This page is still under construction!
I am living in a country with right-hand traffic – car steer left – so…
- This page is about the conversion of a set of two ATB’s to e-bikes or ATEB’s.
- The bikes are described here: Review ATB low cost high fun.
- This project is really fun, but I am having my share of trouble too, like a complete destroyed battery pack and wrong design choices.
- I was really not informed about e-bikes at all before.
- This page should give people planning something similar a basic understanding.
- It will take some additional hours to prepare your own project – unless you go this way…
Table of Contents
A few things are written on other pages, otherwise it gets too long.
- 48V_Wiring About wire choice, connectors
- Li-ion_Batteries Theories
- Battery_Management_System_BMS Theory and tips
- Programming_your_e-bike How to
- Bike_things_that_went_South The things that went wrong or will be changed next time.
- Bike_things_that_went_North The things that went surprisingly well.
What we want
It turns out that “what we want” is different from “what we get”, read on.
- A cost effective solution. Two parts are not cheap: the battery and the ordered wheel assembly. It turns out that all bells and whistles are a considerable part of the total sum.
- A modular approach, all components should be easy to service, detach, attach.
- Rear drive, keeping existing brake disc and spur functionality.
- Actually, front drive is easier to service and for very low power applications (equals low weight) an option.
- Enough power for terrain, steep hills, 500W is not enough. A powerful brush-less motor (1500W is barely enough for a 10% hill) has a longer lifetime – if not abused. High torque is important. Alternatively, a geared middle motor is an option for slower riding with steep slopes.
- Limited total weight gain. That one is hard to achieve, but not an excuse to just add unlimited kilograms.
- A battery with spot welded 18650 elements, powerful for a proper range.
- Easy conversion. But tinkering will be necessary from time to time.
- Proper quality. Time will tell.
- Support when pedalling but also power without pedalling.
- This is not an excuse for irresponsible behaviour, when driving more than 20 km/h:
- Technical defects like tire punctures or a breaking fork become life threatening.
- Fellow road users do not expect speeding e-bikers so please, copy what I crammed in my head: “I am a cyclist and I am extremely vulnerable”.
- It is just very, very relaxed!
- This is not an excuse for irresponsible behaviour, when driving more than 20 km/h:
Things to consider
- About Safety: 18650 cells are dangerous, DON’T UNDERESTIMATE! Shorting can result in explosions. Work with batteries with safety in your mind.
- About Energy: Wh (Watts during an hour) are more interesting than mAh (milliamperes during an hour).
cell of 3.7 V and 3400 mAh does have an amount of energy of 3.7 * 3400 /
1000 = 12.6 Wh. This means that this cell can deliver 12.6 Watt during
an hour. That is 3600 seconds, equals 12.6 * 3600 / 10e3 = 45 kJ (or
0.045 MJ). Let’s consider two cells, totalling 25.2 Wh:
- In series two cells are still 3400 mAh, but 7.4V.
- Parallel two cells are 6800 mAh, and only 3.7V.
- So, concluding, the amount of mAh is only interesting when you know the voltage.
you build a battery of 48 V, based on 3400 mAh cells, with 52 cells,
the amount of Wh is 3.7 * 3400 / 1000 * 52 = 654 Wh, not bad!
- Calculating backwards, this equals a 48 V battery of 654 Wh / 48 V = 13.6 Ah (Amperes for an hour), or,
- this also equals a 36 V battery of 654 Wh / 36 V = 18.2 Ah.
- To give you an idea, it also equals a 12 V car battery of 54 Ah.
- And that is the reason to stop thinking in Ampere-hours and always calculate Watt-hours.
- A cell of 3.7 V and 3400 mAh does have an amount of energy of 3.7 * 3400 / 1000 = 12.6 Wh. This means that this cell can deliver 12.6 Watt during an hour. That is 3600 seconds, equals 12.6 * 3600 / 10e3 = 45 kJ (or 0.045 MJ). Let’s consider two cells, totalling 25.2 Wh:
- About Weight: a cell is <50 grams. 52 cells are ~2.5 kg without housing and BMS. Not that much.
- About Nickel: A word about nickel conductors:
- Take care that you use pure nickel and not nickel plated metal.
- About “how much” nickel: 1mm2 is okay for 5 Ampere, 7.5 A as max, but 10 A gets warm.
- A strip of 7 mm * 0.15 mm = ~1 mm2.
- A 1.5 kW motor on ~50 V is peaking at 30 A. So you need 6 mm2 conductors, at least 4 mm2 if peaks are very short.
- It may help you to draw arrows on nickel strips to see how currents are divide over strips, in order to find weak spots.
- If you need more conductivity, just layer up extra nickel.
- Nickel is not used primary used because of it conductivity, that is a common misunderstanding. Resistance in ohm.m: Copper = 1.68*10−8, Aluminium = 2.65*10−8, Nickel = 6.99*10−8 (Iron = 9.71*10−8, Steel (carbon) = 14.3*10−8, Steel (stainless) = 69*10−8).
