[chbot] HBRobotics/Micromouse

chchrobotics@lists.linuxnut.co.nz chchrobotics@lists.linuxnut.co.nz
Tue, 9 Jan 2007 08:37:59 +1300


This is a multipart message in MIME format.
--=_alternative 006BD95FCC25725D_=
Content-Type: text/plain; charset="US-ASCII"


Apologies if this appears twice.
My Email address has been changed & I received a message from the list 
saying I wasn't subscribed.




Hi All,
It's good to see interest in Micro Mouse starting up again.
I used to race micro mouse at the "Nelcon" competitions back in the 90s.

At the time I used steppers as they have inherent positioning which 
simplifies a lot of the control problems for a beginner.

The square shaped steppers from old 5.24' drives are ideal to get started 
with. 
I made wheels by turning up acetal plastic disks that pushed on to the 
band drum on the stepper shaft. A groove for an O ring makes a serviceable 
tyre.

When using steppers there are a number of things to watch for. Max speed 
is determined by both applied voltage & flyback voltage. A zenner diode to 
allow the flyback to reach 30 volts allows the 12V steppers to run much 
faster and have higher torque at the high speeds.

The other thing is that they do NOT start & stop instantly. When pushing 
performance they must be ramped up and down or you will loose steps.

Dead reckoning is a good starting point but auto re-zero on the fly is 
absolutely critical if you hope to survive more than one or 2 corners.

The advise I received when I first started proved very useful.

Complete each of  the following steps in turn  and don't proceed to the 
next step until the current step is fully and repeatedly completed.
1) make sure the mouse will fit in the maze. (my first mouse wouldn't fit 
round some corner types) 
2) make it go forward in a straight line and then stop.
3) make it go forward exactly one square and stop.
4) make it go forward 1 square, stop, and return to its original position.
5) make it turn 90 degree and stop.
6) make it turn 90 degree, stop, turn back 90 and stop at the same heading 
as original.
7)  make it go forward 1 square, stop, turn 90 degree, stop.

At this stage you will have good basic control and all the basic 
functionality. 
This is a good starting point on which to build in the navigation and wall 
sensing.

Only after you have basic navigation and wall sensing going do you need to 
worry about solving the maze.
If your mouse won't wander aimlessly around the maze for hours on end 
without hitting walls then the most advanced maze algorithm in the world 
won't help. 
(well at least 10 minutes.)

Once you get a basic mouse running the next challenge is a test maze.
I build my test maze out of 2 sheets of 12mm chip board, not MDF. 
(1200x2400)
By cutting 50mm strips off the side and ends of the sheets I ended up with 
a 10 X 10 maze. (2 sheets 1900 x 950)
I found out the hard way that you need a test maze that includes a centre.

10x10 seems to be about the best size for a home maze as it will stand up 
against the garage wall when not in use.
If you have a bigger garage you can just scale it up.
 My maze is a single piece but all the full size competition mazes I've 
seen are made in 4 quarters and then dowelled & bolted together.
Beware the joints if you take this approach. The APEC maze was more like a 
4W drive track than a micromouse maze.

I joined the 2 sheets by gluing one of the strips to the back of the 
joint.
The other strips were cut to length as wall segments.  Don't just cut them 
all to 180mm though.
A full outer wall was glued to the base and a few key wall sections were 
also glued in place.
I could then use bluetac to position lengths of walls as required to 
change the maze around.
I drew a grid in pencil to help layout. the lines were left/below the wall 
positions. 
Remember the maze is a grid on 180mm centres but the walls are 12mm so the 
passage ways are only 168mm wide. (See step 1 above)

Chip board is quite reflective to infra-red, which I was using for 
sensors. 
I didn't bother to paint it at all as I was using over wall down sensors 
and could easily distinguish the wall tops from the floor due to the much 
greater distance.

I had an infra-red reflectometer that I used to check competition maze 
reflectivity before a contest so I could re-calibrate my sensors if 
necessary. 
Never required in practice as reflectivity was usually within about 20% of 
my maze.


Happy mousing.

Regards,
Robin





Michael Pearce <mike@kiwacan.co.nz> 
Sent by: chchrobotics-admin@lists.linuxnut.co.nz
02/01/2007 06:55 p.m.
Please respond to
chchrobotics@lists.linuxnut.co.nz


To
chchrobotics@lists.linuxnut.co.nz
cc

Subject
Re: [chbot] HBRobotics/Micromouse






I forgot the following...

