Converting B'twin Bike to Electric

ningaman151

100 mW
Joined
Jul 21, 2019
Messages
38
Hi all I am a final year electronic engineering student and I want to convert my B'twin Triban 500 se road bike to electric. What I'm currently thinking is using the following kit: https://www.aliexpress.com/item/32956364535.html?spm=a2g0s.8937460.0.0.76402e0eSTCwL1(the 6 in 1 option with foot)

I've used Arduino many times before, so I plan to use it again for this project to create the battery charging system. I will create my own battery out of 18650 batteries salvaged from laptop batteries. (I will buy them on ebay). I plan to salvage instead of buying because all the ones I found for sale are cheap fakes.

I also plan to use Arduino to create the bicycle computer and throttle system. I wont be using the throttle handle posted in the link above as I wont be able to mount it on the bike (due to the bike having drop bars).

Any thoughts, ideas, or recommendations? Thanks!
 
That motor looks like a pain in the ass to mount to a road bike. If you do manage to mount it, one problem is that it will disbalance the bike, as one side will have to hang out.

The motor is not particularly great in itself either. OK if you are bolting it to some utility trike where weight and efficiency aren't a priority.

There's tons of legitimate 18650 sellers out there. Properly matching 18650's that someone else considered garbage worthy is a time consuming task that requires special balancing equipment, spreadsheets, and dozens of hours of charger/discharger time. Only a job for those who have a lot of time, and very little money.

A road bike also cannot handle this kind of speed anyhow. 3000w translates into about 45mph speeds. Above 25mph, suspension goes from being optional to mandatory for both safety and comfort.

I understand you're new to this forum and maybe ebikes as a whole. I'd recommend checking out some other people's builds and finding what works and wasn't a bit better before taking the plunge. There is a lot to learn from ground zero.
 
I had to google the bike to see what it looks like:

BTwin Triban.jpg

For something like that I'd go with a small hub motor.
 
Yup, i'd run one of the new 3-6lbs dual reduction geared hub motors.. they're perfect for anything with a 700c wheel.
 
neptronix said:
That motor looks like a pain in the ass to mount to a road bike. If you do manage to mount it, one problem is that it will disbalance the bike, as one side will have to hang out.

The motor is not particularly great in itself either. OK if you are bolting it to some utility trike where weight and efficiency aren't a priority.

There's tons of legitimate 18650 sellers out there. Properly matching 18650's that someone else considered garbage worthy is a time consuming task that requires special balancing equipment, spreadsheets, and dozens of hours of charger/discharger time. Only a job for those who have a lot of time, and very little money.

A road bike also cannot handle this kind of speed anyhow. 3000w translates into about 45mph speeds. Above 25mph, suspension goes from being optional to mandatory for both safety and comfort.

I understand you're new to this forum and maybe ebikes as a whole. I'd recommend checking out some other people's builds and finding what works and wasn't a bit better before taking the plunge. There is a lot to learn from ground zero.

Thank you for your reply. Yes I had many double thoughts before taking on this experiment, especially due to the complications of the battery. However my university project deals with a bicycle computer, and I would love to extend it to involve an electric bike. I don't mind buying a mountain bike from Decathlon and converting that. Regardless, I will look into hub motors.

I've found a supplier of 18650 batteries after a search: https://18650.uk/

So how does using 18650 battery packs work exactly? For a 72 volt battery pack that would require 17 or 18 batteries in series. To get the current requirement I would need extra batteries in parallel with each battery in series. How would that work? Do I just place the extra batteries in parallel as is there's only one battery in each stage? Or would I require manual discharge balancing? I know that for charging the battery pack I need to have a connection between the positive and negative terminals on each battery, and I would charge each battery independently.

Any clarification/discussion on the above matters would be appreciated :)
 
fechter said:
I had to google the bike to see what it looks like:

BTwin Triban.jpg

For something like that I'd go with a small hub motor.
Sorry about not posting a picture, I thought that unnecessary images were not appreciated. I will be more thorough from now on :) .

