Is this a viable bike for conversion? (Updated)

2wGlider

1 mW
Joined
Jul 16, 2019
Messages
16
Hi everyone :)

I´m planning on building my second e-bike. Hope this one turns out successful as the first one :mrgreen:

So this is my first ebike. I intend to do a similar build with a few differences.




It was built 3 years ago.
58,8V 14,5Ah 14S 5P Panasonic PF cells, with 130 cycles to date.
Cheap 1000W motor/ 20A controller, shunt modded to 45A, giving a peak of 2646W.
Hits 50-55kmh with 75kg person.

Now, for this build, we have this bike to work with:







My friend needs to cover more ground in each run, speed needs to be confortably around 60km/h.
The frame fits confortably 120 cells with space to spare, so battery capacity shouldn´t be a problem.
To gain the extra speed, I was thinking about using the same 48V motor but with a 15S battery, shunt modd controller to around 40A and replace capacitors for 100V ones. So power is around 2520W.
He weighs 55kg, so that might help with speed.

My main concern about this bike is the hard tail integrity: The derailleur side is 5mm (the broken aluminum piece was from the derailleur). The other side is 9mm. If I use torque arms, will the frame hold fine?

Regarding motor choice, is there a better alternative?

What would you guys change to meet the speed and safe requiremnts?
Thanks!
 
Looks good from where I'm sitting. Maybe check out what you're going to get into mounting a battery? Check the dimensions of a few batteries to see if you'll be able to mount them in the available space.
 
Assuming the drivetrain and brakes are in decent shape, then yes for a rear hub or mid drive, and maybe OK for a small front hub. Battery space is limited unless you go with a back pack. Might be challenging to add a rear rack, if you want to carry stuff.
 
AHicks said:
Looks good from where I'm sitting. Maybe check out what you're going to get into mounting a battery? Check the dimensions of a few batteries to see if you'll be able to mount them in the available space.
AHicks said:
Looks good from where I'm sitting. Maybe check out what you're going to get into mounting a battery? Check the dimensions of a few batteries to see if you'll be able to mount them in the available space.

I hit submit insted of preview! Sorry for that guys! The thread is now updated :)
 
2wGlider said:
I´m planning on building my second e-bike. Hope this one turns out successful as the first one :mrgreen:

So this is my first ebike. I intend to do a similar build with a few differences.

So this one was not yours? ;)
https://endless-sphere.com/forums/viewtopic.php?f=49&t=110161


My friend needs to cover more ground in each run, speed needs to be confortably around 60km/h.
Make sure you've got brakes, wheels, and tires and good suspension that are capable of handling these speeds. Note that without good suspension, anything other than smoothly paved roads could be dangerous at those speeds; hitting a pothole or piece of debris could break a wheel or throw you off the bike. Part of it depends on rider experience, but sometimes it's just dumb luck what happens in those circumstances.

To gain the extra speed, I was thinking about using the same 48V motor but with a 15S battery, shunt modd controller to around 40A and replace capacitors for 100V ones. So power is around 2520W.
Note that a shunt mod means the controller can no longer correctly detect the current in the system, and therefore cannot protect itself (or anythign else) against overcurrent, so sudden catastrophic failures are possible at any time; with a rear DD hubmotor this could effectively lock up your wheel and cause a skid (and crash) (but with a front hubmotor it could throw the rider over the bars at whatever speed they were going, to land headfirst at that speed). Sometimes nothing bad ever happens...sometimes it fails immediately.

Note that since 15s is an unusual number of cells in series, you'll probably need to customize a BMS and charger for it. Using a more standard number of 14s or 16s or 20s would probably make it easier to find off the shelf parts to build or repair it.


He weighs 55kg, so that might help with speed.

Weight doesn't really change top speed, just acceleration and ability to go up slopes. Air resistance will change speed, so if the rider is in a more aerodynamic position they'll be able to go faster for the same power usage, so as long as the motor will spin the wheel fast enough to reach the needed RPMs to get to that speed, it would be able to do so.