Since you are stepping, you know exactly how many 
revolutions the wheel has done so you can calculate 
position easily. Using a standard 1.8degree/step motor you 
can easily calculate 1/200 of a rev using single stepping.

Using micro-stepping 1/200,000 of a rev is possible to 
determine (but not at a sudden stop)

Up Sides:
 1. Step counting for accurate positioning
 2. Easy to build basic drivers (Full and half step)
 3. High output torque (Can drive wheel directly)
 4. Instant stop and hold at current position
 5. Very repeatable.
 6. Easily obtained in old dot-matrix printers, 
    5.25" floppy drives and old hard drives (<200MB)


Down sides:
 1. weight
 2. power draw
 3. losing steps (if hardware design bad or running to fast)
 4. slow top speed  (but dont need gearbox)
 5. Not cheap to buy new compared to standard DC motor.
 

I may get stuck into my Pick and place project later this 
month, and will hopefully be able to demo it at the next 
meeting to show standard and microstepping techniques.


Mike


On Tuesday 02 January 2007 18:24, Michael Pearce wrote:
> That design is nice and simple.
>
> The steppers do have the advantage of being able to stop
> immediately, and have a basic stepping top speed of
> around 200 - 800rpm (higher if you microstep) and
> relatively high torque depending on motor choice.


-- 
KIWACAN 
Electronics Design
Concept to Production

Mobile:  +64-21-249-5674
Phone:   +64-3-347-3311
Fax:     +64-3-347-2123
Website: http://www.kiwacan.co.nz
Skype:   kiwacan


_______________________________________________
Chchrobotics mailing list
Chchrobotics@lists.linuxnut.co.nz
http://lists.ourshack.com/mailman/listinfo/chchrobotics




This email may contain privileged/confidential information. You may not copy or disclose this email to anyone without the written permission of the sender.  If you have received this email in error please kindly delete this message and notify the sender.  Opinions expressed in this email are those of the sender and not necessarily the opinions of the employer. 

This email and any attached files should be scanned to detect viruses.  No liability will be accepted by the employer for loss or damage (whether caused by negligence or not) as a result of email transmission.
--=_alternative 006BD95FCC25725D_=
Content-Type: text/html; charset="US-ASCII"


<br><font size=2 face="sans-serif">Apologies if this appears twice.</font>
<br><font size=2 face="sans-serif">My Email address has been changed &amp;
I received a message from the list saying I wasn't subscribed.</font>
<br><font size=2 face="sans-serif"><br>
</font>
<br>
<br>
<br><font size=2 face="sans-serif">Hi All,</font>
<br><font size=2 face="sans-serif">It's good to see interest in Micro Mouse
starting up again.</font>
<br><font size=2 face="sans-serif">I used to race micro mouse at the &quot;Nelcon&quot;
competitions back in the 90s.</font>
<br>
<br><font size=2 face="sans-serif">At the time I used steppers as they
have inherent positioning which simplifies a lot of the control problems

