You can see how the lads from USQ got on when they flew their robot to the
UK - unless they've been edited out!

-----Original Message-----
From: Tracy Lightfoot [mailto:T.Lightfoot@...]
Sent: Friday, 19 November 1999 2:03
To: robot_sig@egroups.com
Subject: [robot_sig] Robot Wars on TV


Hi there,

Apparently "Robot Wars" - the big beat-em-up overseas robot
competition - is on TV on Channel 9, Sunday 6.30pm.  These are not
true 'robots', as they are remote controlled, but lots of fun to
watch!

Tracy

-----------------------------------------------------------------
Tracy Lightfoot                                   Multimedia Unit
Multimedia Developer          Griffith Flexible Learning Services
Ph +61 7 3382 1310               Division of Information Services
Fx +61 7 3382 1333                            Griffith University
-----------------------------------------------------------------

------------------------------------------------------------------------
Looking for the latest consumer electronic gadgets or computer
equipment? eBay has thousands of audio equipment, computer
games & accessories.  You never know what you might find at eBay!
http://clickhere.egroups.com/click/1142

--
If you wish to be removed from this mailing list
Go to http://www.eGroups.com and remove yourself
or email aarond@...

-- Create a poll/survey for your group!
-- http://www.egroups.com/vote?listname=robot_sig&m=1

Hi Tracy,

I guess you mean 27 of November not march
I should be there as a spectator

John Carter
Climate Impacts and Natural Resource Systems
Resource Sciences Centre
Department of Natural Resources
Floor 4 Forestry Building
80 Meiers Road Indooroopilly

Phone (07) 38969588
Fax     (07) 38969606

> ----------
> From:  Tracy Lightfoot[SMTP:T.Lightfoot@...]
> Sent:  Friday, 19 November 1999 14:01
> To:  robot_sig@egroups.com
> Subject:  [robot_sig] Robot Challenge 1999
>
>    RRRR    CCCC   Robot
>    R   R  C
>    RRRR   C       Challege
>    R   R  C
>    R   R   CCCC   1999
>
>
> When:  Saturday 27th March
>
> Where: Room 0.06 of Humanities building, Nathan Campus, Griffith
> University
>
> Events:
>
>    Stockcar Derby Qualifying  starts 1.30pm
>
>    This is a unique competition. Autonomous robotic cars
>    race against each other for speed on a Nascar shaped track.
>
>    Sumo Wrestling Match starts at 3:00pm.
>
>    Sumo Wrestling Based on real Sumo Wrestling rules, each
>    autonomous robot is required to locate and push, flip,
>    roll, shove, throw, hurl, or otherwise remove the opposing
>    robot from a 1.5m ring in the shortest period of time.
>
> For More Info:
>
> Phone Tracy Lightfoot on 3382 1310
> WWW http://members.xoom.com/robot_sig/
>
> Everyone is welcome!
>
> -----------------------------------------------------------------
> Tracy Lightfoot                                   Multimedia Unit
> Multimedia Developer          Griffith Flexible Learning Services
> Ph +61 7 3382 1310               Division of Information Services
> Fx +61 7 3382 1333                            Griffith University
> -----------------------------------------------------------------
>
> ------------------------------------------------------------------------
> Looking for the latest consumer electronic gadgets or computer
> equipment? eBay has thousands of audio equipment, computer
> games & accessories.  You never know what you might find at eBay!
> http://clickhere.egroups.com/click/1142
>
> --
> If you wish to be removed from this mailing list
> Go to http://www.eGroups.com and remove yourself
> or email aarond@...
>
> -- Talk to your group with your own voice!
> -- http://www.egroups.com/VoiceChatPage?listName=robot_sig&m=1
>
>

Hi there,

Apparently "Robot Wars" - the big beat-em-up overseas robot
competition - is on TV on Channel 9, Sunday 6.30pm.  These are not
true 'robots', as they are remote controlled, but lots of fun to
watch!

Tracy

-----------------------------------------------------------------
Tracy Lightfoot                                   Multimedia Unit
Multimedia Developer          Griffith Flexible Learning Services
Ph +61 7 3382 1310               Division of Information Services
Fx +61 7 3382 1333                            Griffith University
-----------------------------------------------------------------

RRRR    CCCC   Robot
    R   R  C
    RRRR   C       Challege
    R   R  C
    R   R   CCCC   1999


When:  Saturday 27th March

Where: Room 0.06 of Humanities building, Nathan Campus, Griffith University

Events:

    Stockcar Derby Qualifying  starts 1.30pm

    This is a unique competition. Autonomous robotic cars
    race against each other for speed on a Nascar shaped track.

    Sumo Wrestling Match starts at 3:00pm.

    Sumo Wrestling Based on real Sumo Wrestling rules, each
    autonomous robot is required to locate and push, flip,
    roll, shove, throw, hurl, or otherwise remove the opposing
    robot from a 1.5m ring in the shortest period of time.

For More Info:

Phone Tracy Lightfoot on 3382 1310
WWW http://members.xoom.com/robot_sig/

Everyone is welcome!

-----------------------------------------------------------------
Tracy Lightfoot                                   Multimedia Unit
Multimedia Developer          Griffith Flexible Learning Services
Ph +61 7 3382 1310               Division of Information Services
Fx +61 7 3382 1333                            Griffith University
-----------------------------------------------------------------

Hiya,

Lots of people have said "tell me what you got" so I will just post if to

the robot list.

Please express interest before sunday or I wont bother.

mail me back at

andrewm@...

with a PH# and I will give you details then (don't really want to give

my address and all on the mailing list)



There is about two boxes full of "broken" stuff that is free to

good homes.  The have many components on them to pull off.

There is also stuff that I got at the "computer graveyard" that I

have never go around to using as well as stuff I got at auctions

that I will never do anything with.

All that kind of stuff I will get rid of for what I got it for.

Here is a bit of a list.

There is also various electonic things I have collected over the years.