My main concern about this bike is the hard tail integrity: The derailleur side is 5mm (the broken aluminum piece was from the derailleur). The other side is 9mm. If I use torque arms, will the frame hold fine?
In the case of unknown used potentially-damaged frame dropouts, then rather than just torque arms, I would make torque plates that bolt / clamp to the frame and original dropouts, to ensure that regardless of the frame condition you don't have an axle spinout or wheel separation. There are various examples around the forum such as in the Torque Arm Picture Thread, if it helps.
 
[/quote]
So this one was not yours? ;)
https://endless-sphere.com/forums/viewtopic.php?f=49&t=110161

The one I showed now is the first one I built, and owned ;)
I have also the sondors bike and a foldable one. We live in a hilly area unfortunatly...

Make sure you've got brakes, wheels, and tires and good suspension that are capable of handling these speeds. Note that without good suspension, anything other than smoothly paved roads could be dangerous at those speeds; hitting a pothole or piece of debris could break a wheel or throw you off the bike. Part of it depends on rider experience, but sometimes it's just dumb luck what happens in those circumstances.

All that will be taken into consideration, for sure! The Berg´s front suspension is meant for a small axel wheel, hope it´s strong enough :confused:

Note that a shunt mod means the controller can no longer correctly detect the current in the system, and therefore cannot protect itself (or anythign else) against overcurrent, so sudden catastrophic failures are possible at any time; with a rear DD hubmotor this could effectively lock up your wheel and cause a skid (and crash) (but with a front hubmotor it could throw the rider over the bars at whatever speed they were going, to land headfirst at that speed). Sometimes nothing bad ever happens...sometimes it fails immediately.

Had no ideia that could happen... Besides the power output, my controller is working the same as before the modd, It seems to give the same reading everytime I throttle it. (well actualy now is acting weird, when I throttle the motor goes backwards an inch, thengoes forward lol)
Do you mean it can go over 45A despite what the amp meter says?
How exacly can the DD motor lock up ???

Note that since 15s is an unusual number of cells in series, you'll probably need to customize a BMS and charger for it. Using a more standard number of 14s or 16s or 20s would probably make it easier to find off the shelf parts to build or repair it.

I was a litle afraid that my motor would burn up with 14S but its doing fine. Could I feed it with 15 or 16S safely if I keep the temperature low? Correct me If I´m wrong, with more amps you need thicker cables, with more voltage you need thicker insulation.


Weight doesn't really change top speed, just acceleration and ability to go up slopes. Air resistance will change speed, so if the rider is in a more aerodynamic position they'll be able to go faster for the same power usage, so as long as the motor will spin the wheel fast enough to reach the needed RPMs to get to that speed, it would be able to do so.

Oh ok, makes sense. I´ll still lose in a drag race anyway :cry:


In the case of unknown used potentially-damaged frame dropouts, then rather than just torque arms, I would make torque plates that bolt / clamp to the frame and original dropouts, to ensure that regardless of the frame condition you don't have an axle spinout or wheel separation. There are various examples around the forum such as in the Torque Arm Picture Thread, if it helps.

I´ll take a look thanks!
 
2wGlider said:
Had no ideia that could happen... Besides the power output, my controller is working the same as before the modd, It seems to give the same reading everytime I throttle it.
Do you mean it reads the same current (using an external ammeter with it's own shunt, not reading the controller's shunt) after the mod as before?

If so, it means the mod didn't actually change the shunt resistance, or the controller is not using that shunt to control current flow.

If you mean instead that after the mod it reads the same (now higher than before the mod) reading using that external ammeter whenever it's used under the same conditions, then that is normal.

If you mean the controller itself reports the same (now lower than before the mod) reading whenever it's used under the same conditions, that's also normal, because the shunt resistance doesn't keep changing once it's modified, so it will keep reading the same (wrong, lower) current when it's placed under the same load conditions.



(well actualy now is acting weird, when I throttle the motor goes backwards an inch, thengoes forward lol)
That last is usually a sign of a hall sensor problem. Either a connection issue, a wiring problem, or an actual sensor failure, causing the controller to be unable to determine the rotor position, and so sending the wrong phase order signals and causing it to spin backwards until it reads the other sensors still working normally reporting the wrong spin direction and then it reverses and spins forwards.

