SMALL - espritmodel.com SMALL - Telemetry SMALL - Radio
Reply
Thread Tools
Old Feb 03, 2004, 08:16 AM
now that's a wattmeter...
simingx's Avatar
Singapore
Joined May 2002
1,136 Posts
Quote:
Originally posted by KreAture
Mr DIY
One of my biggest headaches is finding a good chip for the BEC.
Wich one are you using ?

Low dropout and high current usually implies large, bulky non SMD parts.
Try the Rohm BA05 series. Available in DPAK package, 0.3V dropout, 1.2W max dissipation. Use 3 in parallel... Reasonably priced from RS Components (not RadioShack!), their US rep is Allied Electronics.
simingx is offline Find More Posts by simingx
Last edited by simingx; Feb 03, 2004 at 08:19 AM.
Reply With Quote
Sign up now
to remove ads between posts
Old Feb 03, 2004, 08:37 AM
Impossible? Hah!
KreAture's Avatar
Oslo Fornebu, Norway
Joined Jun 2003
3,074 Posts
3 regulators would take up a lot of room.
KreAture is offline Find More Posts by KreAture
Reply With Quote
Old Feb 03, 2004, 12:47 PM
Dreamin' in 3D
Blackhawk3D's Avatar
Lodi/San Luis Obispo
Joined Oct 2003
1,584 Posts
Is there a smaller project, even a brushed one, that I could do to get my feet wet before I start with a complicated project such as this. I probably wouldn't program anything (I would use a preprogrammed or hardware solution) but some of this is still complicated. Are there any brushed controllers I could build to get an idea on how all this works and everything? Thanks.
Blackhawk3D is offline Find More Posts by Blackhawk3D
Reply With Quote
Old Feb 03, 2004, 01:37 PM
homo ludens modellisticus
Ron van Sommeren's Avatar
The Netherlands, GE, Nijmegen
Joined Feb 2001
10,791 Posts
DIY brushed controller designs:
http://www.rcgroups.com/forums/showt...hreadid=157193
Ron van Sommeren is offline Find More Posts by Ron van Sommeren
Reply With Quote
Old Feb 03, 2004, 02:09 PM
Registered User
Baltimore Bwi Rr Stn, Maryland, United States
Joined Oct 2003
520 Posts
Quote:
Originally posted by Ron van Sommeren
I heard from Jo Aichinger, of Speedy-BL ESC fame, that Microchip will introduce a new PIC for motorcontrollers, 30MHz (MIPS?) He did not have a part number.

Met vriendelijke groet Ron van Sommeren
diy brushless e-motor homebuilders discussion.
Electric fly-in & lrk meet, June 27th, 2004, Nijmegen, the Netherlands.
The parts are 40mhz clock and 30MIPS. They are the dsPIC30 series and are now in production.
whanderson is offline Find More Posts by whanderson
Reply With Quote
Old Feb 04, 2004, 01:19 AM
Registered User
Cape Town, South Africa
Joined Feb 2004
6 Posts
PWM question

I have been following this thread for a while now, and am considering doing a controller design for myself.
I have one question though, does the PWM just get used to get the motor moving, or is it used throughout the speed range while the motor is running?
CobusM is offline Find More Posts by CobusM
Reply With Quote
Old Feb 04, 2004, 01:54 AM
Registered User
Mr DIY's Avatar
South Africa
Joined Aug 2003
587 Posts
Quote:
I have been following this thread for a while now, and am considering doing a controller design for myself.
I have one question though, does the PWM just get used to get the motor moving, or is it used throughout the speed range while the motor is running?
I got to smile at this question ...as it was my very first question and headache as well. Unlike others, who have had access to a commercial controller to analyze, I had nothing. I had to figure all this out by experimentation.

Answer: PWM is used throughout the whole drive sequence and not just under starting.
Mr DIY is offline Find More Posts by Mr DIY
Reply With Quote
Old Feb 04, 2004, 09:37 AM
Registered User
Baltimore Bwi Rr Stn, Maryland, United States
Joined Oct 2003
520 Posts
Got another app note for you folks. This one is not for the timid, worth reading though.
http://www.microchip.com/download/ap...ic18/an901.pdf
whanderson is offline Find More Posts by whanderson
Reply With Quote
Old Feb 04, 2004, 10:35 AM
Impossible? Hah!
KreAture's Avatar
Oslo Fornebu, Norway
Joined Jun 2003
3,074 Posts
Nice one... And notice how limited it will be in rpm ?
A 4 pole motor can be ran at 4500 rpm...