1 wishmaker digital prototype station (you all would love this) $200
1 Modem cabinet (needs a trailer ot truck to move)  $200
1 motorised credit-card reader/writer with software I wrote   $70
1 DSO adaptor MK2 (EA june 94 Dick smith price $199)  $80
1 DSO adaptor MK1 (not completed) 		 $20
1 VHF/UHF spectrum analyser (EA sept 92 dicky price ~200) $110
1 Tektronics 4663 interactive digital plotter (two people to lift) $60
1 Jaycar UV EPROM Eraser (very little use and 170new)  $60
2 heavy duty beat-baby-harp-seals type UV eraser (no bulbs) $40ea
2 large per.mag. servo motors (~2Kg) with sloted sensors  $5ea
2 large per.mag. servo motors (~2Kg) with encoder wheels $5ea
1 Big-arse heavy-duty monitor stand 		 $15
2 IBM PC XT computers (4Mhz 8086) with KBD and monitors $10ea
1 486 - 8Mb RAM No HDD no Floppy 		 $5
1 Broken pioneer CD player (not yet canabalised) 	 $5
1 Strobe light (faulty trigger transformer) 		 $30
? little IDE and SCSI hard drives 			 $1-5
? sundry computer cards 			 $0-5
1 adaptec 1542A SCSI card 			 $10
2 386MB with ram 				 $2ea
2 Aura interactors (Bought these at jaycar for $40)  $25ea
1 Epson LX400 					 $5
1 comodore MPS1250 printer 			 $5
1 14400 modem with power supply ($5 for the PSU really)  $5
1 FAX converter/modem (turn laser or dotty into plain.pap.fax) $100
1 broken atari lynx (LCD OK) 			 $5
1 microfiche printer 				 $30
2 RS232 terminals 				 $5 for both

lots of ADPCM digital voice recorder chips 	 $5ea
lots of little steppers 				 $?
several broken IBM KBDs 			 $0
comodore 64 and 128 carcasses 			 $0

Lots of other broken shit just to be carried away so I don't have to

look at it any more.

Later,
andrewm

S#!T Damn Mailing List,

Sorry that was just for andrewm

Josh

-----Original Message-----
From: Wherrett, Joshua [mailto:Joshua.Wherrett@...]
Sent: Tuesday, 9 November 1999 15:19
To: robot_sig@egroups.com
Subject: [robot_sig] Re: Message from andrewm


Ok andrewm,
hit me with the treasure list

Aarrrrgg

Josh


-----Original Message-----
From: andrewm [mailto:andrewm@...]
Sent: Tuesday, 9 November 1999 1:28
To: robot_sig@egroups.com
Cc: andrewm@...
Subject: [robot_sig] Message from andrewm


Hiya all,

This is a message to any of the people at the last

monthly meeting who expressed interest in a

"garage sale" I was planning on having.

I wil prob. hold it this sunday.

Instead of wasting everybodies time on the mailing

list if anyone is interested please mail me back at

andrewm@...

and I will tell you more about what kinds of motors/

computers/electronics I will have there.

Later,
andrewm

------------------------------------------------------------------------
--
If you wish to be removed from this mailing list
Go to http://www.eGroups.com and remove yourself
or email aarond@...

-- Create a poll/survey for your eGroup!
-- http://www.egroups.com/vote?listname=robot_sig&m=1

Ok andrewm,
hit me with the treasure list

Aarrrrgg

Josh


-----Original Message-----
From: andrewm [mailto:andrewm@...]
Sent: Tuesday, 9 November 1999 1:28
To: robot_sig@egroups.com
Cc: andrewm@...
Subject: [robot_sig] Message from andrewm


Hiya all,

This is a message to any of the people at the last

monthly meeting who expressed interest in a

"garage sale" I was planning on having.

I wil prob. hold it this sunday.

Instead of wasting everybodies time on the mailing

list if anyone is interested please mail me back at

andrewm@...

and I will tell you more about what kinds of motors/

computers/electronics I will have there.

Later,
andrewm

Hiya all,

This is a message to any of the people at the last

monthly meeting who expressed interest in a

"garage sale" I was planning on having.

I wil prob. hold it this sunday.

Instead of wasting everybodies time on the mailing

list if anyone is interested please mail me back at

andrewm@...

and I will tell you more about what kinds of motors/

computers/electronics I will have there.

Later,
andrewm


----------
> From: Brett Lymn <blymn@...>
> To: robot_sig@egroups.com
> Subject: [robot_sig] Re: h-bridge
> Date: Thursday, September 30, 1999 10:27 AM
>
> According to Chris Parker:
> >
> >Can anyone recommend some mosfets to use in an H-bridge?
>
> I used to use some excellent mosfets in a switch mode power supply I
> was making - they were very robust.  I cannot for the life of me
> remember the manufacturer at the moment.  I shall have a look at home.
>
> > The motors I am
> >using draw around 50Amps.
> >
>
> What is the failure mode of the driver devices?  Are they going open
> circuit or short circuit (and then blowing fuses)?  This gives a clue
> as to what is happening.  If the drivers are going OC then you are
> overloading them and burning out the connection wire between the die
> and the external connection.  If they are going short circuit then you
> are getting a spike that is punching through the layers on the die and
> creating a short circuit - this indicates that your snubbing circuitry
> is not doing is job well/fast enough.
>
> --
>
============================================================================
===
> Brett Lymn, Computer Systems Administrator, British Aerospace Australia
>
============================================================================
===
>
>
>
> ------------------------------------------------------------------------
> Join theglobe.com the premiere online community! Take advantage of
> FREE homepages, Stock quotes, News, Chat, Games and more! Interact
> with people just like you! http://clickhere.egroups.com/click/1042
>
> --
> If you wish to be removed from this mailing list
> Go to http://www.eGroups.com and remove yourself
> or email aarond@...
>

Hi there,

MEMBER MEETINGS

The next Member Meeting is on Tuesday 2nd November. We will be
holding pre-competition tests and discussions, with the Sumo ring and
a segment of the Stock Car track being available.  An update will
also be given about the SIG's own Sumo-bot development.

The meeting will start as usual at 7.30pm, and will be held at
Griffith University's Nathan Campus in the Technology building, room
0.15.  A map with directions can be found on the web at

      http://members.xoom.com/robot_sig/map.gif

Parking is free.

For those interested, the committee will be having a meeting just
prior to this at 6:30pm (same venue) so please come along if you want
to help out in this area.

