells that I charge at 4.18 and never discharge lower than 3.7. I have an average life of 500 cycles with my Lipo. They start losing performance after 200 cycles, then I use the pack in parallel with a fresh one for extended range. I run 24s, controller set to 150A battery current in the summer, 100A in the winter. Most of my cells are balanced only 4 or 5 times in their lifespan. I bulk charge quickly, up to 3 times a day in the summer.

Cost of Lipo for me is about 10 miles per $, for in a year I ride an average of 10 000 miles with 1 000 $ of batteries. That is probably the most expansive chemistry on the long term, but I like the advantage of weight, C rate and fast charging. Performance has a price.

FS bikes of course, I agree 100%. Upgraded suspension and brake components, a must.

I don't do anything special for puncture proofing, other than using fresh DH racing tires and tubes.
My Demo, for an example, is using 8 tires a year. That is for avg 6 000 miles, more than half being off road on relatively nice, fast mountain trails. All tyres I am using are 2.7 to 3.0 width, dual ply clinchers (Nevegals, Gazzaloddi, Duro, Michelin). The road racer has Hookworms, but I ride little mileage with this one.

I build motor wheels with 12 Ga spokes, rim and flange washers. I use DH racing rims, ideally 34mm wide. I jump 2 ft often, ride the stairs once in a while. My Demo uses 3 rims a year avg.

I do very frequent maintenance; wheels, brakes, bearings and suspension components are always in perfect order. Lube almost everyday, inspection after every mountain ride. I'd say my dirt bikes average 600 $ a year each in parts and maintenance.

Overall, I find my bikes are much better fun and freedom to ride than a gasser, at much lower cost.

 

[ian.mich]

Or: get a sensor/less controller from greentime. sensors broke? no changing controllers.

 

[Hillhater]

After 1.5 years or so, I can say Lipo is NOT a good battery for the long run. As of now I am at 255 cycles. It is spring now and the batteries are doing a bit better. But in the winter, it sagged from 82V to 76V if I was drawing 50A. Keep in mind, 50A < 3C. Yet this thing was sagging > 0.2V per cell. Totally UNACCEPTABLE! When it was new, it would go from 82V to 79V at 50A and in the cold. Moreover, even with balancing everytime, I now have one weak cell which discharges faster than others. But for some reason, it also come back up faster than the others. In any case, Lipo is NOT the way to go. Look for other chemistry which are still high C but last longer in terms of cycles and usages..

Whilst i agree your battery performance is disappointing , i struggle to see what options you could find to give the same performance, capacity, size, weight and cost . ?
Are you convinced "ALL" your packs have degraded equally, or is it possibly just a few individual cells that are handicapping the others ?
Possibly simply using a cheaper 20C standard lipo would do the same job at a lower cost ?

 

[icecube57]

WInter+ Cold Weather = Higher cell RI which means more sag. Unavoidable... Warm you packs up before you ride out probably would have had much less sag.

 

[miro13car]

I hope this doesn't offend, this is merely my inexperienced opinion. But I've seen a lot of people trying to turn a bicycle into a motorcycle. I like the idea in some ways, reliability is hard to beat in terms of transportation when when all may fail you can still pedal.
However, I am not completely certain it is wise to turn a bicycle into a motorcycle.

bowlofsalad,
I thought exactely same thing.
His is not ebicycle for sure, rather motorcycle if we can call it that way.
Pedalling at 40mph ??
the smallest cog on the back can be 11T , in the front I cannot imagine how BIG chain ring do you need to pedal at 40mph?
It would be great commuter scooter if protected from elements of the weather like rain.
I cannot imagine what shown in the pictures in lasting rain.
It is rather proof of concept technical exercise, educatonal tool type but not practically build. how to wash it for example.
great build should not only work but be practical also.
But on other hand to nicely weather proof build battery pack to last less than 1000 cycles?
Exercise on 1000W unrestricted ebike is possible at 25-30mph MAX speed.
my Eplus bike proofs that everyday.
The thing is that my final speed depends COMPLETELY on me and it is hudge motivation to exercise.
Once I stop pedalling I am not on ebike anymore and all pleasure to increase my heart beat is not there, exercise is not there.

