Hi guys,we have just read your discussions about our e-truck and that there is maybe a lack of information. Thus we will give you more details about the technic.
We know that the whole truck indicators sounds too crazy, so it is difficult for us to demonstrate the benefits.
We try to explain the whole truck with calculation examples and support it with videos - if our truck is back from an exhibition in Berlin.
First about the batteries:
Fundamentally the energy and the power of a cell goes opposite.
So for range we need energy, and for acceleration we need current. To combine range and acceleration (especially needed at the drive up) in one cell, we develop a serial / parallel switch to have 2S2P for high torque and acceleration at low speed and 4S1P for high speed.
Four batteries are integrated in each hanger of the truck. So we have in two trucks 8 cells in sum. We use the new industrial cell size 2170(0) to get almost the latest and best cells for the truck.
At the beginning we tested our trucks with the LG HG2 (18650, 20A continuous, 3 Ah) and had very good results regarding acceleration and good regarding range. Now we test the Panasonic NCR20700B (2070(0), 15A continuous, 4,25 Ah), cause the focus is to increase the range.
With this cell we generate a conservative calculated capacity of 100 Wh with a moderate cycle life.
At this point we haven’t consider the new 2170(0) cells, which will be available very soon.
EMC
The cells are directly mounted on our PCB inside the truck. So we have a very short distance between cells, EMC and Motor and have only one plug between the PCB and the motorphases. So the wire losses are minimal. We use a four layer PCB with 70 µm cooper, so there is for every single motorphase a own PCB layer with lowest resistance. We tested the PCB successful with 30 Amps (temp rise at 20 A is about 7° C). We have at both ends our motormosfets with a low RDSon of 1 mOhm and a QGD of 13,5 nC for highest efficient.
We have never made a compromise between cost and quality, we almost take the best for the best result (OK, no silverwinding ;))
The PCB includes:
for each motor a mosfet circuit with driver, a back emf measurement, current measurement and capacitor bank
a balancer circuit for each cell with voltage measurement and total current measurement
a cell switch 2S2P <-> 4S1P, to have till about half of speed high torque or if the torque is not required a quarter of the cell losses.
example, 10 km/h and 10A for each motor, 20 mOhm cell resistance
4S1P:
(20 A)² x 4 x 20 mOhm = 32 W cell loss
2S2P
2 x (10 A)² x 2 x 20 mOhm = 8 W cell loss
so the controller switches if possible in 2S2P to reduce internal losses.
an automatic on/off switch with self holding (using the voltage from a rotating motor to switch the circuit on at about 2-3 km/h)
a DC powerline communication, to communicate and power the PCB in the baseplate with only one power pogopin (mass connection goes through the housing like in a car) so you can mount the hanger in the baseplate without plug a cable.
a high-performance, low-power 32-bit controller with 50 MHz to manage all the tasks
and other stuff like voltage regulators and temperature measurement.
All parts are under 1 mm height and so the total PCB height with parts is max. 1,8 mm.
Motor
We use an ironless permanently excited synchronous machine.
This kind of motor has following benefits:
no cogging and very less eddy currents, cause of the less iron in the stator and the thin lamination.
We want to use the e-truck like a normal truck, so we don’t want to use electric energy for freewheeling. This is the only permanently excited synchronous machine who can do this. Other possible motors for freewheeling are without magnets, like reluctant- or asynchron. But these types are heavier at the same power and have no ideal powercurve for our application.
best ratio of weight to torque (in hub configuration without gearbox) cause of the less iron which is needed to close the magnetic circuit.
high efficiency, in our manufacturing process we press the cooper winding with high pressure on the stator and get a coil resistance of 50 mOhm of each phase!
In one motor we generate a torque of 0,06 Nm / A.
The three motorphases would be automatic connected by putting the stator on the truck, so you haven’t connect the motor with a cable.
In sum we get a very high system efficiency, which generates less heat and more range!
PCB Baseplate
In the baseplate we include a second PCB. This small PCB includes the following features:
powerline communication with the hanger PCB, cause the hanger PCB is inside a faraday cage and can’t send via bluetooth.
communication via Bluetooth with the second truck, the remote and possible with a smartphone or -watch
handle the USB PD protocol to unlock the PD charger. We can charge the e-truck with profile 4 and 5, this means 3 A @ 20 V or 5 A @ 20 V (look for further Information at USB Power Delivery)
Charge inductively the e-truck. We integrated in the cover a primary Qi coil. With this coil the board could be charged inductive. The holder is under development, but other tasks have actual higher priority.
As gadget you can charge your qi-able phone if the truck is drop through mounted, otherwise you can charge via the type C connector. Both connections are limited with 5 V @ 1 A. We see this connection as a emergency charge station, to take a call.
NFC chip for fast Smartphone connection and fast Truck profiling
Truck housing
The truck itself is made of a high quality aluminum alloying, which is normally used in automotive axis. Unsuccessful we tried to break it with a 2 t car jack.
The truck is housing and cools the rest of the motor and electronics.
Wonderful explanation - although to be perfectly honest I believe both 2s2p and 4s1p even in a 4wd setup is much too weak to impress anyone on this forum (we are used to 6-12s and 3-8p). What I do find impressive (and I’m sure others can agree) is that you have taken the initiative to respond to each and every one of our criticisms in clear detail, and I think that this shows extremely good communication and reliability on the part of your company. Good luck to you, and I wish you the success that you hope for your product to achieve.