COMPETITION

Our annual competition "Robot Challenge" will be held at Griffith
University on Saturday 27th November.  The competition web site is at:

    http://members.xoom.com/robot_sig/comp99/index.html

If you have a robot for the competition, you must register it in
advance - the form is available from the competition page.

More details will be posted to the competition web site as they are
confirmed.  A reminder email with full details will be sent around 1
week before the event also.

Don't forget the SIG has a web site at

   http://members.xoom.com/robot_sig/

This contains an events schedule to find out when things are on,
information about the SIG, a page of resources links and pointers to
the competition pages.   I am always on the lookout for links so if
you see something others would be interested in, let me know.

Looking forward to seeing you there!

Have fun,

Tracy

-----------------------------------------------------------------
Tracy Lightfoot                                   Multimedia Unit
Multimedia Developer          Griffith Flexible Learning Services
Ph +61 7 3382 1310               Division of Information Services
Fx +61 7 3382 1333                            Griffith University
-----------------------------------------------------------------

The Department of Computer Science and Electrical Engineering
at The University of Queensland is having its annual Open Day on
Thursday, 28th October.

Open Day is used to demonstrate the products of undergraduate
research in the department. The following projects will be of
particular interest to SIG members:

  - RoboRoos robot soccer team

  - Thecodont - a biped walking robot

  - An Autonomous Vacuum Cleaner

  - A Reborn PUMA Robot Arm - now equipped with the latest digital
controllers

  - An Amazing Floating Ping-Pong Ball

These projects will be on display in Room 311 of the Axon building.
There will, of course, be a wide variety of other student projects
throughout
the Axon and GP South Buildings. The displays run from 1pm to 6pm.

Hope to see you here.

Gordon Wyeth
Department of Computer Science and Electrical Engineering
University of Queensland     4072 AUSTRALIA
Phone: +61 7 3365 3770         Fax: +61 7 3365 4999
wyeth@...         http://www.elec.uq.edu.au/~wyeth

Hi

Could you please remove me from this listserver.  Thanks.

At 10:16 4/10/99 +1000, you wrote:
>Hi there,
>
>Ignore my last email.  Doh!
>
>Tracy
>
>-----------------------------------------------------------------
>Tracy Lightfoot                                   Multimedia Unit
>Multimedia Developer          Griffith Flexible Learning Services
>Ph +61 7 3382 1310               Division of Information Services
>Fx +61 7 3382 1333                            Griffith University
>-----------------------------------------------------------------
>
>------------------------------------------------------------------------
>--
>If you wish to be removed from this mailing list
>Go to http://www.eGroups.com and remove yourself
>or email aarond@...
>
>
>
>


Regards
Lana


Lana Conic
Project Officer (Health and Safety)
Office of Human Resource Management
Griffith University

Telephone: (07) 3875 7387
Facsimale: (07) 3875 5371
International Prefix: ISD 61 7
Email: L.Conic@...

Could you please put a link under the heading
Robotics conferences - calls for papers
to http://www.m2vip.com
(The hosting will hae been set up by Monday - in the meantime, you can see
the page at
http://www.usq.edu.au/m2vip

Cheers

John

PS We have a number of mechatronics/mobile robots projects at
http://www.ncea.org.au - but the 'best' ones such as Robocow have not got
much data.


-----Original Message-----
From: Aaron Dwyer [mailto:AaronD@...]
Sent: Friday, 1 October 1999 2:21
To: 'ACS Robotics SIG'
Subject: [robot_sig] october 5th monthly meeting


hi all

the next robotics meeting will be on next tue the 5th of october.

we will be watching a video of this years robocon and robosoccer that
was on early this year.

this should make for some very interesting viewing.

this meeting will start as usual at 7.30pm, and will be held at Griffith
University's Nathan Campus in the Technology building, room 0.15.  A map
with directions can be found on the web at

http://members.xoom.com/robot_sig/locations/gu_monthly.html

Parking is free.

for those interested, the committee will be having a meeting just prior
to this at 6:30pm (same venue) so please come along if you want to help
out in this area.

don't forget the SIG has a web site at

http://members.xoom.com/robot_sig/

this contains an events schedule to find out when things are on,
information about the SIG, a page of resources links and pointers to the
competition pages.  I am always on the lookout for links so if you see
something others would be interested in, let me know.

looking forward to seeing you there!

ciao
--
Aaron Dwyer                           Nulec Projects
Systems Engineer                      37 South St
Ph  : +61 7 3249 5635                 Lytton  Qld  4178
Fax : +61 7 3348 6260                 Australia

------------------------------------------------------------------------
--
If you wish to be removed from this mailing list
Go to http://www.eGroups.com and remove yourself
or email aarond@...

hi all

the next robotics meeting will be on next tue the 5th of october.

we will be watching a video of this years robocon and robosoccer that
was on early this year.

this should make for some very interesting viewing.

this meeting will start as usual at 7.30pm, and will be held at Griffith
University's Nathan Campus in the Technology building, room 0.15.  A map
with directions can be found on the web at

http://members.xoom.com/robot_sig/locations/gu_monthly.html

Parking is free.

for those interested, the committee will be having a meeting just prior
to this at 6:30pm (same venue) so please come along if you want to help
out in this area.

don't forget the SIG has a web site at

http://members.xoom.com/robot_sig/

this contains an events schedule to find out when things are on,
information about the SIG, a page of resources links and pointers to the
competition pages.  I am always on the lookout for links so if you see
something others would be interested in, let me know.

looking forward to seeing you there!

ciao
--
Aaron Dwyer                           Nulec Projects
Systems Engineer                      37 South St
Ph  : +61 7 3249 5635                 Lytton  Qld  4178
Fax : +61 7 3348 6260                 Australia

The seventh conference on Mechatronics and Machine Vision in Practice will
be held in Hervey Bay September 19-21 (during the first week of the
Olympics)

There will be a special session on Robot Games - and we are looking for some
'happenings' to combine with the conference.

Have a look at

http://www.usq.edu.au/m2vip

to see more about the conference.

Cheers

John

According to Chris Parker:
>
>Can anyone recommend some mosfets to use in an H-bridge?

I used to use some excellent mosfets in a switch mode power supply I
was making - they were very robust.  I cannot for the life of me
remember the manufacturer at the moment.  I shall have a look at home.