I'll look into hub motors! :D
 
ningaman151 said:
Sorry about not posting a picture, I thought that unnecessary images were not appreciated.
More pics is always better. ;)


The kit you link to in the OP is going to require quite a lot of custom made parts to fit it to a bicycle frame, regardless of type or design. (same for a trike, but at least the trike has more places to put stuff like that that isn't in the way of pedals, feet, legs, etc.)

About the only place there really is to put it is up on the top of a rear rack, or *maybe* under the back of the seat. If the rack isn't VERY stiff, the chainline will keep moving laterally as the rack sways with bike angle changes, and will likely ahve derailment issues. Then there is integrating it's power into your pedal chain, to get it to the rear wheel (or modifying or replacing your rear wheel to drive the left side instead). You *could* drive the front wheel instead, but that also requires modifying or replacing the wheel, and also building a frame / rack to mount the motor to.

There are a number of threads and pages here and around the web showing how those types of systems have been built, if you're really wanting to do it.


But a small hubmotor is a lot easier to mount, and is no different in the electronics to control it, so your bike computer could do the same things there that it could with the first kit.

The bigger the hubmotor, the more torque it can make, and the more you have to worry about good hardware (torque arms or plates) to prevent that from destroying your dropouts/frame.


A middrive (BBSxx, TSDZ2, etc) would also work, and again you could control it the same as anything else. (additionally, there is open-source firmware to replace the OEM firmware in some of those units, here on ES by Casainho / Stancecoke / et al).
 
ningaman151 said:
So how does using 18650 battery packs work exactly? For a 72 volt battery pack that would require 17 or 18 batteries in series. To get the current requirement I would need extra batteries in parallel with each battery in series. How would that work? Do I just place the extra batteries in parallel as is there's only one battery in each stage? Or would I require manual discharge balancing? I know that for charging the battery pack I need to have a connection between the positive and negative terminals on each battery, and I would charge each battery independently.
If you're going to build your own pack, with no prior experience, I would recommend looking up the many 18650 build (and question) threads, and reading as many of them as you can stand, because the answers to every question you can think of (and a whole lot you don't yet know enough to ask ;) ) are in there.

The simplest search that finds most of them is this:
https://endless-sphere.com/forums/search.php?keywords=18650&terms=all&author=&sc=1&sf=titleonly&sr=topics&sk=t&sd=d&st=0&ch=300&t=0&submit=Search
but it also finds a bunch that you won't need (yet). Mostly you can see by title which ones to start with.

A slightly filtered list
https://endless-sphere.com/forums/search.php?keywords=18650+&terms=all&author=&fid%5B%5D=14&sc=1&sf=titleonly&sr=topics&sk=t&sd=d&st=0&ch=300&t=0&submit=Search
shows only the threads in the Battery Technology section, but there's good info not in those.
 
So how does using 18650 battery packs work exactly? For a 72 volt battery pack that would require 17 or 18 batteries in series. To get the current requirement I would need extra batteries in parallel with each battery in series. How would that work?

I'm NOT a pack builder, but even a casual e-bike rider like me will tell you not to go with 72 volts for a build like that. 48 volts would probably be overkill, but either that or 36 volts would work fine. A small front hub motor would also work well with that bike's rigid forks.
 
LeftieBiker said:
I'm NOT a pack builder, but even a casual e-bike rider like me will tell you not to go with 72 volts for a build like that. 48 volts would probably be overkill, but either that or 36 volts would work fine. A small front hub motor would also work well with that bike's rigid forks.
The ESC says it is rated between 72 and 48 volts. So at full charge the battery pack should provide 72v. Correct me if I am wrong :D
 
understand nominal voltage rating

battery voltages perform

across wide voltage range

72v lithium pack envelope

usually between 60-84v

almost 25v difference

full to empty soc



next thesis

understand voltage changes

under load vs open circuit



ohms law is your friend



that platform

front geared hub

good advice
 
ningaman151 said:
The ESC says it is rated between 72 and 48 volts. So at full charge the battery pack should provide 72v. Correct me if I am wrong :D

That just means it will run on a 48v pack (13s), 52v (14s), 60v (16s), or 72v (18s). (for common Li- cells, not LiTi or LiFePO4). You can look around for what voltage range that will give you on each type of pack, but at the lowest a 48v pack will be lower than the 48v, and at the highest a 72v pack will be higher than 72v.