This also happens in sensorless-mode of controllers taht support that, when they can't read the BEMF phase signals from the motor correctly at startup; as soon as the motor is moving fast enough it's much easiier to read and so then works normally. (which is why on (probably many) sensorless controllers, pedal-start is either an option that makes it work more reliably, or isn't optional but forced, whether or not the controller uses PAS input).



Do you mean it can go over 45A despite what the amp meter says?
A separate external ammeter will read whatever it is supposed to, as long as it is using it's own external shunt. But any readings the controller itself makes using it's own modified shunt are not correct, and since this modification is essentially uncontrolled, it's not even really easy to predict what the actual shunt resistance is, and thus whether the controller can even read current at all or do any protection (if the resistance is low enough, it effectively doesn't read current flowing, and the controller literally has no idea what is happening regarding current flow).

Some controllers have phase-current monitoring, via shunts or other sensors on the actual phases before they exit the controller, but most don't. Those that do are usually FOC controllers. So if the battery current measurement shunt is modified or unable to read battery current, it is still able to read motor currents, and protect the motor and the phase FETs, but can't protect the battery against excessive discharge currents.

Controllers without phase current monitoring (meaning most controllers) can't reliably monitor any current flow in the system at all once the shunt is modified, because the current they think is flowing is smaller (often far smaller) than what's really happening. They "guess" based on some math what the phase currents are based on the battery current and some info about what it's doing with the motor at the moment...but without the real battery current, all it's guesses are wrong.

Whether anything ever fails because of it depends on how close to the edge of the parts' capabilities the controller was designed to work to start with, and then how much the shunt mod changed it's resistance and thus how far off the readings are from reality, and then how hard the system is pushed and how often. Sometimes stuff blows up the first time it's tried, sometimes it never fails. :)




How exacly can the DD motor lock up ???
If the FETs in the controller fail, they will usually fail shorted. This essentially shorts the motor phases together. Doing that creates braking forces in the motor...shorting two phases is pretty severe drag but all three is like jamming on a brake. The faster the motor speed, the more current that tries to flow and the harder the braking forces that occur.


If you've never experienced this, and have the desire and time, you could disconnect the motor entirely from the controller, and connect the *motor* phase wires together, first a pair, then all three, and try to manually turn the wheel. The more powerful (torque) the motor design, the harder it will be to turn it this way. If you try spinning the wheel by hand and *then* short them together, you'll find it's even stronger. Don't try it while riding....

Just don't accidentally connect the *controller* phase wires together, because if power is present it may fry your controller and you'll need a new one. :( ).



I was a litle afraid that my motor would burn up with 14S but its doing fine. Could I feed it with 15 or 16S safely if I keep the temperature low? Correct me If I´m wrong, with more amps you need thicker cables, with more voltage you need thicker insulation.
More cells in series isn't more amps. It's more volts. That doesn't burn up motors. It just makes them spin faster for the same throttle setting.

Your controller shunt mod is what makes more amps. That could, though if it hasn't already it's unlikely to. ;)

The insulation on the existing wires, including the windings, is probably capable of handling a couple hundred volts or more, possibly up to a few hundred.


There's not really much specific reason other than finding a BMS to suit the unusual number of series cells to protect it, for using any particular number of series cells, beyond how much voltage you want your system to have to spin the motor at a specific RPM for top speed (which depends on the motor's kV (RPM/volt)), and how much voltage your controller is designed to take.
 
amberwolf said:
If you mean instead that after the mod it reads the same (now higher than before the mod) reading using that external ammeter whenever it's used under the same conditions, then that is normal.
^This. It´s working properly, just with more output. barely gets warm at all


That last is usually a sign of a hall sensor problem. Either a connection issue, a wiring problem, or an actual sensor failure, causing the controller to be unable to determine the rotor position, and so sending the wrong phase order signals and causing it to spin backwards until it reads the other sensors still working normally reporting the wrong spin direction and then it reverses and spins forwards.
Got it... I´ll reasearch about it, hope its fixable.
Apart from the problem it self, theoreticly, am I loosing "Power On" time because of that sensor? Could this be caused by overheating? I have a Temp sensor on the stator, and rarely went over 70ºC. Althought I had a few runs where I pushed too hard on slopes and I bet the coils were boiling hot....