> The motors I am
>using draw around 50Amps.
>

What is the failure mode of the driver devices?  Are they going open
circuit or short circuit (and then blowing fuses)?  This gives a clue
as to what is happening.  If the drivers are going OC then you are
overloading them and burning out the connection wire between the die
and the external connection.  If they are going short circuit then you
are getting a spike that is punching through the layers on the die and
creating a short circuit - this indicates that your snubbing circuitry
is not doing is job well/fast enough.

--
===============================================================================
Brett Lymn, Computer Systems Administrator, British Aerospace Australia
===============================================================================

Hi Andrewm,

You have confused me with the original person who asked the question. I
believe Aaron Dwyer asked the first question. I haven't blown anything up as
yet, but found I had a similar problem so I just put my question in too.

I am building an underwater device and it will have 4 x 12 volt motors. So
far with them under load I have them pulling over 30A each. They are going
to be in pairs so my driver circuit would need to handle over 60A
continuous. I am changing the propeller design to reduce the load to more
like 25A each.

Coincidentally I bought some BUK456-60As yesterday, put some on a
breadboard last night and couldn't get them to work correctly. I will try
again tonight. Cost is not really a problem, within reason, and I have lots
of space for the circuit. Eventually I will need to build a pwm circuit to
go with it.

Thanks for your help and any further tips would be greatly appreciated.

Chris Parker

-----Original Message-----
From: andrewm <andrewm@...>
To: robot_sig@egroups.com <robot_sig@egroups.com>
Cc: andrewm@... <andrewm@...>
Date: Thursday, 30 September 1999 9:23
Subject: [robot_sig] Re: h-bridge


>Hiya,
>
>There are a few other things we need to know.
>
>What voltage are you using.
>
>How cheap do you want them
>
>How much space and support circuitry do you want.
>
>What kind of current feed back you wish to use (if any)
>
>
>But apart fromt hat
>
>DOES YOUR MOTOR REALLY DRAW 50Amps
>
>or is that just the peak stall current.
>
>Sounds like an impresive project.
>
>
>Later,
>andrewm
>
>----------
>> From: Chris Parker <cjparker@...>
>> To: robot_sig@egroups.com
>> Subject: [robot_sig] Re: h-bridge
>> Date: Thursday, September 30, 1999 9:01 AM
>>
>> Hi,
>>
>> Can anyone recommend some mosfets to use in an H-bridge? The motors I am
>> using draw around 50Amps.
>>
>> Regards
>> Chris Parker
>>
>> -----Original Message-----
>> From: Aaron Dwyer <AaronD@...>
>> To: 'robot_sig@egroups.com' <robot_sig@egroups.com>
>> Date: Wednesday, 29 September 1999 5:51
>> Subject: [robot_sig] Re: h-bridge
>>
>>
>> >> PS  Does anyone know of a good H-bridge chip that will do
>> >> largish motors
>> >> justice?  I've just blown up 4 in a row (well, I didn't do
>> >> it, my students
>> >> did!).  We're trying to drive a pair of motors that propel a
>> >> 10kg machine
>> >> at a fair speed.  We'd prefer to use a single chip to do both
>> >> motors if
>> >> possible, since we're runningout of real-estate quite quickly on this
>> >> machine!!!
>> >>
>> >
>> >you'll have to build your own.
>> >
>> >andrew might have published a webpage about it. maybe??
>> >
>> >but i suggest a mosfet h-bridge, that way you can make it as strong
>> >(current handling) as you need.
>> >
>> >real-estate wise. if it for one of your big toys then it's no problem.
>> >if it's for a small pcb then your screwed.
>> >
>> >ciao
>> >--
>> >Aaron Dwyer                           Nulec Projects
>> >Systems Engineer                      37 South St
>> >Ph  : +61 7 3249 5635                 Lytton  Qld  4178
>> >Fax : +61 7 3348 6260                 Australia
>> >
>> >------------------------------------------------------------------------
>> >
>> >eGroups.com home: http://www.egroups.com/group/robot_sig
>> >http://www.egroups.com - Simplifying group communications
>> >
>> >
>> >
>> >
>> >
>>
>>
>> ------------------------------------------------------------------------
>> Join theglobe.com the premiere online community! Take advantage of
>> FREE homepages, Stock quotes, News, Chat, Games and more! Interact
>> with people just like you! http://clickhere.egroups.com/click/1042
>>
>> --
>> If you wish to be removed from this mailing list
>> Go to http://www.eGroups.com and remove yourself
>> or email aarond@...
>>
>
>------------------------------------------------------------------------
>Join theglobe.com the premiere online community! Take advantage of
>FREE homepages, Stock quotes, News, Chat, Games and more! Interact
>with people just like you! http://clickhere.egroups.com/click/1042
>
>--
>If you wish to be removed from this mailing list
>Go to http://www.eGroups.com and remove yourself
>or email aarond@...
>
>
>

According to Gordon Wyeth:
>
>You will find that you will nearly always overshoot the required speed
>when you use proportional and integral together. If this is a problem
>introduce a derivative part. Again determing the derivative is easy:
>

Wheee lots of control theory I had happily forgotten :-)

One thing to note with adding the derivative is that it is very prone
to noise - you can get very large values out of the derivative so you
need to be careful to clamp this one, more so than the integral.


while (one_legged_ducks_swim_in_circles) {
     v_req = SetMySpeed ();
     v_act = ReadMySensor();
     v_err = v_req - v_act;
     v_int = v_int + v_err;
     if (v_int > BIG_NUMBER)
         v_int = BIG_NUMBER;
     v_der = v_err - v_olderr;
     if (v_der > BIG_NUMBER)
         v_def = BIG_NUMBER;
     v_olderr = v_err;
     PWM = Kp * v_err + Ki * v_int + Kd * v_der;
}

--
===============================================================================
Brett Lymn, Computer Systems Administrator, British Aerospace Australia
===============================================================================

Hiya,

Sorry, I did not read the original message about the 10Kg

machine and you allready blowing up a few and the

real estate premium.

50 amps is do-able in a single pakage h-bridge.

However its getting close to most big devices MAX

raitings and they also won't be cheap.

If you don't mind the real estate then as Arron D said,

and you have allready asked, use discreete mosfets.

for 50 amps (big ask anyway) and cheap devices I would

AGAIN recomend are 50N05 / BUK456-60As as long as your

Vds is not going to exceed 20V or so (the FETs are rated

to 50/60 volts but no point tempting the devil).