Actually I am fascinated by how much electronics you apparently put into such a tiny footprint. My Antispark alone is apparently larger than your whole electronics setup
4,25 Ah x 3,6 V x 4 cells = 61,2 Wh per e-truck. Means 122,4 Wh with both e-trucks as maximum.
In this case we “only” use 80 % of the cell capacity, about 100 Wh, to guarantee the 12 km range and maximize the cycle life of the cells.
When we charge the e-truck via USB, the cells are switched in 4S1P, so the maximum cell voltage is 4,2 V x 4 = 16,8 V. For this we have on the PCB in the hanger a step down converter.
i will load up a picture of the PCB and explain the components
you can recuperate and push energy back to the cells. The energy from the motors go through the charging circuit to regulate it. Limited is this “brake” force by the maximum charge current of the cells and the cell state. If the required “brake” force is higher, we burn the current in the motorcoil till the temperature is too high. This goes very fast (depending on brakeforce) and then there is no more possibility to break. We thought about a brake resistor, which is cast in the aluminum hanger, but this would only put off the problematic to handle the heat.
Fundamentally a “brake” via a recuperate isn’t a real break cause it depend on cell voltage and the system temperature.
A brake have a defined brake torque and work in every case, so we do not brake, we decelerate if possible.
So the driver of a electric board without a brake should be able to brake by himself.
We fit four cells with electronic in each truck.
Tomorrow I open for you the accupack to show you who it fits. Cause our board is exposed I send a uphill video as soon as possible.
i don’t understand as you say the motor will start to be powered if it is moved by pushing right? and then once you’re moving it will continue to move until you brake as done with a non-electric board, which I guess means put your foot down and drag it, but I must be missing something as how is the motor to know if it’s an uphill resistance or the rider putting their foot down?
With the USB Power Delivery it is,
There is a controller in the device and one in the power adapter. The device sends a request what it need for a profile (means voltage and current) and the power adapter send a answer message with his possible profiles. So they handle out and start charge if they match. For profile 5 (meanse 5 A) the active cable must be detected and these cabels are build for 5 Amps. So you can charge a Smartphone, which needs 5 V or a notebook which needs 20 V with the same adapter without destroing one of the devices.
Sorry, I misunderstood.
You control the board with a remote.
The remote has three basic functions: accelerate, decelerate and freewheel.
You push once the board, then only the electronic circuit switch on. Now you can accelerate with the remote or push further like a non electric board.
If you want to stop you decelerate till you stop or you choose freewheel and brake by foot.
This drive is comparable with other drives in the 100 Wh class.
You can push the board and assist if you don’t want to push at long distances or uphill. Futher you can drive pure electric. It’s like a hybrid with foot and electric. For bashing around the streets with raw power, the capacity is to low like other drives in this class. It’s perfect for cruising through the city or on the sea promenade.
Nice work, as I understand two of the cells goes inside the motor? The larger diameter section of the hanger is where the motor goes? What about heat on these cells? Lithium degrade faster at higher temperatures, and a hub motor will be severely limited if you have to keep the current down to prevent cell overheat, not saying it doesn’t work, I just want to hear what was the thinking behind this setup
Hello Pedrodemio,
yes, the both outer cells are inside the motor. We know from the beginning that we had a challenge to handle the heat, cause our electronic and the motors are directly mounted near the cells. Because we want anyway range and no heat, we spend a lot of time to develop a high efficient setup.
So we designed the PCB circuits so, that the power loss are low as possible.
We use four motors, so there is half of the current compared with a 2WD and so we have a quarter of the losses per motor.
We designed the Motor that we could implement a winding with only 100 mOhm. For this we press under high pressure a litz wire (although to minimize eddy currents) on the stator and use in the rotor N52 Neodym magnets for a strong magnetic field. At nominal voltage the back inductance of the motor limit the maximal speed at 35 km/h. So you see the magnetic flux can’t be higher!
Cause of the less stator iron this motor type have much lower impedance compared with other synchron motors.
Additionally we use between the stator iron and the aluminium hanger an aluminium spacer, to lead the heat to the complete hanger.
So we don’t create heat hotspots in the coil like other hub motors.
Cause of the thermal conductance of aluminium the heat spreads, so the temperature between the inner and outer cells are only a few degree Celsius.
The result is, that we have at our last test ride a hanger temperature at about 40° C. It feels only warm.
At the promised uphill test ride, I take pictures for you with a thermal camera.
We use only the iron that there is no saturation of the flux.
Or do you mean why this motor type need a lower quantity of iron?
If we use more iron, there would be a slightly higher flux and a noticeable heavier engine.
Although the motor would slightly increase the torque and degrees the speed.
If you design a motor, there are many parameters to adjust.
If you change a parameter on the one side, there is a change on some other sides.
So we give the parameters priorities regarding the engine requirements.
I’d think for a hub motor especially you’d want a high magnetic flux that would come from iron at the expense of weight. Is there any iron in the motor on the stator or rotor side? I’d think you could do a hallbach array on the rotor and avoid any need for iron but the stator side it seems unavoidable if you have so slow a motor with no gearing to have torque.
Can u show pics of the motor? And how it attaches?
