help on starting a project on BLDC

[pmm]

Hello every body
I told to start up a project on BLDC motor postion control system this
would be my first project based on motor control,can any body please
guide me in some aspects which are needed to start the project ,I have
browsed through the net and downloaded few of the documents on BLDC
motor makes ,few on PID implementations . My design should be based on
CAN and Devicenet protocols . the Motor should be 3-4amps 24v . the
thong which is stopping me is where to start and what to start i cant
run into it blindly with out a proper direction . so i am looking into
this group to give me a start up idea of designing it and embedding
code into it

the aspects are what encoder shall be selected for the protoboard shall
be cheap and easy to work on
do i need the tachometer or just counting the encoder reply will do
and any suggestion on your own which help me to execute this task

thanks & regards
PMM

 

[pmm]

I am adding one more

how to test the algorithms is there any demonstration board existing
,can one provide me the name of it and the provider

 

[Paul Burke]

Hello every body
I told to start up a project on BLDC motor postion control system this
would be my first project based on motor control,can any body please
guide me in some aspects which are needed to start the project ,I have
browsed through the net and downloaded few of the documents on BLDC
motor makes ,few on PID implementations .

Good start. The BLDC motor is just like a brushed motor except it lacks
the mechanical switch to do the commutation- you get a set of rotor
position feedback optos so you can do it electronically. Google for
'brushless motor controller' and you should find a bunch of devices to
make this easy. They'll also have design examples with them.

The hardest bit in hardware will be making an output stage that doesn't
go pop too easily.

the aspects are what encoder shall be selected for the protoboard shall
be cheap and easy to work on
do i need the tachometer or just counting the encoder reply will do
and any suggestion on your own which help me to execute this task

That depends on how much control you want. The motor feedback is coarse,
just a few changes per rev, so speed control can be expected to be
lumpy, and position control coggy. smooth and/ or good positioning, and
you'll need an encoder with appropriate resolution - NOT a tacho for
position control.


Paul Burke

 

[pmm]

Hello Mr Paul
Thanks for your reply ,
I was reading thru the PIDs questions posts
in the forum,and I feel like its pretty tough to implement them , how
do I test them, is there any demonstration tool ready so that I can get
one to test them.
any ideas from your side on those algos , will help me a lot

thanks and regards
PMM

 

[Paul Burke]

I was reading thru the PIDs questions posts
in the forum,and I feel like its pretty tough to implement them , how
do I test them, is there any demonstration tool ready so that I can get
one to test them.
any ideas from your side on those algos , will help me a lot

PIDs are really easy to implement, it's the tuning that's hard! But a
lot depends on your hardware/ software setup. With a really fast
processor, and plenty of resolution on the DAC (or PWM or
whatever)things get easier.

Proportional? Subtract position from setpoint. Multiply by gain. Output
to DAC.

Integral? Calculate error as above. Add to integrator Don't let it
overflow so keep it within a settable limit (both ways). Multiply by
integral gain. Add to P and output.

Derivative? Subtract last error from current error. Multiply by D gain.
Add to the others and output.

Think about the resolution required- typically you will be talking about
16 or 32 bit bit setpoints and feedback, 16 bit error, 32 bit
integrator and 16 bit gains. It's useful to have some idea what gain
range and resolution you will be using- so you can shift the multiplied
results appropriately.

The best demonstration tool is a motor, encoder, amplifier and processor!

Paul Burke