 

[TheBeastie]

Here is my update on > 45mph ebike after 4700 miles. ~4500 on the image with CA below

This will be a long post, but I think it is warranted for all those who want to get into this hobby for the long run.

Cool interesting post, should be quite useful to newbs. Just out of interest what battery charger do you use?

 

[grndslm]

I'd also recommend dropping that inner [Schwalbe] tire. Not necessary, just adding way more inertia than needed.

I'd also recommend to NOT even bother attempting to go over 45... and ALWAYS pedal from a dead stop.

I don't have an eBike yet (have gas-powered bicycles, however), but I can assure you I will be building it with efficiency in mind, over speed... similar to the way I've bought my laptops. Every laptop I've bought in the past 10 years has started out with a claimed battery life of 5 hours. In 2003, my Sony Vaio TR1A claimed that 5 hours, but it probly started out closer to 4.5 hours. By the end of year one, it was surely just below 4 hours, maybe 3.8 hours. But it started leveling off around 3 hours or so, I'd say. Even tho, I haven't touched it in years, I'm sure I could pull it out and get it to run for close to 2 hours, if not more. Everybody around me, however, wanted the fastest CPU with the biggest and baddest graphics card for playing games [which they typically played on xbox], and they were lucky to get 1 hour without a plug nearby. Eventually, those times get down to 15 minutes or less, more like 10 minutes. All I'm trying to get at is... if you want a portable electric device, think of how often you really want to use the device without it being plugged in. For the case of an eBike, this is even more of a primary concern than laptops -- where prices now rival desktops.

If you want to run your bike for 3 hours at a time, day in and day out for any number of years, then you might as well start out with "6 hours" worth of batteries on Day 1. You'll actually make it 5 years without wanting to upgrade your bike. And also, pedaling saves your juice. There's a chart around here, I'm sure (haven't seen it, but one must exist), that shows good results from both speed gains *and* realized distance (battery output) will never be able to co-exist. For example, at 10 mph, you can go 20 miles. At 15mph, you can go 16 miles. And at 20mph, you can go 12 miles. These are numbers I just made up in my head, but there is an inverse relationship of some sorts. If you want speeds of 45 miles per hour, day in and day out, for years to come... put a 2-stroke with a shift kit on your bike (full suspension, yes!), or just buy a motorcycle / scooter.

 

[mvly]

Great write up.

I hope this doesn't offend, this is merely my inexperienced opinion. But I've seen a lot of people trying to turn a bicycle into a motorcycle. I like the idea in some ways, reliability is hard to beat in terms of transportation when when all may fail you can still pedal.

However, I am not completely certain it is wise to turn a bicycle into a motorcycle.

Suspension is something I am very interested in so it's nice to hear a strong opinion on the matter.

That level of anti-flat is about as strong as you can go. What effect on watt hour consumption would you say that has? I am sure it increases it, but to what degree.

Thanks for sharing.

Not sure how much the Watt/hour increase but before I usually get around 45Wh. Now I get around 45-50Wh. So 5Wh?

 

[mvly]

Here is my update on > 45mph ebike after 4700 miles. ~4500 on the image with CA below

This will be a long post, but I think it is warranted for all those who want to get into this hobby for the long run.

Cool interesting post, should be quite useful to newbs. Just out of interest what battery charger do you use?

I use 2 1010b+ icharger with HP power supplies. It takes around 2-3 hours to balance charge. One pack is much easier to balance. The other is not.

 

[mvly]

After 1.5 years or so, I can say Lipo is NOT a good battery for the long run. As of now I am at 255 cycles. It is spring now and the batteries are doing a bit better. But in the winter, it sagged from 82V to 76V if I was drawing 50A. Keep in mind, 50A < 3C. Yet this thing was sagging > 0.2V per cell. Totally UNACCEPTABLE! When it was new, it would go from 82V to 79V at 50A and in the cold. Moreover, even with balancing everytime, I now have one weak cell which discharges faster than others. But for some reason, it also come back up faster than the others. In any case, Lipo is NOT the way to go. Look for other chemistry which are still high C but last longer in terms of cycles and usages..