Each device can handle 50amps (but I woudl put a few in

parallel - 'casue satan is still there)


You are going to need support hardware and protection

for the fets.  This is going to take a few square inches of

board space.


Unfortunatly it is all a bit more complex than just throwing

a few fets on a board.

Later,
andrewm


----------
> From: Chris Parker <cjparker@...>
> To: robot_sig@egroups.com
> Subject: [robot_sig] Re: h-bridge
> Date: Thursday, September 30, 1999 9:01 AM
>
> Hi,
>
> Can anyone recommend some mosfets to use in an H-bridge? The motors I am
> using draw around 50Amps.
>
> Regards
> Chris Parker
>
> -----Original Message-----
> From: Aaron Dwyer <AaronD@...>
> To: 'robot_sig@egroups.com' <robot_sig@egroups.com>
> Date: Wednesday, 29 September 1999 5:51
> Subject: [robot_sig] Re: h-bridge
>
>
> >> PS  Does anyone know of a good H-bridge chip that will do
> >> largish motors
> >> justice?  I've just blown up 4 in a row (well, I didn't do
> >> it, my students
> >> did!).  We're trying to drive a pair of motors that propel a
> >> 10kg machine
> >> at a fair speed.  We'd prefer to use a single chip to do both
> >> motors if
> >> possible, since we're runningout of real-estate quite quickly on this
> >> machine!!!
> >>
> >
> >you'll have to build your own.
> >
> >andrew might have published a webpage about it. maybe??
> >
> >but i suggest a mosfet h-bridge, that way you can make it as strong
> >(current handling) as you need.
> >
> >real-estate wise. if it for one of your big toys then it's no problem.
> >if it's for a small pcb then your screwed.
> >
> >ciao
> >--
> >Aaron Dwyer                           Nulec Projects
> >Systems Engineer                      37 South St
> >Ph  : +61 7 3249 5635                 Lytton  Qld  4178
> >Fax : +61 7 3348 6260                 Australia
> >
> >------------------------------------------------------------------------
> >
> >eGroups.com home: http://www.egroups.com/group/robot_sig
> >http://www.egroups.com - Simplifying group communications
> >
> >
> >
> >
> >
>
>
> ------------------------------------------------------------------------
> Join theglobe.com the premiere online community! Take advantage of
> FREE homepages, Stock quotes, News, Chat, Games and more! Interact
> with people just like you! http://clickhere.egroups.com/click/1042
>
> --
> If you wish to be removed from this mailing list
> Go to http://www.eGroups.com and remove yourself
> or email aarond@...
>

Hiya,

There are a few other things we need to know.

What voltage are you using.

How cheap do you want them

How much space and support circuitry do you want.

What kind of current feed back you wish to use (if any)


But apart fromt hat

DOES YOUR MOTOR REALLY DRAW 50Amps

or is that just the peak stall current.

Sounds like an impresive project.


Later,
andrewm

----------
> From: Chris Parker <cjparker@...>
> To: robot_sig@egroups.com
> Subject: [robot_sig] Re: h-bridge
> Date: Thursday, September 30, 1999 9:01 AM
>
> Hi,
>
> Can anyone recommend some mosfets to use in an H-bridge? The motors I am
> using draw around 50Amps.
>
> Regards
> Chris Parker
>
> -----Original Message-----
> From: Aaron Dwyer <AaronD@...>
> To: 'robot_sig@egroups.com' <robot_sig@egroups.com>
> Date: Wednesday, 29 September 1999 5:51
> Subject: [robot_sig] Re: h-bridge
>
>
> >> PS  Does anyone know of a good H-bridge chip that will do
> >> largish motors
> >> justice?  I've just blown up 4 in a row (well, I didn't do
> >> it, my students
> >> did!).  We're trying to drive a pair of motors that propel a
> >> 10kg machine
> >> at a fair speed.  We'd prefer to use a single chip to do both
> >> motors if
> >> possible, since we're runningout of real-estate quite quickly on this
> >> machine!!!
> >>
> >
> >you'll have to build your own.
> >
> >andrew might have published a webpage about it. maybe??
> >
> >but i suggest a mosfet h-bridge, that way you can make it as strong
> >(current handling) as you need.
> >
> >real-estate wise. if it for one of your big toys then it's no problem.
> >if it's for a small pcb then your screwed.
> >
> >ciao
> >--
> >Aaron Dwyer                           Nulec Projects
> >Systems Engineer                      37 South St
> >Ph  : +61 7 3249 5635                 Lytton  Qld  4178
> >Fax : +61 7 3348 6260                 Australia
> >
> >------------------------------------------------------------------------
> >
> >eGroups.com home: http://www.egroups.com/group/robot_sig
> >http://www.egroups.com - Simplifying group communications
> >
> >
> >
> >
> >
>
>
> ------------------------------------------------------------------------
> Join theglobe.com the premiere online community! Take advantage of
> FREE homepages, Stock quotes, News, Chat, Games and more! Interact
> with people just like you! http://clickhere.egroups.com/click/1042
>
> --
> If you wish to be removed from this mailing list
> Go to http://www.eGroups.com and remove yourself
> or email aarond@...
>

Hi,

Can anyone recommend some mosfets to use in an H-bridge? The motors I am
using draw around 50Amps.

