Alt-BEAM Archive
Message #13399
To: beam@sgiblab.sgi.com
From: Ben Hitchcock beh01@uow.edu.au
Date: Fri, 14 Apr 2000 08:22:02 +1000 (EST)
Subject: Re: Core-less walker
Bruce, Wilf and others:
> Now isn't that an elegent circuit? It's just too bad you have to mess it
> up with motor drivers.
>
> How much current do those motors draw, anyway? The reason I ask is, you
> could pair up the inverters, and even THAT can be done elegently. Three
> inverters down each side of the chip, so just tie the inputs and outputs
> across the chip. That's something even I could freeform.
I tried that unfortunately :-( With the servos I have (the cheapest,
nastiest servos I could find) you need at least two gates of a 74HC245 to
make it move. This isn't so bad, however, because you can reverse the
gait with the addition of one more inverter on the control side of
things...
> Now wouldn't it be nice if you could do that? Two servos, one chip, and
> a decoupling capacitor (which you might not even need).
Too true. Maybe if you used micro servos... but they're a bit out of my
budget at the moment.
Now to wilf's observations:
> I love it! The basic principle of your design is a clear case of
> parallel
> evolution when compared to my "earlyBEAMservo1" circuit but your
> arrangement
> of using the three 74HC14 inverters to generate the right phase
> relationship
> between the legs is unique.
Geez, I wish I'd actually seen (read understood) your circuits before
rushing off to design
mine! I particularly like the way that you used a capacitor across the
wiper of the pot to generate a phase delay with non-schmitt trigger
inverters. Very nice!
By moving the inverter on the diagram shown pointing left to between the
two motors, I can reverse the gait of the walker. I think what I'll do is
make the circuit using four inverters, two of them with resistors on the
outputs, and have a transistor per 'reversing' inverter to bypass it.
Only one of these transistors will be turned on at any one time. These
transistors will be controlled by another gate (or pair of gates, I have
two left) which will trigger the reverser when the bot gets stuck. I
believe that this will mean that the robot will be able to navigate its
environment using no other sensors than the feedback pots on the servos!
More to come, pending experiments...
Ben
13400 Thu, 13 Apr 2000 15:50:43 -0700 (PDT) Re: Shipping times :p beam@sgiblab.sgi.com Daniel Grace Ok, I'll take the oppurtunity to say, HELP! I
understand that someone on this list runs solarbotics,
maybe Dave Hyr...(however you spell it). I just have
to say that I've tried to get a catalog about 2 or 3
times, and I have yet to get a catalog. Could you
maybe send me one if I sent to you off the list? Maybe
tell me who you are, and your e-mail addy, and then
I'll e-mail you my info. I've tried the form on your
page several times.
~Daniel
--- Gabriel DeVault wrote:
> Just wanted to say that solarbotics shipped faster
> than digikey to santa cruz, CA. The order was also
> perfect and everything was
> packaged beautifully. Hats off to solarobtics,
> excellent service :)
>
> Gabriel DeVault
>
>
=====
ICQ # 39402143
__________________________________________________
Do You Yahoo!?
Send online invitations with Yahoo! Invites.
http://invites.yahoo.com
13401 Fri, 22 May 1998 14:31:41 -0700 It's alive !! STOP HACKING THOSE SERVOS!! Throw away your springs. "beam@palladium.corp.sgi.com" "David@Winchell.Net" Hello World, Tomorrow we WALK. It's time for the next step!!
It's alive !!
STOP HACKING THOSE SERVOS!!
Throw away your springs.
Stop falling over in mid stride.
Stop trying to build pager motor gear boxes
and start building killer walkers.
A given from M. Tilden:
Watch as machine evolution occurs.
In other words, robogenetics through robobiologics.