Our cdrom motors are generally 12 pole. Instead of 2:1 gearing we have 6:1 and max rpm on a cdrom motor would then be 1500 rpm. I wouldn't be happy.

This is why I am trying to simpify the commutation-sensing so that faster rates are possible.

They use a 16 kHz PWM in the appnote ADC synched to it. At 4500 rpm (4 pole motor) the field is rotating at 9000 rpm wich gives us a timeframe for the rising or falling EMF of approx 18.52 us and this encompasses 60 degrees. The distance between the emf intervals is 180 degrees. To achieve an accuracy of 1 degree timing-wise, the dsPIC would have to sample once every 308.6 ns or at a speed of 3.24 MHz. In addition, theese data need to be handled and used to calculate the required changes in commutation rate to keep motor in synch.

I wonder how fast the dsPIC30 can sample ?

I did a grave miscalculation here... Instead of hounding me for my view, how about looking at the numbers? It turns out I forgot to divide rpm by 60 to get rps before I continued my rant.
Thus 9000 rpm = 150 rps and thus a timeframe for rising or falling edge of 1.11 ms. This results in a need for a sample every 18.5 us not every 308.6 ns or at a rate of 54 kHz. (1/60 as fast)
This is well within the scope of the chip. You say it can do 500 ksps and thus we could teoretically do almost 10 times as high rpm. This translates to 83333 rpm for a 2 pole motor or 13888 rpm for a 12-pole cdrom motor. If this can actually be achieved, then this would be great. (Enough for a cdrom motor on a slowflyer.)

I wonder about the price of theese chips ?
KreAture is offline Find More Posts by KreAture
Last edited by KreAture; Feb 04, 2004 at 03:41 PM.
Reply With Quote
Old Feb 04, 2004, 01:41 PM
Registered User
Baltimore Bwi Rr Stn, Maryland, United States
Joined Oct 2003
520 Posts
Quote:
Originally posted by KreAture
Nice one... And notice how limited it will be in rpm ?
A 4 pole motor can be ran at 4500 rpm...

Our cdrom motors are generally 12 pole. Instead of 2:1 gearing we have 6:1 and max rpm on a cdrom motor would then be 1500 rpm. I wouldn't be happy.

This is why I am trying to simpify the commutation-sensing so that faster rates are possible.

They use a 16 kHz PWM in the appnote ADC synched to it. At 4500 rpm (4 pole motor) the field is rotating at 9000 rpm wich gives us a timeframe for the rising or falling EMF of approx 18.52 us and this encompasses 60 degrees. The distance between the emf intervals is 180 degrees. To achieve an accuracy of 1 degree timing-wise, the dsPIC would have to sample once every 308.6 ns or at a speed of 3.24 MHz. In addition, theese data need to be handled and used to calculate the required changes in commutation rate to keep motor in synch.

I wonder how fast the dsPIC30 can sample ?
Come on and get real. It is an app note on a specific implementation. It is intended to provide insights on how to use the chip. Are you so closed minded that you can't see beyond the app note's specific implementation? Have you no imagination? The PWM module clocks to 10Mhz with up to 16 bit resolution. There are 8 possible PWM channels with 4 possible timebases. Sounds like this guy will do a little better than 4500 rpm to me. The A/D is 10 bits at 500ksps, is that fast enough for you? The part clocks up to 40Mhz and does 30mips. It has internal configurations specifically intended for BL motors. Your failure to see farther than your nose expains why you are having trouble reaching your controller design goals.
whanderson is offline Find More Posts by whanderson
Reply With Quote
Old Feb 04, 2004, 02:13 PM
Registered User
anfarol's Avatar
Forli, Italy
Joined Dec 2001
242 Posts
yes whanderson the DsPic has very powerful feature, but the bet is doing the job with poor microcontroller.
Hacker controller have an ATMega8... and no other intelligence...
Speedy BL also...

I work a lot with ATMega8, but i would like do the job with a common PIC, at the moment, I'm just downloading datasheet of 16F767. 3 separated PWM, 8Mhz internal (20 ext) and other good features...
I'll se...