Regards
Chris Parker

-----Original Message-----
From: Aaron Dwyer <AaronD@...>
To: 'robot_sig@egroups.com' <robot_sig@egroups.com>
Date: Wednesday, 29 September 1999 5:51
Subject: [robot_sig] Re: h-bridge


>> PS  Does anyone know of a good H-bridge chip that will do
>> largish motors
>> justice?  I've just blown up 4 in a row (well, I didn't do
>> it, my students
>> did!).  We're trying to drive a pair of motors that propel a
>> 10kg machine
>> at a fair speed.  We'd prefer to use a single chip to do both
>> motors if
>> possible, since we're runningout of real-estate quite quickly on this
>> machine!!!
>>
>
>you'll have to build your own.
>
>andrew might have published a webpage about it. maybe??
>
>but i suggest a mosfet h-bridge, that way you can make it as strong
>(current handling) as you need.
>
>real-estate wise. if it for one of your big toys then it's no problem.
>if it's for a small pcb then your screwed.
>
>ciao
>--
>Aaron Dwyer                           Nulec Projects
>Systems Engineer                      37 South St
>Ph  : +61 7 3249 5635                 Lytton  Qld  4178
>Fax : +61 7 3348 6260                 Australia
>
>------------------------------------------------------------------------
>
>eGroups.com home: http://www.egroups.com/group/robot_sig
>http://www.egroups.com - Simplifying group communications
>
>
>
>
>

> Hiya,
>
> PWM is the method that you use for varying power to
>
> the motor - NOT THE MOTOR SPEED.
>

Um, actually, for a given load, the PWM value DOES set the
motor speed. It categorically does NOT set motor power.
Of course, for varying loads you need a control loop as per
your eminently sensible stuff that follows.


> One of the other guys (Brett) mentioned the magic word
>
> CONTROL LOOP.
>
> He also gave you the basic idea behind it.
>
> A DESIRED set speed.
>
> A FEED BACK of ACTUAL speed.
>
> Some CONTROL acting on the FEEDBACK to ALTER
>
>      the PWM signal that will drive the motor.
>
>
> BUT to answer your question about the thing Norbert spoke
>
> about and what it was called.
>
> NOT PIM but PID
>
> P = proportional
> I = integral
> D = differential.
>
> But there is prob. no need to go all the way to a PID system
>
> when a P (proportinal) one will do yo just fine.
>
> As Norbert said - there is plenty of info on control sys.
>
> on the net.
>
> If you go to infoseek (www.infoseek.com) and type
>
> +control +PID
>
> you will get thousands of hits.
>
> Best of luck.
>
> Keep the mailing list posted (and you can also ask us more Q.).
>
> Later,
> andrewm
>
>
>
> ----------
> > From: Paul L Nielsen <P.Nielsen@...>
> > To: 'robot_sig@egroups.com'
> > Subject: [robot_sig] Speed Control Question
> > Date: Wednesday, September 29, 1999 2:56 PM
> >
> > Hi Folks,
> >
> > I'm looking for a little advice on speed control for the stock cars. I
> > recall Norbet gave us a talk on how to effectively carry this out. The
> > system has some acronym like PIM and he said there's heaps of info on
the
> > web. Basically I'm trying to get my car to reach and maintain a specific
> > speed as efficiently as possible. I can get it to speed up and slow down
> > based using a series of if statements but I was hoping to figure out the
> > formular to get the beast to adjust the increment/decrements based on
the
> > gap between the existing speed and the required speed. It sounds like
> > integral calculus to me and I haven't touched that for 20 years.
> >
> > Regards,
> >
> > Paul Nielsen
> > Information Technology Services (ITS)
> > University of Queensland
> > Email: p.nielsen@...
> >
> > ------------------------------------------------------------------------
> >
> > eGroups.com home: http://www.egroups.com/group/robot_sig
> > http://www.egroups.com - Simplifying group communications
> >
> >
> >
>
> ------------------------------------------------------------------------
> --
> If you wish to be removed from this mailing list.
> Go to http://www.eGroups.com and remove yourself.
> or email aarond@....
>
>

The acronym that Norbet used I'm sure would have been PID -
meaning proportional-integral-derivative. The idea is like this:

You set your required speed => v_req
You measure your actual speed from your sensor => v_act
You find the difference between the two
which is the error in speed => v_err = v_req - v_act

v_err is your PROPORTIONAL error. You can use it on its own for
crude speed control, but you will find that your car never actually
reaches v_req. This is called "steady state error".

To use the proportional error simply multiply the error by a constant
determined by experimentation (lets call it Kp) and send that value
to your PWM register.

PWM = Kp * v_err

Low values of Kp will cause sluggish response,
a high value will cause oscillation about the required speed. Low
values will also have worse steady-state error.

If you add the integral part, you will get rid of the steady-state error.
Determing the integral error is very easy, simply set a variable to
zero at the start and then add on the proportional error as you go.
This integrates the error over time. If you call the variable v_int:

v_int = v_int + v_err

is all you need to calculate the integral error. If you use the integral
error
on its own, your response will likely be unstable to combine with the
proportional error thus

PWM = Kp * v_err + Ki * v_int

where Ki is the intgral error of your choice.

Words of warning:
You will need to choose a new value of Kp.
In some circumstance you will find that you are getting extremely large
values of v_int. If this is the case, limit the values to which it can
integrate
i.e. if (v_int > BIG_NUMBER) v_int = BIG_NUMBER

You will find that you will nearly always overshoot the required speed
when you use proportional and integral together. If this is a problem
introduce a derivative part. Again determing the derivative is easy:

v_der = v_err - v_olderr
v_olderr = v_err

You will now need a third constant Kd

PWM = Kp * v_err + Ki * v_int + Kd * v_der

Tuning these three variable together is bit like tuning a twin- dual barrel
carby. When you tweak one thing the rest all changes. Now that I've given
you the basics, you can probably find the tuning techniques on the web
(or enrol in my third year engineering control subject ;-) ).

In summary:

while (one_legged_ducks_swim_in_circles) {
     v_req = SetMySpeed ();
     v_act = ReadMySensor();
     v_err = v_req - v_act;
     v_int = v_int + v_err;
     if (v_int > BIG_NUMBER)
         v_int = BIG_NUMBER;
     v_der = v_err - v_olderr;
     v_olderr = v_err;
     PWM = Kp * v_err + Ki * v_int + Kd * v_der;
}

Cheers

Gordo

Gordon Wyeth
Department of Computer Science and Electrical Engineering
University of Queensland     4072 AUSTRALIA
Phone: +61 7 3365 3770         Fax: +61 7 3365 4999
wyeth@...         http://www.elec.uq.edu.au/~wyeth
----- Original Message -----
From: Paul L Nielsen <P.Nielsen@...>
To: <robot_sig@egroups.com>
Sent: Wednesday, 29 September 1999 14:56
Subject: [robot_sig] Speed Control Question


> Hi Folks,
>
> I'm looking for a little advice on speed control for the stock cars. I
> recall Norbet gave us a talk on how to effectively carry this out. The
> system has some acronym like PIM and he said there's heaps of info on the
> web. Basically I'm trying to get my car to reach and maintain a specific
> speed as efficiently as possible. I can get it to speed up and slow down
> based using a series of if statements but I was hoping to figure out the
> formular to get the beast to adjust the increment/decrements based on the
> gap between the existing speed and the required speed. It sounds like
> integral calculus to me and I haven't touched that for 20 years.
>
> Regards,
>
> Paul Nielsen
> Information Technology Services (ITS)
> University of Queensland
> Email: p.nielsen@...
>
> ------------------------------------------------------------------------
>
> eGroups.com home: http://www.egroups.com/group/robot_sig
> http://www.egroups.com - Simplifying group communications
>
>
>
>

Hiya,

PWM is the method that you use for varying power to

the motor - NOT THE MOTOR SPEED.