A separate external ammeter will read whatever it is supposed to, as long as it is using it's own external shunt. But any readings the controller itself makes using it's own modified shunt are not correct, and since this modification is essentially uncontrolled, it's not even really easy to predict what the actual shunt resistance is, and thus whether the controller can even read current at all or do any protection (if the resistance is low enough, it effectively doesn't read current flowing, and the controller literally has no idea what is happening regarding current flow).
Controllers without phase current monitoring (meaning most controllers) can't reliably monitor any current flow in the system at all once the shunt is modified, because the current they think is flowing is smaller (often far smaller) than what's really happening. They "guess" based on some math what the phase currents are based on the battery current and some info about what it's doing with the motor at the moment...but without the real battery current, all it's guesses are wrong.
Whether anything ever fails because of it depends on how close to the edge of the parts' capabilities the controller was designed to work to start with, and then how much the shunt mod changed it's resistance and thus how far off the readings are from reality, and then how hard the system is pushed and how often. Sometimes stuff blows up the first time it's tried, sometimes it never fails. :)



Great info, thanks :) In this case, my controller doesn´t have display, only had 3 leds as voltage indicators, so I used this external ampmeter. Its very consistent and I trust it (I used the low voltage red led as an On indicator :)


If the FETs in the controller fail, they will usually fail shorted. This essentially shorts the motor phases together. Doing that creates braking forces in the motor...shorting two phases is pretty severe drag but all three is like jamming on a brake. The faster the motor speed, the more current that tries to flow and the harder the braking forces that occur.

If you've never experienced this, and have the desire and time, you could disconnect the motor entirely from the controller, and connect the *motor* phase wires together, first a pair, then all three, and try to manually turn the wheel. The more powerful (torque) the motor design, the harder it will be to turn it this way. If you try spinning the wheel by hand and *then* short them together, you'll find it's even stronger. Don't try it while riding....

Just don't accidentally connect the *controller* phase wires together, because if power is present it may fry your controller and you'll need a new one. :( ).

You bet I won´t try it while riding!! shock: :lol:


More cells in series isn't more amps. It's more volts. That doesn't burn up motors. It just makes them spin faster for the same throttle setting.
Your controller shunt mod is what makes more amps. That could, though if it hasn't already it's unlikely to. ;)
The insulation on the existing wires, including the windings, is probably capable of handling a couple hundred volts or more, possibly up to a few hundred.

Well, apart from the faulty sensor, yes it´s doing "fine". It has became noisier, It does a loud squiking sound when throtteling also, might be the bearings idk

There's not really much specific reason other than finding a BMS to suit the unusual number of series cells to protect it, for using any particular number of series cells, beyond how much voltage you want your system to have to spin the motor at a specific RPM for top speed (which depends on the motor's kV (RPM/volt)), and how much voltage your controller is designed to take.

Because my friend has to cover 15km daily and tackle a few steep slopes, we are thinkg about using a more capable motor that could handle 15-16S safely without having to slow down to much when climbing. thats about it.
What motor would you recommend?
 
2wGlider said:
Apart from the problem it self, theoreticly, am I loosing "Power On" time because of that sensor?
Not sure what you mean by "power on time". Could you provide details?


Could this be caused by overheating? I have a Temp sensor on the stator, and rarely went over 70ºC. Althought I had a few runs where I pushed too hard on slopes and I bet the coils were boiling hot....

Where is the sensor physically located? If it is at the windings themselves (the part most likely to be damaged by sudden heat spikes, since they are what generates the heat), then you should see any of those heat spikes on your temperature measurement device, assumign it is capable of reading them (meaning, it can read that high and it refreshes it's display often enough to catch short-term excursions).


Because my friend has to cover 15km daily and tackle a few steep slopes, we are thinkg about using a more capable motor that could handle 15-16S safely without having to slow down to much when climbing. thats about it.
What motor would you recommend?
As noted, the motor itself will handle whatever voltage you use on there (as it's unlikley you'll be in the hundreds-of-volts category). The batttery voltage has more to do with the top speed you can reach given a particular motor, wheel size, and any gearing between them. Battery current delivery capability is likely to be much more important in your situation.

So what you need is a motor that can handle the power demand of the work you are asking of it, *at the speed* you need to do that work. (if you have to slow down a lot and still push the motor near it's limits it may overheat).