NOTE: THE FOLLOWING CIRCUIT DESCRIPTION IS COPYRIGHTED BY
DAVID WINCHELL AS OF 22 MAY 1989 AND MAY ONLY BE USED BY
BEAM MEMBERS FOR ENTERTAINMENT. NO COMMERCIAL APPLICATIONS
ARE AUTHORIZED WITHOUT WRITTEN PERMISSION. WHAT BELONGS TO
MARK TILDEN IS HIS AND MINE IS MINE. OH JUST HAVE FUN
BEAMer'S
My new walker NON-MODIFIED SERVO brain has run for 24 hours
without going in to saturation.* I used 2 * 74HC14's and one
74HC125. One 74HC14 acts as a dual channel SERVO POSITION
UNIT "SPU" Two of the inverters are wired to form a BI-core.
two more inverters are wired to form half monostable one
shots. R1*C1 AND R2*C1 are set to generate a 45 HZ rate
triggers to the R3*C3 and R4*C4 half mono's. The R3*C3
half mono is set to generate a 1.25 ms positive going pulse
or 45 degree CCW position. The R4*C4 half mono is set to
generate a 1.75 ms positive going pulse or 45 degree CW
position. By using an SPU you can build and test your frame,
legs and test it with one chip on a tether.
--------------------------------
| 1 2 3 4 |
----|>--+--C1--+--|>--+--C2--+--
| | | |
| R1 | R2
| | | |
| G | G
| |
| | 5 6 9 8
| ----C3--+--|>----|>---- CCW POSITION
| |
| R3
| |
| G
| 11 10 13 12
------------------C4--+--|>----|>---- CW POSITION
|
--|>-- 74HC14 R4
|
G
If you use two push buttons you can test run your servo.
___________
+6VDC | |
----------- | Servo red wire ---------
| | -------------------+ |
| | CW___ | |
| +------o o----- Servo white wire | |
| SPU | CCW___ +--------------------+ Servo |
| +------o o----- | |
| | | |
| | -------------------+ |
----------- | Servo white wire ---------
G
Now that the SPU is working add the 74HC125 SERVO CONTROL
UNIT or SCU.
1 ___
---RS-CW--o o--G
2 | 3
----------- --------|>----------------------+
| | | 4 ___ |
| | | ------- FS-CW--o o--G |
| | CW | 5 | 6 |
| +------+--|>---+ |
| | | |
| SPU | +--Front Servo White +--Rear Servo White
| | CCW | |
| +------+---|>--+ |
| | | 9 | 8 ___ |
| | | -------FS-CCW--o o--G |
| | | 10 |
----------- ---------|>---------------------+
12 | 11 ___
--RS-CCW--o o--G
13
74HC125
---- Tri-State Control, Ground to enable.
|
In --|>-- Out
Note: I put 10K pull ups in 74HC125-PINs 1,4,10 & 13
74HC125-PIN 1 = G Then Rear Servo = CW RS-CW
74HC125-PIN 10 = G Then Rear Servo = CCW RS-CCW
74HC125-PIN 4 = G Then Front Servo = CW FS-CW
74HC125-PIN 13 = G Then Front Servo = CCW FS-CCW
Now for the final step add the a Stabilized BEAM Quad Core. This
design may be expanded to drive a 12, 16 and so on servo walker.
Lookout MIT this may be the new wave. But the TIMING may Vary
with user and cause variant and deviant behavior.
--------------------------------
| 1 2 3 4 | Core Stabilizer
----|>--+--C5--+--|>--+--C6--+--
| | |
--------- R5 R6
| | |
| G
|
|
| Open loop Quad Core.
| G1 G2 G3 G4
--C7-+-|>-+-C8-+-|>-+-C9-|>-+-C10-+-|>---
| | | | | | |
R7 | R8 | R9 R10 |
| | | | 1N914 Diodes
| | | --+--|>|-- TO 74HC125-PIN 4 - FS-CW
| | | |
| | | ---|>|-- TO 74HC125-PIN 10 - RS-CCW
| | |
| | |
| | --------+--|>|-- TO 74HC125-PIN 4 - FS-CW
| | |
| | ---|>|-- TO 74HC125-PIN 1 - RS-CW
| |
| |
| ----------------------+--|>|-- TO 74HC125-PIN 13 - FS-CCW
| |
| ---|>|-- TO 74HC125-PIN 1 - RS-CW
|
|
--------------------------------+--|>|-- TO 74HC125-PIN 13 - FS-CCW
|
---|>|-- TO 74HC125-PIN 10 - RS-CCW
Bi-Core State:
Quad-Core Front Servo Rear Servo
===================================
G1 = LOW CCW CCW
G2 = LOW CCW CW
G3 = LOW CW CW
G4 = LOW CW CCW
* Use .01 ceramic caps on all power leads.
Use 10uf electrolytic on servo power leads.
Home