One of the other guys (Brett) mentioned the magic word

CONTROL LOOP.

He also gave you the basic idea behind it.

A DESIRED set speed.

A FEED BACK of ACTUAL speed.

Some CONTROL acting on the FEEDBACK to ALTER

      the PWM signal that will drive the motor.


BUT to answer your question about the thing Norbert spoke

about and what it was called.

NOT PIM but PID

P = proportional
I = integral
D = differential.

But there is prob. no need to go all the way to a PID system

when a P (proportinal) one will do yo just fine.

As Norbert said - there is plenty of info on control sys.

on the net.

If you go to infoseek (www.infoseek.com) and type

+control +PID

you will get thousands of hits.

Best of luck.

Keep the mailing list posted (and you can also ask us more Q.).

Later,
andrewm



----------
> From: Paul L Nielsen <P.Nielsen@...>
> To: 'robot_sig@egroups.com'
> Subject: [robot_sig] Speed Control Question
> Date: Wednesday, September 29, 1999 2:56 PM
>
> Hi Folks,
>
> I'm looking for a little advice on speed control for the stock cars. I
> recall Norbet gave us a talk on how to effectively carry this out. The
> system has some acronym like PIM and he said there's heaps of info on the
> web. Basically I'm trying to get my car to reach and maintain a specific
> speed as efficiently as possible. I can get it to speed up and slow down
> based using a series of if statements but I was hoping to figure out the
> formular to get the beast to adjust the increment/decrements based on the
> gap between the existing speed and the required speed. It sounds like
> integral calculus to me and I haven't touched that for 20 years.
>
> Regards,
>
> Paul Nielsen
> Information Technology Services (ITS)
> University of Queensland
> Email: p.nielsen@...
>
> ------------------------------------------------------------------------
>
> eGroups.com home: http://www.egroups.com/group/robot_sig
> http://www.egroups.com - Simplifying group communications
>
>
>

Hiya,

PWM is the method that you use for varying power to

the motor - NOT THE MOTOR SPEED.

One of the other guys (Brett) mentioned the magic word

CONTROL LOOP.

He also gave you the basic idea behind it.

A DESIRED set speed.

A FEED BACK of ACTUAL speed.

Some CONTROL acting on the FEEDBACK to ALTER

      the PWM signal that will drive the motor.


BUT to answer your question about the thing Norbert spoke

about and what it was called.

NOT PIM but PID

P = proportional
I = integral
D = differential.

But there is prob. no need to go all the way to a PID system

when a P (proportinal) one will do yo just fine.

As Norbert said - there is plenty of info on control sys.

on the net.

If you go to infoseek (www.infoseek.com) and type

+control +PID

you will get thousands of hits.

Best of luck.

Keep the mailing list posted (and you can also ask us more Q.).

Later,
andrewm


----------
> From: Paul L Nielsen <P.Nielsen@...>
> To: 'robot_sig@egroups.com'
> Subject: [robot_sig] Speed Control Question
> Date: Wednesday, September 29, 1999 2:56 PM
>
> Hi Folks,
>
> I'm looking for a little advice on speed control for the stock cars. I
> recall Norbet gave us a talk on how to effectively carry this out. The
> system has some acronym like PIM and he said there's heaps of info on the
> web. Basically I'm trying to get my car to reach and maintain a specific
> speed as efficiently as possible. I can get it to speed up and slow down
> based using a series of if statements but I was hoping to figure out the
> formular to get the beast to adjust the increment/decrements based on the
> gap between the existing speed and the required speed. It sounds like
> integral calculus to me and I haven't touched that for 20 years.
>
> Regards,
>
> Paul Nielsen
> Information Technology Services (ITS)
> University of Queensland
> Email: p.nielsen@...
>
> ------------------------------------------------------------------------
>
> eGroups.com home: http://www.egroups.com/group/robot_sig
> http://www.egroups.com - Simplifying group communications
>
>
>

> PS  Does anyone know of a good H-bridge chip that will do
> largish motors
> justice?  I've just blown up 4 in a row (well, I didn't do
> it, my students
> did!).  We're trying to drive a pair of motors that propel a
> 10kg machine
> at a fair speed.  We'd prefer to use a single chip to do both
> motors if
> possible, since we're runningout of real-estate quite quickly on this
> machine!!!
>

you'll have to build your own.

andrew might have published a webpage about it. maybe??

but i suggest a mosfet h-bridge, that way you can make it as strong
(current handling) as you need.

real-estate wise. if it for one of your big toys then it's no problem.
if it's for a small pcb then your screwed.

ciao
--
Aaron Dwyer                           Nulec Projects
Systems Engineer                      37 South St
Ph  : +61 7 3249 5635                 Lytton  Qld  4178
Fax : +61 7 3348 6260                 Australia

>Hi Folks,
>
>I'm looking for a little advice on speed control for the stock cars. I
>recall Norbet gave us a talk on how to effectively carry this out. The
>system has some acronym like PIM and he said there's heaps of info on the
>web. Basically I'm trying to get my car to reach and maintain a specific
>speed as efficiently as possible. I can get it to speed up and slow down
>based using a series of if statements but I was hoping to figure out the
>formular to get the beast to adjust the increment/decrements based on the
>gap between the existing speed and the required speed. It sounds like
>integral calculus to me and I haven't touched that for 20 years.
>
>Regards,
>
>Paul Nielsen
>Information Technology Services (ITS)
>University of Queensland
>Email: p.nielsen@...
>
>------------------------------------------------------------------------
>
>eGroups.com home: http://www.egroups.com/group/robot_sig
>http://www.egroups.com - Simplifying group communications

Paul,

What you're looking for is PWM - pulse width modulation.  Yes, you are
right, it does look scarily like integral calculus, and that is the best
way of accurately incrementing/decrementing speeds to a particular desired
speed.   You can however use PWM and skip the calculus.  What you want to
do is use a hunting method.  Basically the way you do it is send a bunch of
pulses out to your motor (should be able to do this in code from the
handyboard).  Now read in from your shaft encoders.  Do this for a
specified length of time, say 200 microseconds.  Compare the number of
shaft encoder leading edges to what you would expect if the wheels actually
were going the right speed.  If its less than what you expect, increase the
the pulse width (duty cycle) of your PWM generator, and if its more,
decrease your pulse width.