You'll need to know the specifics of the slopes you must go up, and at least guesstimate the total weight of the entire system, bike, rider, and anything else they'd be carrying, to determine the approximate power required of the system to do the climbing work.

I recommend using the simulator at http://ebikes.ca/tools/simulator.html to see how this works, experimenting with various setups that might fit your needs.
 
Not sure what you mean by "power on time". Could you provide details?

" That last is usually a sign of a hall sensor problem. Either a connection issue, a wiring problem, or an actual sensor failure, causing the controller to be unable to determine the rotor position, and so sending the wrong phase order signals and causing it to spin backwards until it reads the other sensors still working normally reporting the wrong spin direction and then it reverses and spins forwards."

When the bike is moving, does that burned sensor affect speed or torque at all? In your explanation you said the burned sensor causes a "misfire" of the controller, so I assume the overall On time would be less now.

Where is the sensor physically located? If it is at the windings themselves (the part most likely to be damaged by sudden heat spikes, since they are what generates the heat), then you should see any of those heat spikes on your temperature measurement device, assumign it is capable of reading them (meaning, it can read that high and it refreshes it's display often enough to catch short-term excursions).



Idk, can you tell from this picture?
The temp sensor is glued to the rim, close to the coils but not touching them (the picture was taken before I intalled it). The refresh time is a bit slow but works fine. If I go up a hill quite hard for 3-4min min and then stop the motor, the temp continues to rise for like 10sec maybe. (that might be due to the heat being dispersed slowly to the stator)

Battery current delivery capability is likely to be much more important in your situation.
So what you need is a motor that can handle the power demand of the work you are asking of it, *at the speed* you need to do that work. (if you have to slow down a lot and still push the motor near it's limits it may overheat).
You'll need to know the specifics of the slopes you must go up, and at least guesstimate the total weight of the entire system, bike, rider, and anything else they'd be carrying, to determine the approximate power required of the system to do the climbing work.
I recommend using the simulator at http://ebikes.ca/tools/simulator.html to see how this works, experimenting with various setups that might fit your needs.

Sounds good but I have no idea what to put in half the parameters :?
Ive been reaserching 48V 1000-2000W motors, but they all look the same so I don´t know if they are what they claim to be.
We just need a motor that beffier than the standart 48V 1000W, that could take more Amps, but still fit in the Berg´s dropout
 
60kph will require a big battery!

I will be the nay sayer based on 60kph on a bicycle frame.
- Brand name bicycles with millions spent of testing and engineering, arent meant to go 60kph with a heavy hub motor and the battery weight. Department store bicycles would be even worse, generic bicycles are the worst of the worst.

Enduro/Stealth box bikes have zero money spent on testing and engineering. What I like about those bicycles is it cleans everything up, puts a big battert box in the triangle part of the bicycle.

The conclusion would be to buy an electric motorbike (like a Zero) that is engineered and designed to go 60kph all day long, and most important is designed and engineered to go 100kph+ to a cutoff of 170-180kph which was what my old 2001 car did back in 2002.

If your bicycle has holes in the bicycles tubing to hide cables then good luck to you sir, I had a top tube crack on me right where the hole was, only had the battery hanging off it, 1000wh battery. Imagine tdoing that at 60kph vs my 25-30kph. I am just lucky I caught it!

My life and my health is to important to me, even wtih full medical coverage and full dental coverage.

Done nay saying, sorry but I feel it as a public service announcement to other readers.
Do as you wish sir.
 
2wGlider said:
" That last is usually a sign of a hall sensor problem. Either a connection issue, a wiring problem, or an actual sensor failure, causing the controller to be unable to determine the rotor position, and so sending the wrong phase order signals and causing it to spin backwards until it reads the other sensors still working normally reporting the wrong spin direction and then it reverses and spins forwards."

When the bike is moving, does that burned sensor affect speed or torque at all? In your explanation you said the burned sensor causes a "misfire" of the controller, so I assume the overall On time would be less now.

I didn't say anything about a burned sensor. Is one of your sensors actually burned up, with smoke markings on the objects around it, and / or wire damage on it from fire?