It also means that unless there is a programming function for which pack it's set to, or a jumper on the PCB inside, it will only have an LVC for a 48v pack--all the others above that will be destroyed if you run them until the controller shuts down from LVC.

I'd guess that if your bike computer university project is about ebikes, you have a fair bit of research and lots of reading yet to do. ;)
 
amberwolf said:
ningaman151 said:
The ESC says it is rated between 72 and 48 volts. So at full charge the battery pack should provide 72v. Correct me if I am wrong :D

That just means it will run on a 48v pack (13s), 52v (14s), 60v (16s), or 72v (18s). (for common Li- cells, not LiTi or LiFePO4). You can look around for what voltage range that will give you on each type of pack, but at the lowest a 48v pack will be lower than the 48v, and at the highest a 72v pack will be higher than 72v.

It also means that unless there is a programming function for which pack it's set to, or a jumper on the PCB inside, it will only have an LVC for a 48v pack--all the others above that will be destroyed if you run them until the controller shuts down from LVC.

I'd guess that if your bike computer university project is about ebikes, you have a fair bit of research and lots of reading yet to do. ;)
The esc has a cuttoff voltage of 42v. That would make packs of less than 16s unable to run with the full discharge range.
 
ningaman151 said:
The esc has a cuttoff voltage of 42v. That would make packs of less than 16s unable to run with the full discharge range.
That depends on whether you want to kill the cells or not. ;)

Normally the lowest voltage you'd run the cells down to is around 3.0 to 3.2v for the common Li chemistries (other than LiTI and LiFePO4). There's not usually much capacity below that.

If you run down to the spec sheet minimum, especially at a high current, then you can easily overdischarge the cells, especially if they are not well-balanced. Each excursion below the limit damages the cells, so it's cumulative.

Additionally, the cells will unbalance more easily the lower you run them, making the problem more likely to happen again, each time you do it.


So, assumign a 3.0v cell-level lower limit, then the LVC for a 16s pack is 48v.

The LVC for a 14s pack is 42v.

The LVC for a 13s pack is 39v.

So for an LVC of 42v, on a controller listed as compatible with a 48v (13s) pack, it is using a 3.23v per cell lower limit, which is pretty safe.



A BMS, on the ohter hand, often uses a much lower per-cell LVC (as low as 2.8v), because it is a last-ditch line of defense against destroying a pack or starting a fire.

If you always run a pack down until the BMS LVC kicks in and shuts the pack off, it's really really hard on the pack, and it's useful lifespan will be much shorter than if a higher LVC is used.


YOu may "lose out" on 10% (up to 20%) of the total cell capacities...but they will probably last several times as many cycles.


So...specs for individual parts (like cells) aren't necessarily applicable when those parts are part of a system. The specs may have to be derated for actual usage scenarios.

Lots of research and reading still go go. ;)
 
amberwolf said:
YOu may "lose out" on 10% (up to 20%) of the total cell capacities...but they will probably last several times as many cycles.


So...specs for individual parts (like cells) aren't necessarily applicable when those parts are part of a system. The specs may have to be derated for actual usage scenarios.

Lots of research and reading still go go. ;)

Yes correct there isn't much capacity gain after draining beyond 3.1V for high current scenarios, as can be seen in the image below.

