[kreutz]

Hello:

Yesterday I received one BLDC motor and controller set. I ordered it a few days ago ($90.00 + tax) just to find out how to modify it for my own purposes.
(Tried to post the photos but they are to large and I will have to reduce their size)
The idea is to use, if possible, some or most of, the original components in order to make it do what I want, without having to completely redesign the controller board and user interface. But, it all depends on having access to each individual component's data sheet and finding out how the circuit works, before deciding what to change.

Specs known so far:
550W on the motor label.
110Vac line input
500 - 4500 rpms controlled by a hall sensor-magnet contraption. (yes, it is one of the myriad "servo controllers" for a sewing machine floating around the internet) Not a real servo, of course.
Real BLDC motor specs are completely unknown. We will have to measure them, but first, I have to reverse engineer the main board and the user interface board. Apparently not a difficult task.

What I want it to do:

Isolate my control signals from the controller which is not isolated from the power line.
Controllable by step-direction, or -10V/10V signal in CNC mode
Manual Switch for forward-off-reverse, and a Potentiometer for speed control in Manual mode
Emergency stop (both remote and manual button)
Motor tuned by isolated usb connection to PC, if possible, using auto-tuning for the PI gains of the speed and torque control.

I know that some of you will tell me to use the magnetic contraption as seen on you-tube videos, and forget about all of the above, but that is where the fun is!

Stay tuned....


kreutz

[kreutz]

Photos:

[kreutz]

The controller board has:

line voltage input connectors, fuse, bridge rectifier, 2x 680 uF/250Volts capacitor bank, 1 dc/dc converter ic (Viper22a-E) and associated circuit for +15Volts power supply, and a 5 volts regulator ic.
3 channels IGBT driver IC (ID7T6036), 6 x 20A/600V IGBTs (Xiner XNF20N60T)

1x 0.015 ohms 3 Watt sense resistor for motor current overload measurement, 1 x 510 ohms 1 Watt resistor (seems to be part of a soft start circuit to charge the capacitors bypassed by one of the relays when the 15 volts is present).

1 x 32 pin microcontroller IC (MM32SPIN05PT 32-bit Micro controller based on ARM Cortex M0 , 32K flash, 4k Ram) the only thing I know so far is that it is supported by Keil.

The heatsink is pathetically thin.

[kreutz]

Photo with labeled components. Bottom left are the two relays.

[Al_The_Man]

Seems to be a little bit of a masochistic task
If you intend to modify the board, I would have thought that a program listing of the 32-bit Micro would be needed.
It may have been easier to design from scratch, Incidentally, what pole count does the motor have?

[kreutz]

6 magnets, 3 pole pairs.
I don't think the program listing will be available, Looking at the microcontroller's datasheet, there is little chance they are doing anything else than 6 steps commutation, using the 3 hall effect position sensors mounted on the stator, trapezoidal PWM speed control or in the best case sinusoidal commutation, . So, starting writing code from scratch would be better than to wait and pray for the source code.

Resistance measured between two phases (Y connection) = 5.1 ohms (including cable and connector resistance), Inductance @ 1khz between two phases = 35-36 mH, inductance changes with rotor position.

Regards,
kreutz

[kreutz]

Found another guy trying to reverse engineer a similar BLDC controller, possibly an older model of the one I have. Look at: https://www.candrian.gr/index.php/re...or-controller/

I am currently trying to reverse engineer the PCB and produce an schematics, also connecting test points for testing the gate driving signal waveforms. For anyone trying to do the same, please, remember to use an isolation transformer (1000VA minimum) to feed the controller, there is a potentially lethal line voltage present at the pcb ground otherwise!! Protect yourselves and your test equipment!

[kreutz]

The relay to the right on the pcb (top side) is used for the voltage doubler function for 120Vac operation, the voltage doubler raises the filtered rectified voltage to approx 310-320 Vdc, the two big capacitors are in series connection across the 320Vdc bus. In case of 220Vac line input the relay on the right is not populated, and, of course, there is no need for voltage doubling.

Here is a photo of the bottom side of the pcb:

[kreutz]

The documentation on the micro controller's manufacturer page is mostly in chinese, no english translated useful documents on their page, apparently they have their own ISP programmer, MMlink emulator and programmer, but I wan't able to get more info, not even a user's manual. So, the option of replacing the firmware on the original pcb's micro-controller is closed, so far.
I have a few ST Semiconductor's evaluation boards at hand, as well as their high voltage motor evaluation board, maybe that is the way to go and let's forget about modifying the original board for the moment. I checked on the internet and the ST micro controllers, on the evaluation board I have (STM32G474RE), are "out of stock" . So it will take time to make prototyping boards and populate them unless I desolder the micro controller from the evaluation board to test my prototype pcb. I am retired, so there is no hurry at all.

I'll keep you posted, on the status of the project, from time to time.

Regards,

kreutz

[kreutz]

Here is the photo of the controller's front panel. I just realized that there are many of this sets for sewing machines around but there are differences between them.

