08/09/2018 at 20:06 •
The new PCBs have arrived! The rest of the components should be delivered by next week
The silkscreen on the coil traces looks a little crappy but its my fault for not checking the manufacturer's print resolution.
- 08/09/2018 at 00:19 • 0 comments
07/15/2018 at 06:23 •
This is the drone concept that gave me the idea of trying to create the PCB motor and spherical folding propeller! Both of these projects need a lot more work and improvement to make this drone feasible but at least now I have a starting point :)
07/12/2018 at 23:24 •
I finally ordered a new PCB motor with an integrated ESC. I have managed to package the circuit in a very small space, 30x16mm including the stator. For now I have ordered the same star winding configuration that i've used in the first prototype. I'll be ordering more configurations soon.
The circuit basically consists from:
Hall Sensor (US1881) to detect the magnets inside the rotor.
- MCU (PIC16F1503) - I have shifted from the DSPIC33EP128 to the PIC16F1503. This MCU has less computational power (not much is required since I am no longer considering sensorless control) but is packaged in 3x3mm chip and is around $2 cheaper
- 3-Phase Motor Driver (STSPIN230) - which is rated at 1.3Arms, and has several types of fault protection build-in.
- A filtering circuit to supply the micro.
I will upload all the source code and gerber files once the PCB arrives and verify its functionality. Keep tuned!
06/19/2018 at 23:49 •
So torque is the biggest weakness of my tiny PCB motor. This was measured it to be 0.9gcm.
But it is something that can be improved. These are three ways how I tried to improve it:
- Double Rotor - Some people in the comments suggested to try and use a double rotor. This will increase the magnetic field produced by the neodymium magnets and therefore will also increase the motor's torque. But in practice this was not the case and it barely had any effect. The measured torque was 0.9gcm (the same as with one rotor).
- Ferrite Sheet - This was used as a core to increase the strength of the magnetic field inside the printed windings. Two different shaped cores were tested. The uncut one gave a higher torque value and has increased it from 0.9gcm to 1.5gcm. The only down side of this is that It has also increased the pcb's temperature from 70°C to 90°C due to eddy current losses.
- Delta configuration - Other simple way to increase the torque is by changing the configuration to a delta winding. This way the coils will be powered with a higher voltage, so it will also increases the torque and temperature (hopefully not by much). This approach was not tested yet.
Check out the full tested video:
06/08/2018 at 22:35 •
A quick update on my PCB Motor prototype:
- Dimensions - 16x17x5mm (excluding the shaft)
- Weight - 1.6grams
- PCB Stator - 6 Poles connected in a delta configuration, with 40 turns each with 4/4mil traces.
- 3D Printed Rotor - 4 Poles
- Bearings - Stainless Steel
- Phase Resistance - 19.3 ohms
- Power Rating - 1W
- Maximum Temperature - 70°C (at 5V)
- Maximum Torque - 0.9 g.cm
Improvements that I'm planning to test:
- Increase the winding's magnetic field strength by reducing the PCB thickness
- Use ceramic instead of stainless steel bearings to reduce friction
- Experiment with different trace winding's clearance and thickness. Increasing it to 5/5mil from 4/4mil will reduce the pcb's cost at low volumes. However, this would also enlarge the motor's diameter to reach same parameter's of the 40turn 4/4mil coil.
05/27/2018 at 21:21 •
How hot is my pcb motor getting? With a direct 5v supply on the windings it reached a maximum of 70°C after 13 minutes of continuous voltage.
How to improve this? For the next prototype I'm going to use more vias and a copper plane on all four layers to conduct heat on a larger area.
03/27/2018 at 20:09 •
My plan for speed controlling the PCB Motor was to implement a sensorless back-emf speed controller, which works just like every other brushless ESC. It measures the time it takes to detect the zero-crossing point from the under-driven phase, and adjust the commutation waveforms according. However, during testing the back-emf generated in the windings of the PCB motor was a little weak.
Plan-B is to use a hall sensor to implement the closed loop speed controller. This will be a little more pricey but will also include positional sensing.
03/05/2018 at 00:33 •
This is my new 3D printed rotor prototype with a snap-hook shaft and press-fit circular magnets!
Unlike the old prototype, the new rotor has a uniform magnetic field across its whole area.
The metallic shaft was also eliminated. Instead it was made part of the 3D rotor model, At the end of this shaft are two snap-hooks that lock the rotor inside the bearing.
This feature makes the brushless PCB motor cheaper and very easy to assemble!
02/25/2018 at 13:58 •
This video briefly describes how my brushless PCB motor works.