Whilst i agree your battery performance is disappointing , i struggle to see what options you could find to give the same performance, capacity, size, weight and cost . ?
Are you convinced "ALL" your packs have degraded equally, or is it possibly just a few individual cells that are handicapping the others ?
Possibly simply using a cheaper 20C standard lipo would do the same job at a lower cost ?

Yes the cost will be lower if I go for cheaper Lipo. Maybe the performance will be the same. But I went with Nanotech to get the low voltage sag. It seems to me like the nanotech is only nanotech for the first 30 cycles or so. After that, it's just your regular turnigy. I say just get cheap turnigy to begin with. nanotech is overhyped.

 

[mvly]

I'd also recommend dropping that inner [Schwalbe] tire. Not necessary, just adding way more inertia than needed.

I'd also recommend to NOT even bother attempting to go over 45... and ALWAYS pedal from a dead stop.

I don't have an eBike yet (have gas-powered bicycles, however), but I can assure you I will be building it with efficiency in mind, over speed... similar to the way I've bought my laptops. Every laptop I've bought in the past 10 years has started out with a claimed battery life of 5 hours. In 2003, my Sony Vaio TR1A claimed that 5 hours, but it probly started out closer to 4.5 hours. By the end of year one, it was surely just below 4 hours, maybe 3.8 hours. But it started leveling off around 3 hours or so, I'd say. Even tho, I haven't touched it in years, I'm sure I could pull it out and get it to run for close to 2 hours, if not more. Everybody around me, however, wanted the fastest CPU with the biggest and baddest graphics card for playing games [which they typically played on xbox], and they were lucky to get 1 hour without a plug nearby. Eventually, those times get down to 15 minutes or less, more like 10 minutes. All I'm trying to get at is... if you want a portable electric device, think of how often you really want to use the device without it being plugged in. For the case of an eBike, this is even more of a primary concern than laptops -- where prices now rival desktops.

If you want to run your bike for 3 hours at a time, day in and day out for any number of years, then you might as well start out with "6 hours" worth of batteries on Day 1. You'll actually make it 5 years without wanting to upgrade your bike. And also, pedaling saves your juice. There's a chart around here, I'm sure (haven't seen it, but one must exist), that shows good results from both speed gains *and* realized distance (battery output) will never be able to co-exist. For example, at 10 mph, you can go 20 miles. At 15mph, you can go 16 miles. And at 20mph, you can go 12 miles. These are numbers I just made up in my head, but there is an inverse relationship of some sorts. If you want speeds of 45 miles per hour, day in and day out, for years to come... put a 2-stroke with a shift kit on your bike (full suspension, yes!), or just buy a motorcycle / scooter.

There are many reasons why I didn't go out and buy an electric/petrol motorcyle/scooter.

1st, no way will I support the oil companies that much by buying a petrol scooter or motorcycle.
2nd, I DO NOT like the sound of a petrol engine. One of my reason to go electric is the ultra silent ride.
3rd, Even an electric motorcycles cannot be taken on bike lane and bike trails or sidewalk if need be.
4th, The cops will be gunning for me if I rode a regular motorcycle. I have ridden pass cops clocking me at 40mph and they do nothing, albeit the speed limit on the road is 45mph. Still I get 0 tickets if I ride my bike vs drive my car.
5th, No way can I ride a motorcycle right into lab. If I did have a scooter, I would need a parking permit and such and park like 2 miles from my lab where I work. It is totally not worth the trouble.
6th, Finally with a bike, I can maneuver through traffic better. With motorcycles, you are bigger and from time to time, you will get stuck regardless. With a bike, I can maneuver a bit better.

Yes there is a bit of hassle to get the bike reliable, but in the end, the benefits out way the cost significantly for me.

Like I said before, it seem to be like everything about these high-power ebike works except the battery. Hence we still need to find cheap, reliable, long lasting, widely available and light battery. There is a reason why Lead-Acid is still the battery of choice for many of the things out there. But i guess that is nothing new. People here have been trying their best to find the perfect battery and have not found it yet.

 

[mvly]

Or: get a sensor/less controller from greentime. sensors broke? no changing controllers.

True I can, but i hear it's expensive and hard to tune. Also what if the FET is blown. Then you are shit out of luck. I still like the idea of 2 controller. Keep in mind, it's not only the sensor that can go wrong, the controller itself can also.

