RC Groups

RC Groups
    Power Systems
        My own design Sensor less/ Brush less/ Slot less Motor and driver

#1 Takao Shimizu Nov 05, 2002 07:39 AM

My own design Sensor less/ Brush less/ Slot less Motor and driver
 
1 Attachment(s)
The schematic and software source code are;
http://www.rcgroups.com/forums/showt...96#post1536496

http://www.rcgroups.com/forums/showt...27#post1530527

The motor case is plywood and fuselage is by PET bottle.

#2 Looooeeee! Nov 05, 2002 09:47 AM

Welcome to the E-Zone Takao,


I'm very impressed, I know nothing about programming PIC's or even Amtels for that matter, and here you are with code and a circuit schematic. Trizza, are you following this?

Do you have any rough sketches of how your motor is made?

Looee

#3 vintage1 Nov 05, 2002 10:00 AM

Re: My own design Sensor less/ Brush less/ Slot less Motor and driver
 
Quote:

Originally posted by Takao Shimizu
The schematic and software source code are in my Home Page http://plaza29.mbn.or.jp/~sgl02777/
The motor case is plywood and fuselage is by PET bottle.

Seems a very complex way you have to drive the positive going outputs - all bootstraps and opto-couplers...Wouldn't a P channel FET have been easier? :)

And I don't see what UL1 - 78LD12 - is for either...

However it looks pretty interesting.

Does it work?

#4 stumax Nov 05, 2002 04:52 PM

Vintage1, don't P channel fets have higher on resistance usually? I've designed and built high power h-bridge power amps for prototype automotive electric power steering systems and used International Rectifier fet drivers - able to deliver over one amp to the gate at 15kHz, boosts itself up, quite a neat method. If the voltage is low enough (ie below 20 cells) you could use a boost circuit to get 30V and drive the fets via hcpl3120 optos from Agilent. Mind you, I'm only a mech eng, you'd probably know better than me (I gathered from your previous posts you're an elec eng or similar, correct?), I'm always keen to learn :D

#5 Takao Shimizu Nov 05, 2002 08:19 PM

Just wind the coil 5 turns, 3 phase. That's all.
 
1 Attachment(s)
I was tring to draw the basic figure of windings, very simple.
Wrap the rotor magnet with kitchen food wrapping before winding and CA the winding wires.
After CAed, remove the wrap.

#6 Takao Shimizu Nov 05, 2002 08:34 PM

Re: Re: My own design Sensor less/ Brush less/ Slot less Motor and driver
 
Hi, Vintage1

>Wouldn't a P channel FET have been easier? :)
N-ch is lower Rds-on resistance.
I am testing IRF7324 P-MOSFET, now.

>And I don't see what UL1 - 78LD12 - is for either...

Small Package of 7812 low drop type voltage regulator.

>Does it work?
See the date of design. Still flying without maintenance.

#7 radix2 Nov 05, 2002 10:09 PM

Takao, very nice work !

do you have any pictures of your stator construction?

did you use the rotor as a form for winding the stator ?

a drawing of a cross section of your stator winding method (using an "o" for wires as they come up out of the page and an "x" where the go into the page may be a clearer way to show the pattern.

Thank you for sharing your work.

#8 S55 Nov 05, 2002 11:43 PM

There are good P- channels now. I am using some bellow 7 mohm at work.
The problem is not Rds-on. Stacking a P and an N together might get you in trouble if you do not pay attention to the drive circuitry. You may easily have them turned on simultaneously.

S55

#9 S55 Nov 05, 2002 11:53 PM

Almost twenty years ago or so I was looking for a quiet motor (not for a plane) and decided to use the one from an old cassette recorder. That motor had a mirror concept. Inside a cylindrical magnet there was a rotating plastic cylinder with the windings embedded in the wall of the cylinder. It had brushes of course, but no laminations on the rotor.
S55

#10 Looooeeee! Nov 06, 2002 12:03 AM

S55,


What you're referring to is called a "coreless motor". The Orion Elite is one kind. Also back in the 70's, Futaba used these motors in very high speed, (for the day, yes) servos. I have a couple of these old servos. Also the Fauhauber and Wes Tecnik motors for indoor and real light parkies are the same type..

#11 Trizza Nov 06, 2002 12:18 AM

Great!! I've bookmarked your stuff. I'll make one once exams are over.

AVRs are easy to deal with. Just build a cheapo programmer eg PonyProg and away you go. AT90S2313-10PC available off the shelf here. Looks like none of the extra features of the 2313 are being used, so a 1200 would probably work too (if the code is short enough). I have plenty of SMD 1200s. How many words is the software Takao?

Very impressive code, quite a lot of rcalls involved but otherwise looks reasonably efficient. This is one of those things I was going to try to make after exams, but I doubt I would have been able to do anywhere near as good a job on the schematic.

Takao, do you have the schematic in a file format (eg Protel)? If so it will save me drawing it out again to make a PCB layout.

I'm excited! CD-ROM brushless coming my way :D

Vintage1 - the optocouplers are to help remove inductive kick it says on the diagram.

*exactly* what I was looking for thanks Takao!!

#12 Takao Shimizu Nov 06, 2002 02:55 AM

radix2,

>did you use the rotor as a form for winding the stator ?

No, I did not.
My English was bad.
Just wrap the magnet rotor with food wrap before winding.

>do you have any pictures of your stator construction?
I can not find it.

#13 Takao Shimizu Nov 06, 2002 02:58 AM

S55,

> You may easily have them turned on simultaneously.

Yes, you are!
Read my program. I had turned the timing as you said, include photo cuppler delay.

#14 Takao Shimizu Nov 06, 2002 03:10 AM

1 Attachment(s)
Trizza,

>How many words is the software Takao?

Over 1k.
1200 can not handle it.

>do you have the schematic in a file format (eg Protel)?
I am using Bsh freeware.
http://www.suigyodo.com/online/e/index.htm

>PCB layout.
I am using PCBE freeware.
http://www.vector.co.jp/soft/dl/win9.../se056371.html
(Command is in English)

Follow is my smallest 2 pole brushless motor, 70krpm without load. The size is about Futaba connector.

#15 stumax Nov 06, 2002 03:56 AM

Mr Shimizu, congratulations on your innovations! Can you explain your method of rotor position sensing in the controller? Are you just sensing the zero crossing of the unused phase, or the second differential to get the inflection point? From what I have done so far, the second method is more accurate over all speeds, but I guess you could find the error and allow for it by mapping timing advance and stroing in a lookup table, similar to an engine ECU., although I have not seen this done. Also, do you pwm the fets on startup to reduce current draw? Finally, what starting algorithm have you employed? I hope you don't mind sharing so much with us, there are many who would not be so open, it's great to see the aeromodelling spirit is still alive!:D
cheers,
Stuart.


All times are GMT -5. The time now is 04:01 AM.