for a beginner.</font>
<br>
<br><font size=2 face="sans-serif">The square shaped steppers from old
5.24' drives are ideal to get started with. </font>
<br><font size=2 face="sans-serif">I made wheels by turning up acetal plastic
disks that pushed on to the band drum on the stepper shaft. A groove for
an O ring makes a serviceable tyre.</font>
<br>
<br><font size=2 face="sans-serif">When using steppers there are a number
of things to watch for. Max speed is determined by both applied voltage
&amp; flyback voltage. A zenner diode to allow the flyback to reach 30
volts allows the 12V steppers to run much faster and have higher torque
at the high speeds.</font>
<br>
<br><font size=2 face="sans-serif">The other thing is that they do NOT
start &amp; stop instantly. When pushing performance they must be ramped
up and down or you will loose steps.</font>
<br>
<br><font size=2 face="sans-serif">Dead reckoning is a good starting point
but auto re-zero on the fly is absolutely critical if you hope to survive
more than one or 2 corners.</font>
<br>
<br><font size=2 face="sans-serif">The advise I received when I first started
proved very useful.</font>
<br>
<br><font size=2 face="sans-serif">Complete each of &nbsp;the following
steps in turn &nbsp;and don't proceed to the next step until the current
step is fully and repeatedly completed.</font>
<br><font size=2 face="sans-serif">1) make sure the mouse will fit in the
maze. (my first mouse wouldn't fit round some corner types) </font>
<br><font size=2 face="sans-serif">2) make it go forward in a straight
line and then stop.</font>
<br><font size=2 face="sans-serif">3) make it go forward exactly one square
and stop.</font>
<br><font size=2 face="sans-serif">4) make it go forward 1 square, stop,
and return to its original position.</font>
<br><font size=2 face="sans-serif">5) make it turn 90 degree and stop.</font>
<br><font size=2 face="sans-serif">6) make it turn 90 degree, stop, turn
back 90 and stop at the same heading as original.</font>
<br><font size=2 face="sans-serif">7) &nbsp;make it go forward 1 square,
stop, turn 90 degree, stop.</font>
<br>
<br><font size=2 face="sans-serif">At this stage you will have good basic
control and all the basic functionality. </font>
<br><font size=2 face="sans-serif">This is a good starting point on which
to build in the navigation and wall sensing.</font>
<br>
<br><font size=2 face="sans-serif">Only after you have basic navigation
and wall sensing going do you need to worry about solving the maze.</font>
<br><font size=2 face="sans-serif">If your mouse won't wander aimlessly
around the maze for hours on end without hitting walls then the most advanced
maze algorithm in the world won't help. </font>
<br><font size=2 face="sans-serif">(well at least 10 minutes.)</font>
<br>
<br><font size=2 face="sans-serif">Once you get a basic mouse running the
next challenge is a test maze.</font>
<br><font size=2 face="sans-serif">I build my test maze out of 2 sheets
of 12mm chip board, not MDF. (1200x2400)</font>
<br><font size=2 face="sans-serif">By cutting 50mm strips off the side
and ends of the sheets I ended up with a 10 X 10 maze. (2 sheets 1900 x
950)</font>
<br><font size=2 face="sans-serif">I found out the hard way that you need
a test maze that includes a centre.</font>
<br>
<br><font size=2 face="sans-serif">10x10 seems to be about the best size
for a home maze as it will stand up against the garage wall when not in
use.</font>
<br><font size=2 face="sans-serif">If you have a bigger garage you can
just scale it up.</font>
<br><font size=2 face="sans-serif">&nbsp;My maze is a single piece but
all the full size competition mazes I've seen are made in 4 quarters and
then dowelled &amp; bolted together.</font>
<br><font size=2 face="sans-serif">Beware the joints if you take this approach.
The APEC maze was more like a 4W drive track than a micromouse maze.</font>
<br>
<br><font size=2 face="sans-serif">I joined the 2 sheets by gluing one
of the strips to the back of the joint.</font>
<br><font size=2 face="sans-serif">The other strips were cut to length
as wall segments. &nbsp;Don't just cut them all to 180mm though.</font>
<br><font size=2 face="sans-serif">A full outer wall was glued to the base
and a few key wall sections were also glued in place.</font>
<br><font size=2 face="sans-serif">I could then use bluetac to position
lengths of walls as required to change the maze around.</font>
<br><font size=2 face="sans-serif">I drew a grid in pencil to help layout.
the lines were left/below the wall positions. </font>
<br><font size=2 face="sans-serif">Remember the maze is a grid on 180mm