[kreutz]

Here are the ST evaluation boards I was talking about, I will be starting to write software for this project in a little while, just waiting for parts to repair my 400Vdc/5A lab power supply to start testing step by step. Most probably the first try will be Field Oriented Control of the BLDC motor, will try "sensored", using the 3 Hall sensors present in the motor, or "sensorless" and probably use the sensors for rpm display only. There is time to also try the 6 steps commutation, similar to the actual implementation, just to compare them. I also need to test my low voltage lab power supplies, more than 10 years without use won't make them any better, don't want to risk damaging the eval. boards.

[kreutz]

I was prepared to give a large speech about why "NOT ALL THE PERMANENT MAGNET BRUSHLESS MOTORS AND CONTROLLERS ARE CREATED EQUAL" and describe the different method of "Conmutation Control" available for BLDC motors and their advantages and disadvantages. Terms I already used on the prior posts, such as "6 Steps Conmutation", "Trapezoidal Commutation", "Sinusoidal Conmutation" and "Field Oriented Control" (FOC) need an explanation so you all can follow this thread without overloading your brain with unknown expressions. I will, in the following posts, keep out all the mathematical part and try, as much as possible, to make it easy to understand what is involved in the development of a controller for BLDC motors, in Positioning, Velocity and Torque applications, as required by CNC machines.

Many will think that having a BLDC motor and the cheapest controller in town is a step in advance compared to the induction AC motors employed traditionally, or DC brushed motors which became a trend 10-15 years ago, there are many marketing promises as: "high torque at all speeds", "variable speed advantage without changing gears", which are not always what yo get when you replace your AC induction motor by a DC brushed or BLDC (Brushless DC) motor. Maybe after reading my prepared speech you will be able to understand that different controllers cost have a correlation with different performance as servo axis motor or spindle motor in your machine, and how to analyze, based on the controller manual programming options, if the vendor is inflating the cost but giving you an inferior product disguised as a servo or "idealized" spindle marvel.

I gave up on my speech when I saw the following video, it saved me a lot of time, and it saved you from a lot of reading and photos. It is important for CNC users to understand the implications of a term used in the video: "Torque Ripple" , since it influences the finished surfaces of the machined parts and the vibrations induced in the machine tool. Here is the video:

https://www.youtube.com/watch?v=InzXA7mWBWE

[kreutz]

Let's start with a block diagram description of the future BLDC motor controller.
It will need:
- Power cord receptacle
- Line EMI and common mode noise filter
- Bridge rectifier
- Inrush current Limiter
- Capacitor bank (could be only one capacitor, Two capacitors in series connection, or a real capacitor bank composed of more than two capacitors in series/parallel) depending on the final output power limit for our design.
- Output 3 phase bridge (discrete IGBts, or Integrated power module)
- Phase voltage sensing (resistor dividers, filtering and ESD protectors)
- DC link Voltage sensing (resistor divider, filtering ). Having a DC link voltage sensor allows us to compensate (in software) for the line ripple voltage (which affects "torque ripple") and also protection against DC link over-voltages due to sudden braking an inertial load by using freewheeling or synchronous/quasi-synchronous rectification in the output 3 phase bridge. we will discuss this situation later. This software compensation will allow us to reduce the total DC link capacitance a 20-25% with respect to the traditional designs which calculate the total capacitance based on a 5% - 10% voltage ripple at the output. Dc link capacitance will also be influenced by the motor braking method employed in our design.
- External interface to an external brake resistor (optional)
- 3 phase bridge gate drivers (either individual half bridge drivers or integrated 3 x half bridge drivers.
- Motor coil(s) current sensing and amplifier(s) (could be one or two sensor/amplifiers)
- Microcontroller (requirements depend on the commutation method(s) employed in our design)
- External control interface(s) and the need to electrically isolate the interface from the board's ground, including also the PC communication interface for setup/tuning.

All of the above will be calculated based on our BLDC motor and the data sheets of the components used, Motor parameters will be previously identified using the evaluation boards. So at the time of designing the final pcb those parameters will be known to us. They are required for both, the software and hardware design.

[kreutz]

Let's also be clear from the beginning: I DON'T SELL ANYTHING, NEITHER I WANT TO MAKE THE CHEAPEST, MORE ECONOMICAL SOLUTION FOR "ANY" BLDC MOTOR CONTROL. BLDC motor controller design and motor tuning require the motor parameters to be known beforehand, so this is not for the one of the kind unknown motor, you will have to identify yourself your motor parameters before tuning or are in a painful road to tuning it manually. The more complex the commutation method is, the more you have to know about your motor and application parameters such as rotor inertia, friction, load inertia, motor q and d inductances, coil resistance, speed constant, torque constant, etc.