In my application I think I am going to run 3 cells in parallel, and at maximum throttle the amperage of the batteries will be 16.67A. Assuming an amperage of 15A, one can attain from looking at the image that the cut off the voltage of the batteries should be around 3V or 3.1V. :)

Capture.PNG
 
Hey,

Before talking about the electronics, I would suggest studying the mechanical stress you will add to your bike. A common mistake first time builders ( and even second time builders) do is getting a nice motor and battery, and slapping it on the cheapest bike they can find. You have to understand you'll be adding considerable weight to your bike and ride it at speeds most likely not designed for it.

For instance, with the added weight you'll have to run your tires at a high pressure if you don't want them to have a pinch flat on the first bump you encounter, not having suspension doesn't help. That also means that every vibration will be transmitted to your wiring, motor and battery and can cause a cable or a connection to become loose and create a short circuit.
Also, added weight and rim brakes, combined with added speed means that your rims are more likely to heat up and damage your tires.
That's not very relevant if you're just building a proof of concept. It's more of a problem if you're building something that will be ridden regularly

Personally, I wouldn't go past a 36v motor if I were to use the bike you linked. 72v is a complete no go for me
 
First, why are you making the bike, where will you ride it, how will you ride it? You need to figure out the answers to these questions in depth to decide what kind of bike you will build.

Typical of smart people to get into ebiking and try to reinvent the wheel (arduinos n shit). Don't do that. At this point you don't know what you don't know and your intelligence is actually hurting you, rather than helping.
 
TheBMallory said:
Hey,

Before talking about the electronics, I would
...

Yeah true I remember riding that bike and it's really bumpy, can't imagine having a motor on there haha. Would be disasterious!

At this point I have decided to buy a mountain bike. Preferably not a hard tail (to be on the safe side), I think a fully steel frame and fork should be able to handle the stresses, correctly if I'm wrong please, but from my experience steel is very durable.

I say this because I've got no clue how to calculate stressed on bike other than instantaneous stresses, and even with that I don't know how to make sense of the numbers :mrgreen:
 
flat tire said:
First, why are you making the bike, where will you ride it, how will you ride it? You need to figure out the answers to these questions in depth to decide what kind of bike you will build.

Typical of smart people to get into ebiking and try to reinvent the wheel (arduinos n shit). Don't do that. At this point you don't know what you don't know and your intelligence is actually hurting you, rather than helping.

I am making the bike for fun/university project. It's not really for practical purposes but more of a show off party trick lol. I've been thinking about this project for quite some time so it's just about time I start making it. I will be using it to ride on the road, hence road bike. However, I'll be using a mountain bike instead of the road bike listed in my original post :D for durability and feasibility reasons. I think I'll be riding it without pedalling as I'm physically disabled.

I am already using Arduino for the bicycle computer part of things. Yeah I do feel like I'm trying to reinvent the wheel when thinking of making the battery charger out of Arduino, however this is because that's the only way I know how to do it :?.

Really true, I don't know what I don't know but I am trying to learn. How would you recommend I make the charger? And do I just place the parallel li ions together without protection circuitry?
 
What is your plan to make a charger with the arduino? It's just the front end, the hard work is in the power circuitry.

I don't like protection circuitry in my batteries. To bulk charge low cost you can use a CC/CV boost converter running off server PSU(s). Or, a programmable EV charger like Meanwell or others you find searching "programmable ev charger". Those are high power, moderate cost. Grin satiator is an excellent non-balancing charger that is somewhat slow and expensive.

To balance the cells you can install a BMS or manually balance. If your battery is breakable into sections small enough to use an RC charger that is an option. You could even do it individually to each parallel group using a lab power supply. Balancing is rarely necessary with good cells and avoiding deep discharge.

Riding without pedaling is the way to go.
 
ningaman151 said:
However, I'll be using a mountain bike instead of the road bike listed in my original post :D for durability and feasibility reasons. I think I'll be riding it without pedalling as I'm physically disabled.