centres but the walls are 12mm so the passage ways are only 168mm wide.
(See step 1 above)</font>
<br>
<br><font size=2 face="sans-serif">Chip board is quite reflective to infra-red,
which I was using for sensors. </font>
<br><font size=2 face="sans-serif">I didn't bother to paint it at all as
I was using over wall down sensors and could easily distinguish the wall
tops from the floor due to the much greater distance.</font>
<br>
<br><font size=2 face="sans-serif">I had an infra-red reflectometer that
I used to check competition maze reflectivity before a contest so I could
re-calibrate my sensors if necessary. </font>
<br><font size=2 face="sans-serif">Never required in practice as reflectivity
was usually within about 20% of my maze.</font>
<br>
<br>
<br><font size=2 face="sans-serif">Happy mousing.</font>
<br><font size=2 face="sans-serif"><br>
Regards,<br>
Robin<br>
<br>
</font>
<br>
<br>
<br>
<table width=100%>
<tr valign=top>
<td width=40%><font size=1 face="sans-serif"><b>Michael Pearce &lt;mike@kiwacan.co.nz&gt;</b>
</font>
<br><font size=1 face="sans-serif">Sent by: chchrobotics-admin@lists.linuxnut.co.nz</font>
<p><font size=1 face="sans-serif">02/01/2007 06:55 p.m.</font>
<table border>
<tr valign=top>
<td bgcolor=white>
<div align=center><font size=1 face="sans-serif">Please respond to<br>
chchrobotics@lists.linuxnut.co.nz</font></div></table>
<br>
<td width=59%>
<table width=100%>
<tr>
<td>
<div align=right><font size=1 face="sans-serif">To</font></div>
<td valign=top><font size=1 face="sans-serif">chchrobotics@lists.linuxnut.co.nz</font>
<tr>
<td>
<div align=right><font size=1 face="sans-serif">cc</font></div>
<td valign=top>
<tr>
<td>
<div align=right><font size=1 face="sans-serif">Subject</font></div>
<td valign=top><font size=1 face="sans-serif">Re: [chbot] HBRobotics/Micromouse</font></table>
<br>
<table>
<tr valign=top>
<td>
<td></table>
<br></table>
<br>
<br>
<br><font size=2><tt>I forgot the following...<br>
<br>
Since you are stepping, you know exactly how many <br>
revolutions the wheel has done so you can calculate <br>
position easily. Using a standard 1.8degree/step motor you <br>
can easily calculate 1/200 of a rev using single stepping.<br>
<br>
Using micro-stepping 1/200,000 of a rev is possible to <br>
determine (but not at a sudden stop)<br>
<br>
Up Sides:<br>
 1. Step counting for accurate positioning<br>
 2. Easy to build basic drivers (Full and half step)<br>
 3. High output torque (Can drive wheel directly)<br>
 4. Instant stop and hold at current position<br>
 5. Very repeatable.<br>
 6. Easily obtained in old dot-matrix printers, <br>
 &nbsp; &nbsp;5.25&quot; floppy drives and old hard drives (&lt;200MB)<br>
<br>
<br>
Down sides:<br>
 1. weight<br>
 2. power draw<br>
 3. losing steps (if hardware design bad or running to fast)<br>
 4. slow top speed &nbsp;(but dont need gearbox)<br>
 5. Not cheap to buy new compared to standard DC motor.<br>
 &nbsp; &nbsp;<br>
<br>
I may get stuck into my Pick and place project later this <br>
month, and will hopefully be able to demo it at the next <br>
meeting to show standard and microstepping techniques.<br>
<br>
<br>
Mike<br>
<br>
<br>
On Tuesday 02 January 2007 18:24, Michael Pearce wrote:<br>
&gt; That design is nice and simple.<br>
&gt;<br>
&gt; The steppers do have the advantage of being able to stop<br>
&gt; immediately, and have a basic stepping top speed of<br>
&gt; around 200 - 800rpm (higher if you microstep) and<br>
&gt; relatively high torque depending on motor choice.<br>
<br>
<br>
-- <br>
KIWACAN <br>
Electronics Design<br>
Concept to Production<br>
<br>
Mobile: &nbsp;+64-21-249-5674<br>
Phone: &nbsp; +64-3-347-3311<br>
Fax: &nbsp; &nbsp; +64-3-347-2123<br>
Website: http://www.kiwacan.co.nz<br>
Skype: &nbsp; kiwacan<br>
<br>
<br>
_______________________________________________<br>
Chchrobotics mailing list<br>
Chchrobotics@lists.linuxnut.co.nz<br>
http://lists.ourshack.com/mailman/listinfo/chchrobotics<br>
</tt></font>
<br>
<br>
<table><tr><td bgcolor=#ffffff><font color=#000000>This email may contain privileged/confidential information. You may not copy or disclose this email to anyone without the written permission of the sender.  If you have received this email in error please kindly delete this message and notify the sender.  Opinions expressed in this email are those of the sender and not necessarily the opinions of the employer. <br>
<br>
This email and any attached files should be scanned to detect viruses.  No liability will be accepted by the employer for loss or damage (whether caused by negligence or not) as a result of email transmission.<br>
</font></td></tr></table>
--=_alternative 006BD95FCC25725D_=--