 

[mvly]

You had poor performance with your Lipo, the reason should be investigated.

I use Zippy Flghtmax 5800 mah cells that I charge at 4.18 and never discharge lower than 3.7. I have an average life of 500 cycles with my Lipo. They start losing performance after 200 cycles, then I use the pack in parallel with a fresh one for extended range. I run 24s, controller set to 150A battery current in the summer, 100A in the winter. Most of my cells are balanced only 4 or 5 times in their lifespan. I bulk charge quickly, up to 3 times a day in the summer.

Well there you go, you got it up to 200 cycles. THen you played with it to keep the performance up. I would like to see how much longer your cells can take ALONE without new cells helping it.
ALso how much of 500 cycles did you really get? I was surprised how few cycles I had even though it looks like I recharge a lot more. I would use the CA to keep track.

Finally I would love to see the Ah passed through your batteries. In my opinion batteries should be gauge on that vs this 100% DOD and cycle count.

At this rate, I would be lucky if I can get 10K miles out of this batteries. I am almost 5K into the bike/battery but the batteries seems to only last a few more K miles. My guess is 7.5K miles total before battery is done for or moved to lower C discharge rating applications.

Who knows, maybe balance charging them every time is bad? Maybe you have to abuse it a bit to get the full claimed cycles out of them?

 

[wesnewell]

After 1.5 years or so, I can say Lipo is NOT a good battery for the long run. As of now I am at 255 cycles. It is spring now and the batteries are doing a bit better. But in the winter, it sagged from 82V to 76V if I was drawing 50A. Keep in mind, 50A < 3C. Yet this thing was sagging > 0.2V per cell. Totally UNACCEPTABLE! When it was new, it would go from 82V to 79V at 50A and in the cold. Moreover, even with balancing everytime, I now have one weak cell which discharges faster than others. But for some reason, it also come back up faster than the others. In any case, Lipo is NOT the way to go. Look for other chemistry which are still high C but last longer in terms of cycles and usages.

1.5 years ago the weather wasn't as cold as it is now. All chemistries sag more when cold, so unless you checked it with the exact same temp, your appraisal is worthless. From summer to winter all batteries will have as much as 50% less power depending on how cold the temp gets.

 

[Lebowski]

CONTROLLER CONCLUSION:
Get a decent SENSORED controller. And always invest in a sensorless controller as a backup.

I'm of the opposite opinion ! Sensored is by definition not good as the timing just cannot be correct. Only sensorless can have correct
timing. But it has to be a good (Field Oriented Control) controller, not some piece of cr@p. Which is why I think you had problems
with your sensorless controller. Loss of torque and heating of motor/controller means it is not running with the correct timing, and since
its sensorless there's nothing you can do except chuck it in the bin...

 

[MadRhino]

...Finally I would love to see the Ah passed through your batteries. In my opinion batteries should be gauge on that vs this 100% DOD and cycle count.

Yep, that is the only way to have a precise figure of battery life

Yet, I consistently do 10 000 miles a year, and buy 1000$ of Lipo every year. My 10 mile per $ figure is holding.
After the year, I still have the old bricks but their capacity is about 25 % lower and some weak cells can't keep balance with the others. Of course, the nice season is starting and I start using the new ones. We recently started recycling those older Lipo, making 4 bricks with 5 keeping only the cells that are of similar performance. The Lipo's from last year will be used on a bike that I built for guests to ride, and I can expect to pull more mileage off them with lower performance standards.

I don't know the reason why your Lipo had a shorter life. They should be first compared with other Nanotech.
Then we would have better knowledge of the factors that affect them Lipo's, if we all kept track of their performance over their entire life, along with the various environmental, usage and maintenance factors.

I sure don't take great care of them, other than being careful for LV.
I charge and discharge fast every day, often I charge them whilst they are hot from a hard ride.
I balance them only when they are off by more than 0.1v, some bricks are balanced only once in the summer.
What else? I wire 8 Ga everywhere. I never parallel cells, I use 24s 1p packs, I parallel (or simply carry and swap) extra packs if extra range is needed. My rides are short and fast most of the time, one could say I do 1000 ten mile rides a year.

