Penguin Programing help!!

Zax

Ok first off I have calibrated my STRIDE and TILT servos dead on, and when running these 2 programs it walks straight as an arrow! Penguin-AutonomousNavigation-v1.2.bpx or penguin_happy_feet.bpx But when running Penguin-AutonomousNavigationNoCompas_3.bpx it constantly walks to the Right! even when I recalibrate the TILT and STRIDE to compensate for this, and save it the program acts like its not loading the new settings from EEEPROM memory at all! So I thought that the photoresistors were causing it but even in total darkness it still does it! I am kind of stuck on truble shooting since its a programing issue of some kind thats out of my league it backs up perfectly straight thats what has me stumped, if the photoresistors were removed from the program would that help? I would rather just use the Irs for Avoiding obstacles but don't know how to chop them out of the code here it is
thanks Rick

Post Edited (Zax) : 4/6/2008 4:38:11 PM GMT

Humanoido

Curious, where did you find the code,
Penguin-AutonomousNavigationNoCompas_3.bpx?
At first glance, I see all the compass code remaining.
(Look for HM55B and the word compass.)
If you want to stop the directional navigation, this
code can be removed, and will simplify the overall
program. At this point it will be a chore to modify
the remaining compass turn related statements.
Then its a choice of routine, such as

DO
  * walk forward
  * detect object
  * back up
  * turn
LOOP

Why not write your own code? It will be a wonderful
learning experience, you can post your results, and
everyone will benefit and appreciate your work.
All of these routines have been posted in various
applications for Penguin and can be linked into one
program.

humanoido

Zax

Hello Humanoido, First of all thank you Very much for your Rapid Reply!!!!
But Thank you Very Very Much For al the Great CODES you have provided here!
I downloaded the code here on the Forum somewhere, but yesterday I pasted the wrong one.
This is the No compas 3 code All the codes but the 2 I mentioned before cause a right walk! and when I try to chop out the photoresistors I get some strange results LOL!!! this is the code I would like to modify
Thanks Again, Humanoido for all the GREAT WORK of yours!! you are in a league thats your own!!

Post Edited (Zax) : 4/6/2008 4:38:55 PM GMT

Humanoido

Hey thanks Zax! It's appreciation like yours that
inspires me to write Penguin programs.
This program is a mod by Interact
and can be downloaded here, along with the
modified compass version and text description.

Penguin Another re-work on the Parallax AutonomousNavigation
http://forums.parallax.com/showthread.php?p=672251

As part of the normal functioning of the code,
Penguin will implement turning while walking.
I have highlighted an area of interest in Interact's
quote.

IR detectors, used for obstacle avoidance, receive
reflected IR from the IR transmitter, and work in
total darkness. Cutting out that code would cripple
obstacle detection and give strange results.

humanoido

Interact said...
Another thing about this re-work is that the step routines take into acount which foot is allready forward. this was another reason the poor guy would get "stuck". Also.. object avoidance logic with this simple routine.. The line below in the main loop that has the red hilighted Chirp - 2 can be adjusted to 4 for better results on most objects. The result is that after the penguin takes a few steps backwards while turning away from an object it will go straight for (in this case 2) steps before resuming looking for the light. It will still look for obstructions during this time. If ojects are white the penguin will see them sooner than dark colored objects etc. If objects are big or small also causes a slight change to the ideal logic here. I found that Chirp = 4 is best for most things. Also make sure your penguin CAN walk straight! if it tends go drift left, re-calibrate him/her so that he/she leans a bit to the right. Just setting it up so that it is perfectly vertical dosen't mean it will walk straight.

Zax

Thanks again Humanoido!!! He Who is One with Penguin!!! I just picked up a remote and I am running Penguin Robot Wireless Soccer.
I know the trouble is simple but this code works perfect but it still does not run or walk straight it still moves to the right and the left foot is always half a foot ahead of the right one. When walking during the calibration program he walks straight as an arrow!
I am missing something here, even when I change the calibration to compensate he still turns right during the soccer code he will not go left He goes Straight, but he will sure make a sharp right LOL!!

Humanoido

Thanks Zax for your detailed description. It sounds like
your Penguin Robot has one servo that needs replacing.
Providing that this Penguin is mechanical tuned, with no
binding from side bolts, etc., and the batteries are checked
and new (very important) and because the soccer
code is tested to make Penguin go straight, I would say
it's a done deal about the servo. When one foot is in
front of the other, this is the key. Clearly, confusion
could result when the calibration program is used in an
attempt to center this servo. This happens to me when
I use Penguins so much that a servo wears out!
With such a worn servo, it's possible to bring it back to
the home position, but then running software will cause
the little bot to move in unexpected ways.

humanoido

Zax

O. K. Humanoido I will try fresh batteries.
I have 2 boes, and 3 sumos ,and have considered the servo isue.
I think it would have to be the Stride servo if any, I will change the positions Tilt to sride and stride to tilt and see what happens.
As far as binding no way the little guy is fit as a fiddle there is no binding anywhere.
What has me confused if he walks straight in the other programs such as Penguin-AutonomousNavigation_v1.1 and Happy Feet you would think logic would say it should in the others!
Trust me on this oh GREAT MASTER STAMPER I am eager on you next upcomming codes!!
I am having a Ball with your Soccer Code even if I cant turn Left Lol!! It is to much FUN!!
Is it posible to use a say a RCA remote, so you can have two of the Little guys running at the same time ? I am using A RCA for mine With Sony TV code It would be a blast to have 1 try to to knock the the other down with a good kick in the foot!! LOL
Thanks again Humanoido, for your very VALUABLE TIME
Rick

Humanoido

Let me know what you find after swapping the servos.
Not being able to turn can also be related to the tilt
servo and not just the stride servo. This is because
turning employs a small tilt, one foot goes forward,
the tilt is canceled, and the feet are brought together
or motioned past each other. If both servos have
worn out their positioning ability, swapping may create
some other results.