That's a good choice. You want your center mass as low as possible. :)
 
ningaman151 said:
How would you recommend I make the charger?
You're going to need to learn some electrronics engineering first, as there's a fair bit of stuff to designing chargers and other power control circuitry, not just so they charge the way you want, but so they fail safe rather than burn your house down when they set the battery on fire becuase they failed in a way that let the AC input directly into the battery. ;)

This forum doesn't have the info necessary to teach you what you need to know, but there are places like AllAboutCircuits and similar that have some tutorials that will get you into the basics, and then more engineering-oriented electronics-design forums that can help you work out the details once you've got the idea. It's probably a year or two (or more) and a few hundred to few thousand dollars in blown up parts and wrongly made PCBs, test equipment, etc., to work out a good design for someone that already knows the basics, from the projects I've seen over the years on various forums for similar-complexity projects. Talk to http://ebikes.ca about how long it took to develop the Cycle Satiator, and how much it cost, for example.

If I were you, I'd just buy a charger that's designed for the max charge voltage you're going to use, with the ability to output the current the cells will need to charge at the rate you want (assuming the cells you use can support that rate, when built in a pack, vs what the cell spec sheet says)

If you want a good programmable charger, just buy the Cycle Satiator from http://ebikes.ca.


ningaman151 said:
In my application I think I am going to run 3 cells in parallel, and at maximum throttle the amperage of the batteries will be 16.67A.
So you've measured the controller's current draw, and verified that it only draws a maximum of 16.67A?

If not, you don't know what the maximum throttle current will be. ;)

You can make a guess what it will be based on the advertised current limit for a controller, but with the common cheap ebike controllers, it can be at least several amps higher or lower than the ad says, so it has to be tested under load with the motor you're going to use to find out what it actually is. If you need it to be different, you have to physically modify the controller in one or more of several various ways, or have a programmable-current-limit controller.


Assuming an amperage of 15V, one can attain from looking at the image that the cut off the voltage of the batteries should be around 3V or 3.1V. :)

I"m not sure what the "amperage of 15V" refers to, since you'd have either an amperage or a voltage, but not both, for the same number.
 
flat tire said:
What is your plan to make a charger with the arduino?

I plan to buy some off the shelf by buck converters, and replace their potentiometers with digital potentiometers what will be controlled by the Arduino. Additionally, I'll have current sensors on the output of the buck converters, and this will create a sort of control system, where I adjust the value of the potentiometer to have the correct output current.

All this is for the CC stage, for the CV stage I just set the value of the digital potentiometers to give an output voltage of 4.2V, either by using an analog pin to measure the voltage or using a voltage sensor.

This is repeated for every battery. I plan to use a desktop power supply to power the whole thing.
 
amberwolf said:
ningaman151 said:
How would you recommend I make the charger?
You're going to need to learn some electrronics engineering first, as there's a fair bit of stuff to designing chargers and other power control circuitry, not just so they charge the way you want, but so they fail safe rather than burn your house down when they set the battery on fire becuase they failed in a way that let the AC input directly into the battery. ;)

I don't know why you keep dumbing me down, I'm a final year electronic engineering student like I said, I've used op amps and other electrical components strenuously. I've also done some other electronics projects such as a drone, a buck boost converter, a near field card reader (done on pcb) and other projects that require good knowledge of electronics. So in terms of electronics I have a good ground, no need to relearn what I already know.

I'll be humble and say that I don't know much about batteries and their charging, but I've devoted my time to learning. If I am shown a battery charging circuit it might take me some time to figure how it all works but I'll get it in the end.

I"m not sure what the "amperage of 15V" refers to, since you'd have either an amperage or a voltage, but not both, for the same number.

Obvious mistake of putting V instead of A. I fixed it now.
 
ningaman151 said:
cc/cv arduino

Brilliant plan except there are already a bunch of powerful, inexpensive, off the shelf boost / buck converters that do cc/cv mode.

ningaman151 said:
amberwolf injecting some reality

He's not dumbing you down he's treating you like you have the abilities your posts suggest.
 
Back
Top