- About Thiefs: Consider a good lock. Definite not a cable lock. Abus Bordo can be considered as an example of improved security.
- Consider GPS trackers…
- About choosing between aluminium and stainless steel for constructions:
- Aluminium is light, easy to glue, strong, easy to machine, hard to weld, resulting in weak spots after welding.
- Stainless steel is work-hardening, harder to drill, easy to weld with reliable results, not so strong per kg.
- The battery compartment is an excellent example of how to combine both materials, getting best of both. A stainless steel 304 space frame with an aluminium almg3 battery case.
- Work-hardening can be a PITA. For drilling it is important to cut enough material. So use enough force, combined with reduced speed (r.p.m.). Also, use sharp drills. Grinding your drill tip is what I found to be the best solution. Top angle: sharp, not 120 degrees default angle, books say otherwise. Cut your drill short to prevent slip stick (and hardening). Excellent reading: https://www.bssa.org.uk/topics.php?article=194. Drilling a hardened object? I’ve had a little result by heating with a TIG torch. Grinding towards the right diameter is a last resort. Having short drills with different top angles result in small initial (hardened) contact surfaces.
On Aliexpress.com, there are not many suppliers of complete wheel sets. I ended up with three options, one 1500 W from Conhismotor and two from Pasion E Bike, 1500 and 500 W. The 500 W solution was a Bafang motor. What kept me away from it was the small size in relation to heat production. There are several stories about this on the net. So the final solution for choosing a 1500 W assembly was based on price and wheelsize. Both directed me to Pasion Ebike on details (https://www.pasionebike.com). The price difference was very small and in my case, a 29″ wheel was offered. The difference between 700c and 29″ is merely the width of the rim, 29″ is bit larger while rim diameters match.
Pasion contacted me and proposed to deliver from Germany to the Netherlands. Great news, no extra VAT and import duty’s, paying ~100 euro’s less. It took only 5 working days to receive all. General impression: More than okay. Spoiler: Nah… Just check your package
How much power, for real
I may be wrong, but I read somewhere about a torque of 35.5 Nm. 29″ wheels with a radius / arm of 0.37m (=29/2*25.4/1000) equals someone pushing in your back with 10 kg (=35.5Nm/0.37m/9.8m.s-2). That looks like a lot. But, we’re climbing a hill… If your weight plus bike is 100 kg – the vertical force, and the horizontal force is 10 kg, this is barely enough to get you up on a 10% slope. So 1500W without gears is not a luxury.
In addition, I see power spiking to 1800W with charged batteries, good!
For now I am going to use the included tire, it is certainly not bad, a Maxxis Pace, 29″*2.1, max pressure 4.5 Atm, 65 psi. But it will become the new front tire. Current tires: 29″*2.35 for mud processing and I am going more and more “on road”. Now about the rear tire: I like tires like Schwalbe Marathon MTB because profile on a straight road is not interrupted, their mileage is okay, anti leak features are nice too and they are “e-bike prepared” with an ECE-R75 mark (another hobby from Brussels). By the way, an instant repair can, like “Cyclon Tire Repair Gel”, is always on board. So I bought these to replace the current tires on the rear. I choose 2.1″ width but they come in 2.25″ too with a higher load max and probably a bit more comfort and less direct steering. More on https://www.schwalbe.com/en-GB/offroad-reader/marathon-plus-mtb.html and http://www.maxxis.com/catalog/tire-482-121-pace
Save diameters based on 50 V:
- Copper: 10 A / mm2. Result: 500 W /mm2.
- Nickel: 5 A / mm2. Result: 250 W /mm2.
Lighting is for convenience but also for safety. With a big battery comes great power, so let’s use it. Here are the parts:
- Low beam
- High beam
- Rear light
- Brake light
Bought this one at https://cnsunnylightmotofathion.aliexpress.com/store/1735677:
It has a wide voltage range and a high power. The beam is also very good: Field of view is approx. 40 degrees and it goes up gradually to 0 degrees. Nice! On top of that, there are two power modes: black+blue=6W, black+red=15W. Switching will be one switch for on and off and one for 6W and 15W (using two “on-on-switches”). Oh, did I tell it is only 170 grams?
Some things are hand made:
- Battery compartment
- Battery housing
- Brake switches
The challenge is to keep weight under control. And weight gain is large for the bigger parts. They need extra attention. For example, for the battery housing I have aluminium sheet here, 3mm. But falling back to Aluminium AlMg3 (5754) is tempting: With better strength characteristics I can fall back to 2mm thickness.
Now this is a challenge because there is no out of the box way to attach something to it. It it gets a bit more complicated because we need two switches per brake: One for the engine stop and one for the brake light.
For the brake light, water resistant SPDT (single pole, double throw) micro switches are used. Very small, very nice.