Penguin will need calibration again before running any
programs. Some of these programs may be helpful for
calibrating and study.

11 penguin_automatic_servo_calibrator.bpx
http://forums.parallax.com/showthread.php?p=671204

15 penguin_calibration_companion.bpx
http://forums.parallax.com/showthread.php?p=671346

57 penguin_tiltstride_reader.bpx
http://forums.parallax.com/showthread.php?p=691481

30 penguin_motion_startup.bpx
http://forums.parallax.com/showthread.php?p=695151

52 penguin_startup_instructions.bpx
http://forums.parallax.com/showthread.php?p=690297

You have a very good idea here! To use the RCA codes
and remote on one Penguin and the SONY standard codes
and remote on another Penguin will require one Penguin
to have code running to recognize SONY and another
Penguin to have code recognizing RCA. Someone will
need to write the RCA code for Penguin.

humanoido

humanoido said...
Work is progressing on a new upcoming Penguin
VSC Virtual Sound Chip and additional code
that will demo four methods to create special sound
effects. You will learn how to create sound with the
following coding techniques:

1) FREQOUT
2) BAUD
3) TOGGLE
4) PULSOUT

Zax

Thanks a Bunch, Humanoido!!
Yes, I will let you know what happens!
I wonder if you know Someone Who Would write the RCA code for Penguin, Hint Hint!!
I thought it would be simple just changing a couple numbers for the input data but I have all ready found out there is more to it!
Thanks for all the help Humanoido!!!!!!!
But most of all for "THE SOCCER CODE OUTSTANDING WORK!!!!!!!!!!!!!!!"
Rick

Zax

Hey there Humanoido,
I changed the batteries, and changed the servos, from Tilt to stride and stride to tilt and still get the same results.
IT walks straight as an arrow when running Penguin-AutonomousNavigation-v1.1.bpx, or when I walk in Penguin-ServoCalibration-v1.2.bpx
But when running Penguin-AutonomousNavigationNoCompas_3.bpx or Soccer it constantly walks to the Right! even when I recalibrate the TILT and STRIDE to compensate for this, and save it, the programs acts like its not loading the new settings from EEEPROM memory at all, no matter what position I calibrate the servos in, they remain in the same position in these programs but in the other 2 that it walks straight in, if the stride and tilt are off a little you can see the difference in the walk in other words I think I have narrowed it down Does the Soccer program not move the servos to the set positions I set and save in the Penguin-ServoCalibration-v1.2.bpx? I even went as far to calibrate it with the tilt and stride way way off center, but when I ran Soccer and AutonomousNavigationNoCompas_3.bpx it made no change, the center positions were the same.
Thanks again Humanoido,

Humanoido

Try this calibration program instead:
penguin_automatic_servo_calibrator.bpx
http://forums.parallax.com/showthread.php?p=671204
The calibration changes to the stride motor will be evident
and the position of the feet are easily adjusted about
the front even point. I suggest calibrating with this software
when running any of the humanoid penguin suite programs
that call for calibration. (see attached file below)

After calibration with the above mentioned code, the soccer
program, which is part of the humanoido penguin suite of
programs, will reflect this calibration in the walk and the
effects will be apparent. After the walk is completed,
the home code brings the feet back to the calibrated
center position. The walk, during the walking process,
is straight. However, the home code, normally induces
a slight turn. This is because the foot is not raised when
the two feet are returned to center home position.

I believe the turn you describe in the soccer program is
from the home position written into the code cycle when
the feet are brought together. This happens in the subroutine
Center and is the normal functioning of the program.

You can disable this code, but would need to write new code
subroutines to compensate and remember the ending
position of the feet, and the beginning of the next
start cycle. The key in doing this is to alternate the start
and stop positions from feet, i.e. for example, when stopping
the left foot can be fwd, and when starting, this would
become the starting position. This would be a good upgrade
idea for the next code version.

Zax, thanks for your keen eye in drawing attention to this
feature. Attached is a preliminary list of programs that are
in the humanoido Penguin Robot suite. The reference can
help decide which programs can be calibrated with
penguin_automatic_servo_calibrator.bpx

Also please refer to the Penguin Resources sticky at the
top of the Robot Forum, to determine who created and
modified which code. It will give an idea about who can
best support the various Penguin code.
http://forums.parallax.com/showthread.php?p=675357

humanoido

Post Edited (humanoido) : 4/7/2008 4:51:22 PM GMT

Zax

Thank you thank you Humanoido!!!!!!
I am up and running 100% your penguin_automatic_servo_calibrator.bpx did the trick I had a feeling it was something simple even the other programs I had that were walking to the Right are now PERFECT!!!!
Thank you so much for the help!!!!!!!!!!
I am sorry it took so long for you to trouble shoot it but that was it!
Thanks again Humanoido!!
Rick