There is bucket loads of stuff on the web about it.   Try
www.dontronics.com or www.myke.com  for a start.

Michael.

PS  Does anyone know of a good H-bridge chip that will do largish motors
justice?  I've just blown up 4 in a row (well, I didn't do it, my students
did!).  We're trying to drive a pair of motors that propel a 10kg machine
at a fair speed.  We'd prefer to use a single chip to do both motors if
possible, since we're runningout of real-estate quite quickly on this
machine!!!

*****************************************************************************
Michael Lucas  BE(Mech) P.Eng 	 *email
	 mechanimation@...
Managing Director 			 *Ph  +61 7 3365 4091 (BH)
Mechanimation Technology Services Australia *Ph +61 7 3260 7584 (AH)
PO Box  6018 		 *Fax  +61 7 3266 5449
St Lucia    QLD   4067 		 *Mob  0419 737 869
AUSTRALIA 		 *Web T.B.A.
*****************************************************************************
Mechanimation Technology Services Australia   (BN 7094395)
IT, Mechatronics and Animatronics for Biomedical, Agriculture, Mining and
Cinema
"Let us turn your challenge into an opportunity!"
*****************************************************************************

Australian Young Professional Engineer of the Year 1999
Queensland Young Professional Engineer of the Year 1998
YOUNG ENGINEERS - MAKING IT HAPPEN IN QUEENSLAND
*****************************************************************************

According to Paul L Nielsen:
>
> I was hoping to figure out the
>formular to get the beast to adjust the increment/decrements based on the
>gap between the existing speed and the required speed. It sounds like
>integral calculus to me and I haven't touched that for 20 years.
>

Nope, what you are trying to do is control loop theory.  It sounds
like you know your current speed and your desired speed.  What you are
aiming for is to make the difference between them be zero.  If you
subtract your actual speed (plant reading in control system parlance)
from your desired speed (ka the set point) then you will have a
measure of how far you are away from the speed you want (called the
error in control systems).  This will give you the proportion of the
increment of power you need to apply to reach the desired speed.  You
probably do not want to apply the full increment all at once - I would
suggest you have a scale factor (called a damping factor in control
systems) and sneak up on the desired setting in small steps.  If you
do not do this you may end up with an unstable control loop which will
cause oscillations which you would see as undesired speed variations.
Make sure you cater for going too fast which will give you a negative
error meaning you need to slow down.  Good luck.

--
===============================================================================
Brett Lymn, Computer Systems Administrator, British Aerospace Australia
===============================================================================

Hi Paul,
What you are referring to is Pulse Width Modulation (PWM), this is performed
by adjusting the duty cycle (Time on / Time off) of your motors in order to
make them go faster (Time on >> Time off) or Slower (Time on << Time Off).
There is also another acronym, Proportional Intergral Derivative or PID
controller, which allows the motor to reach the desired speed as quiclky and
efficiently as possible (minimal oscillation).  Norbert Harle has
implemented these functions for his stock car using C.  His code is under
"fire machine", which can be accessed from his web page (along with a lot of
other handy links), his page is: http://www.eese.qut.edu.au/~harle/

and his source code is here:
http://www.eese.qut.edu.au/~harle/projects/68332%20Board%20QUT/firemach/

I hope this helps

Josh

-----Original Message-----
From: Paul L Nielsen [mailto:P.Nielsen@...]
Sent: Wednesday, 29 September 1999 14:56
To: 'robot_sig@egroups.com'
Subject: [robot_sig] Speed Control Question


Hi Folks,

I'm looking for a little advice on speed control for the stock cars. I
recall Norbet gave us a talk on how to effectively carry this out. The
system has some acronym like PIM and he said there's heaps of info on the
web. Basically I'm trying to get my car to reach and maintain a specific
speed as efficiently as possible. I can get it to speed up and slow down
based using a series of if statements but I was hoping to figure out the
formular to get the beast to adjust the increment/decrements based on the
gap between the existing speed and the required speed. It sounds like
integral calculus to me and I haven't touched that for 20 years.

Regards,

Paul Nielsen
Information Technology Services (ITS)
University of Queensland
Email: p.nielsen@...

------------------------------------------------------------------------

eGroups.com home: http://www.egroups.com/group/robot_sig
http://www.egroups.com - Simplifying group communications

Hi Folks,

I'm looking for a little advice on speed control for the stock cars. I
recall Norbet gave us a talk on how to effectively carry this out. The
system has some acronym like PIM and he said there's heaps of info on the
web. Basically I'm trying to get my car to reach and maintain a specific
speed as efficiently as possible. I can get it to speed up and slow down
based using a series of if statements but I was hoping to figure out the
formular to get the beast to adjust the increment/decrements based on the
gap between the existing speed and the required speed. It sounds like
integral calculus to me and I haven't touched that for 20 years.

Regards,

Paul Nielsen
Information Technology Services (ITS)
University of Queensland
Email: p.nielsen@...

The Australian Computer Society (ACS) Robotics Special Interest Group (SIG) aims to bring together people from various backgrounds who are interested in robotics - primarily autonomous mobile robots.  We are based in Brisbane, Queensland, Australia.  Our web site it at http://members.xoom.com/robot_sig/.

Group Manager: robot_sig-owner@egroups.com

To subscribe, send a message to robot_sig-subscribe@egroups.com or go to the e-groups's home page at http://www.egroups.com/group/robot_sig/