If so, then it is almost certain the controller's hall sensor input and/or MCU is also damaged, and since that type of failure would need significant current to happen, which isn't available on the hall power or signal wires, it would quite likely involve wire damage between halls and phases in the cabling, and mean a controller phase could be damaged as well, and associated wiring within the motor-controller cabling. In either case, the controller would need to be replaced as well as the sensor and damaged wiring, to fix the problem. In this case, there could be a different amount of "on time" but what it would be would be dependent on this specific controller's behavior in this failure mode (something you would have to test in the failed mode vs the working mode).

If not, then you'd need to test the connections, wiring, and sensor itself to find out which is actually the cause of the problem (if it is even a hall signal issue at all, though that is likely).

If it's just a signal not being correctly received by the controller, then it again depends on the specific sensorless fallback mode behavior of this specific controller as to what happens, but it is likely that it would have the same duty cycle (on-time) sensorless as sensored, after it has gotten spinning fast enough to correctly use sensorless mode to determine rotor position reliably (usually doesn't ahve to spin very fast for this, a few RPM is typically enough).




Where is the sensor physically located? If it is at the windings themselves (the part most likely to be damaged by sudden heat spikes, since they are what generates the heat), then you should see any of those heat spikes on your temperature measurement device, assumign it is capable of reading them (meaning, it can read that high and it refreshes it's display often enough to catch short-term excursions).



Idk, can you tell from this picture?

The temp sensor is glued to the rim, close to the coils but not touching them (the picture was taken before I intalled it).
I do not understand why you'd ask us if we can tell, from a picture of a motor without the sensor installed, where the sensor is?: No one could possibly tell from that picture, if you had not installed it yet, and thus is not in the picture and thus is not visible to tell where it is installed.

Regardless, here are some notes on what the sensor would actually be reading in various situations, so you can decide what you want to measure and do the appropriate thing to get that:

If you installed it not touching the coils, then it cant' measure the coil temperature. It will measure the temperature of whatever it is touching.

if by "rim" you mean the stator lamination stack, (the layers of metal the coils are wound around), then those take time to heat up from the windings transferring heat to them, and so the sensor will more slowly increase reading than if it were measuring the coils themselves.

If by "rim" you mean the black stator support core, then the heat has to first transfer to the stator lamination stack, and then into the core metal, which will be a lot slower.

If it's on the blue plastic shims of the stator lamination stack, then it'll take even longer, as it has to wait for the air inside the motor to heat up and sense the air temperature.

Etc.


You'll get a more average motor temperature reading the way you have it, but you can never see how hot the coils themselves actually get in operation. If that's important to you to prevent motor damage, you'd need the sensor to be touching the coils instead.


I recommend using the simulator at http://ebikes.ca/tools/simulator.html to see how this works, experimenting with various setups that might fit your needs.

Sounds good but I have no idea what to put in half the parameters :?

The instructions are below the simulator, and explain how it works and what things do what in it. There is also at least one thread about it here on the forum, listed in this search:
https://endless-sphere.com/forums/search.php?keywords=*simulator*&terms=all&author=&sc=1&sf=titleonly&sr=topics&sk=t&sd=d&st=0&ch=300&t=0&submit=Search
that show how others have used it; if it is helpful to you. Since I don't know the numbers for anything you have, or any of the conditions you need things to work in, or what you need them to do, I don't know what to put in there for you either. ;)



Ive been reaserching 48V 1000-2000W motors, but they all look the same so I don´t know if they are what they claim to be.
We just need a motor that beffier than the standart 48V 1000W, that could take more Amps, but still fit in the Berg´s dropout
If you don't know how much power you need to accomplish the task you have to do, then it's not really useful to look for motors yet. You may not need a bigger motor at all, or you may need one much bigger than you are looking for. ;)

If you don't care and just want a bigger, better built motor, go to QSMotors and look at their QS205-50H series.

They can make their motors custom for you, just contact them and tell them what you need it to do and what it has to fit in, and they can probably make you what you need. People have used them for 5-digit wattage power output, so they can probable do whatever you want them to, in 4 digits, as long as you get the right winding/etc for your specific usage scenario (speed vs voltage vs torque vs current vs wheelsize etc).
 
Back
Top