I welcome all of you to this thread and will be willing to explain everything I do and also explain the reason of my design choices, but I won't provide schematics changes for your particular components, you will have to learn to design the changes yourself, I will help you and give you links to design information you will need in that case. In any case, if you ask me a question I can't answer, it will open an opportunity for both of us to learn something new. All the design calculations will be performed by you

At the end of this project, and after tests have been done, and posted for all to see, I will post the schematics, source code and gerber files for anyone who wishes to reproduce this design. There are lethal levels of voltages present in the main pcb that will be designed, you must know what you are getting into and how to protect yourself and your property, I won't be responsible for anything!

Meanwhile I started to post in this thread as a means to document this project for my own use later, unafraid of pc crashes, since this thread will be available as a project log. Previous published designs such as the HP-UHU brushed DC servo drives, and the Mardus-Kreutz Unipolar micro-stepper drives, were lost in my PC crashes over the years, the only surviving information is in the related forum threads and I am glad of having posted most of the details there.

I won't bother you posting design equations, those will be in their corresponding spreadsheets. For those interested into using them for their own purposes or learning, I will make them available after testing.

Best regards,

kreutz

[Al_The_Man]

Are you designing for a BLDC motor, or a 3ph P.M. motor ?
Although the two are almost identical in construction, the commutation is quite different.
In a BLDC motor, only two windings are under power at any one time, (hence BLDC).

[kreutz]

BLDC motor controller, not (IPM) 3 phase motor controller, although they both could be driven with the same hardware.

That is the reason why I only need 1 or 2 current sensors, in 6 steps commutation (trapezoidal) one phase is always high impedance while the other two are driven, so the current in that phase is zero, so only one current sensor is enough, in FOC with SVPWM (Space Vector PWM) all 3 phases are active all the time but the current in the 3rd phase can be calculated knowing the sum of the other 2 phases currents.

3 phase Induction motors, as well as PMSM and BLDC, all of them use a 3-phase Mosfets/IGBTs 3 phase bridge output.

In general, the same controller hardware (with different firmware) could be used to drive 3phase IPM motors , PMSM motors, BLDC motors, and Permanent Magnet Brushed DC servo motors, the only difference is the firmware.

[kreutz]

All work on this project will be stopped for a little while due to hurricane Ian interference, I need to secure and cover everything in the workshop, just in case. Power has already failed two times, the last time for eight hours. Internet connection is still miraculously surviving, but as slow as ever (I live in the country side, no fiberoptic communication around here, a slow (2-3 Mbps) DSL connection over phone lines is all I get), very precarious, a little storm around and it is gone for weeks.

Regards,

kreutz

[kreutz]

The workshop survived the hurricane, now I need to re-organize everything in order to continue working on this project.

Since it is not a "good project" if it doesn't generate at least one sub-project, I need to design and build an electric motor dynamometer in order to evaluate the BLDC motor performance. That will also take some some time and extra repairs (my X1 micro-mill is in need of an overhaul). For a load I am considering using a good AC servo motor and drive (PMSM). Permanent Magnet AC synchronous motor have a very smooth torque curve and almost no torque ripple due to the inherent mechanical construction of the rotor and stator, the more pole pairs the better. That will also solve the positioning encoder problem since it is integrated into the servo motor. The only other optional detail is to get a "Rotary Torque Transducer" in order to directly measure torque instead of having to rely on the AC servo current and the torque constant specified by the manufacturer of the ac servo motor.

kreutz

[kreutz]

Test instruments and materials list (so far):
-3 isolation transformers
-Multimeter
-Low voltage lab power supply
-High Voltage Lab power supply
-ST-linkV2-Isol Isolated debugger/programmer
-USB Isolated interface (USB2- HighSpeed) 5000V isolation
-ST-Nucleo 64 F401 evaluation board (used for BLDC motor parameters' Identification only)
-ST Nucleo G474RE evaluation board (used for FOC firmware development and testing)
-ST IPM10B High Voltage (400V 13A) BLDC motor inverter evaluation board (for Motor parameters' ID and Testing)
-330uF/450 Volts DC link capacitor for the above mentioned board.
-2 channels 100 Mhz oscilloscope (It would be better to have a 4 channel oscilloscope, but this one will do)
-Current probe (ordered)
-4096 ppr (4x) quadrature encoders for testing and motor ID, the end product will use the 3 Hall sensors or be sensorless.
Also:
-TI Launchpad-XL F280049C evaluation board
-TI Boost-XL -DRV8320RS motor power evaluation board (for Low Voltage motors (54Volts-15Acontinuous-20 Amax or less))
Just in case the TI parts are available while the ST parts are still on backorder, None of them are available at this moment.
I have both, the TI development software and the ST development software.

Still working on the dynamometer...

[kreutz]

I forgot to add the following to yesterday's list:

-ST-IHM09M2 board, it is an interface board between the ST-Nucleo boards and the ST-IPM10B power board.
-High Voltage Differential input probe for the oscilloscope. P5205A or equivalent (x50 and x500) must isolate over 700V, better yet 1300Volts.