Bye,
anfarol
anfarol is offline Find More Posts by anfarol
Reply With Quote
Old Feb 04, 2004, 02:15 PM
Impossible? Hah!
KreAture's Avatar
Oslo Fornebu, Norway
Joined Jun 2003
3,074 Posts
whanderson
Please behave. I was merely stating that if one used such a implementation as in the appnote one would need a very high sample-frequency to achieve good accuracy. When it comes to processing power, I agree that it is great. I bet there are a huge number of ways theese chips can be used for motor control and maby we can find one for RC too. It's just the ADC approach that has limits in rpm... Oh, btw: 500 ksps is a bit far from 3.24 msps but I must say it's really fast compared to other PIC chips.
KreAture is offline Find More Posts by KreAture
Reply With Quote
Old Feb 04, 2004, 05:23 PM
Impossible? Hah!
KreAture's Avatar
Oslo Fornebu, Norway
Joined Jun 2003
3,074 Posts
whanderson
Also, please excuse me if I could be misinterpreted earlier.
After looking over what filter order I'd need and what board space it would use to use my method, I think that using a simple ADC system will be better and much lighter for small stuff like I am trying to make.

Sacrifising some max rpm's to get half the board size sounds more and more appealing.

Problem is my chips ADC has a minimum conversion time resulting in about 50 kHz sample time if I run it back2back. If I stay by my hope of 1deg resolution on the sampeling then I'd get a max rpm of about 50000/360*60 = 8333 and a 12 pole cdrom would then go at 1388 rpm.

Ofcource, just to make something that can fly a Tiger Moth or something, I could probably make due with a lot less accuracy.

Mr DYI:
What does your system have for accuracy now?
That is, pr 360 deg, how often do you sample? Or, put another way: In the 60 deg rise-time of the EMF, how
many times do you sample it to find the crossing?

I really want my little $2 chip to do 500 ksps now whanderson. It is sounding a lot better than it did some hours ago.
KreAture is offline Find More Posts by KreAture
Reply With Quote
Old Feb 04, 2004, 07:33 PM
Registered User
Baltimore Bwi Rr Stn, Maryland, United States
Joined Oct 2003
520 Posts
Quote:
Originally posted by anfarol
yes whanderson the DsPic has very powerful feature, but the bet is doing the job with poor microcontroller.
Hacker controller have an ATMega8... and no other intelligence...
Speedy BL also...

I work a lot with ATMega8, but i would like do the job with a common PIC, at the moment, I'm just downloading datasheet of 16F767. 3 separated PWM, 8Mhz internal (20 ext) and other good features...
I'll se...

Bye,
anfarol
The 16F767 has 3 timers and 1 PWM with 2 selectable outputs. A more common PIC would be the 16F87x. This was among the first Flash parts Microchip released. Micrchip's AN857 implements a BL controller with a 16f877.
whanderson is offline Find More Posts by whanderson
Last edited by whanderson; Feb 04, 2004 at 07:36 PM.
Reply With Quote
Old Feb 04, 2004, 08:00 PM
Impossible? Hah!
KreAture's Avatar
Oslo Fornebu, Norway
Joined Jun 2003
3,074 Posts
I think the trick is finding the PIC with th efastest "acquire" delay. This is important for reliable sampling. Then one needs to find a PIC that will do enough samples pr second to keep the angular resolution to acceptable levels.
KreAture is offline Find More Posts by KreAture
Reply With Quote
Reply


Thread Tools

Similar Threads
Category Thread Thread Starter Forum Replies Last Post
Brushless speed control problem TomL Sport Planes 8 Jun 08, 2003 01:23 AM
TMM-18-3ph Brushless Speed Control nelsonjo1 Power Systems 5 Apr 04, 2003 12:33 AM
FS: Model Motors 40 amp Brushless Speed Control. Jim McPherson Aircraft - Electric - Airplanes (FS/W) 3 Dec 10, 2002 05:02 PM
Dragon 35 brushless speed control Steve H. Aircraft - Electric - Airplanes (FS/W) 0 Aug 17, 2002 12:29 AM
Hooking up a Jeti Brushless Speed Control to a Jeti Pashor 15/4 motor. HELP!!! Rob R Power Systems 3 Jan 10, 2002 06:37 PM