 

[electr0n]

I'm running 24s 2p using Turnigy 5ah 20-30c packs. Fully charged at 99.4v, 90 amp battery current. I'm seeing voltage drop down a little below 84v under full load. Packs have been in use for a little under a year. I've put almost 5000km on them. It's below freezing still so I know the cold is contributing to the big sag. I also probably have some losses from poor connections and my phase wires are way too small for the amperage I'm pushing through them.

I don't like the voltage sag. I love the performance at a full charge but voltage drops quickly as does peak performance. I think I'd rather spend money on increased capacity rather than more expensive cells. That should help with the voltage sag and give increased range and maintain top performance a bit longer for a similar cost.

I also almost exclusively balance charge to 4.15v and typically stop before 3.7 but have gone a bit lower a handful of times. I think 1 to 1.5 years of use to be expected from these batteries, although capacity and performance will be reduced. I can live with replacing the batteries annually. It's still one of the most economic forms of transportation around.

I've had way too many flat tires and a few times I've had to walk the bike home and that sucks but overall I'm more than happy with my bike. Most of issues I've had I've been able to improve over time with trial and error.

 

[mvly]

If you replace lipid every year then sure lipid is the way to go, but not everyone has the luxury of replacing every year. My goal was to buy a battery that I won't have to replace until 10K miles or 600 cycles at 80% DOD before problems like weak cells or voltage sag kicks in.

To answer your question about sag in the cold. It is now a bit warmer. The sag is roughly still there. Keep in mind when the battery is new AND in the cold it did NOT sag this much. So I know for sure it is the age and charge cycle affecting it. If you guys do not believe me, prove me wrong and go run you lipo for 250 cycles over a 1.5 year period and get back to me on the sag. No replacing battery. Also make sure you get at least 2300aH through to be consistent with mine. All I know is if I do plan to make and sell these ebike, I would probably NOT offer lipo as battery even if I can somehow make it safe for for he end consumer.

 

[MadRhino]

Of course, I would not offer a production ebike running Lipo.
This is a liability issue, most of those who are buying commercial ebikes are not conscious of battery tech.
Too many people out there, would set them in fire in no time.

A friend of mine running Lifepo, has the same battery for 5 years.
That is what people buying a ready made ebike want.

 

[electr0n]

Wow 5 years? That's a long time. How much degraded is performance after 1, 2, 3 years with lifepo4 I wonder. My only experience is with lead acid and lipo. I'd like something with more longevity but only if it retains high performance over it's life cycle. How long are those a123 batteries lasting for people out there?

 

[MadRhino]

Wow 5 years? That's a long time. How much degraded is performance after 1, 2, 3 years with lifepo4 I wonder. My only experience is with lead acid and lipo. I'd like something with more longevity but only if it retains high performance over it's life cycle. How long are those a123 batteries lasting for people out there?

It is a matter of mileage too.
Lifepo should last at least 4 times the cycles of Lipo.
It is heavier, sagging more, few have an interesting C rate, and they need balancing often if not every time.
But much safer, and cheaper on the long run.
That is why production vehicles use them.

Better chemistry coming soon, hopefully.
Lithium Titanium is promising for long life and reliability, and Lithium Air for high C rate.

 

[mvly]

Personally I am waiting for cells which are quality built. I have a feeling way too many cells are going bad way before their rated cycle charge. I mean for lipo, one cell goes bad, might as well throw the whole pack away. It's too dangerous to savage IMO. I am curious of the reliability rate. Maybe Lipo do last much longer than 250 cycles but because of the imperfection in the manufacturing process, you might only get to 250 before one or more cells go bad.

What I think the industry need is quality control. Too many cells are poorly made and thus fail way too early. I have NOT heard of a lithium technology battery that have not gone bad because of a few cells. I guess this is the price of being the latest and greatest.

 

[MadRhino]

Well, development is taking both ways.

Lithium Air will be even more dangerous than LiCo. I don't know how they control it, but Lithium and Air together are making a violent exothermic reaction. With more than twice the density it is the future for RC and small devices.

Lithium Titanium can be discharged to 0v, probably 5000 times. That is the future for car industry.

Proper quality control in China can't be expected for the next decade. They have the Lithium, so we are left buying batteries like playing lottery. :?