More Problems - Admin if you can help

[soninja8]

Ok, so I got my $50 robot programmed and I plugged in all my servos and photoresistors, and it sorta follows light....I have to point a flashlight directly at the center of the Photoresistor to get it to move that way. Also, one wheel is going a lot faster than the other, even though both are getting the same voltage, I checked with a multimeter. All connections on my board are good....I dont get it. The wheel that goes super slow, goes normal speed backwards, but when it goes forward, it is a snail...Please Help!
Thanks,
Soninja8

[vinniewryan]

output to each motor seperately and monitor the current consumption. If there is a significantly lower amount of current running to that motor, it could be a blown driver. If there's a higher amount of current drawn by the motor then it may be something jamming up the motor (dirt, hair).

Honestly it sounds like a power issue on your board. I would check the impedance of the power lines to the motors and motor drivers, along with common ground and power lines with an ohm meter. If you test the board, you will find a problem somewhere. As for the light sensors, again this could be a power issue or bad connection/ground. Make sure there aren't any shorts or loose solder joints on the board, you might even go through and re-melt some of the connections that look iffy. If everything checks out then I would assume a blown output on the motor driver or a wrong connection.

Check everything, post your results.

[soninja8]

Ok, its the photovore__v1.hex file then....I uploaded hold_servo.hex to my board, and the motors spin at full speed continously (I have the continuos rotation mod) So it appears that I have good welds and power on my board....can someone upload there Photovore_v1.hex file, so I can put it on my board to see if it works properly?

Thanks,
Soninja8

[Conscripted]

Are you using servo's that you modified your self? If so what process did you use to modify them? I think maybe your servo's aren't set at the center position.

Conscripted

[adanvasco]

Hi.

Can you post or upload your Photovore_v1.c file? There might be a problem on the code.
You're not going to find out much information by measuring the voltage on the servos since they are controlled by changing the length of the pulse from 1ms to 2ms.

[soninja8]

Ok, well my problem is the program...the Photovore.c file is not the problem....there is just something wrong with my avrstudio....can comeone please email or upload me their photovore.hex file so i can put it on my robot and see if it works?

Thanks,
Soninja8

[Admin]

Here is my file.

[soninja8]

Hey Admin!! Thanks for the file....but when I try to upload from Avrdude to my atmega8 it says that it autodetected it as an "Intel Hex" and that "ERROR: address 0x2010 out of range at line 513 of me.hex" Every other .hex of mine uploads fine....

Please Reply Quickly!

Thanks!
Soninja8

[Admin]

oops . . . the previous file was for the ATmega168 at 16MHz.

This attached file is an older file I found . . . not sure if it'll work, but give it a try.

[soninja8]

Ok....its the same results....my second servo input moves so slow, I tried both of my servos on it....I have checked all my welding millions of times.....the powers correct. Whats wrong with it?

Heres my pics of my board...

Thanks,
Soninja8

[Admin]

my second servo input moves so slow, I tried both of my servos on it

You mean one servo appears to be working fine, but the other isn't?

What happens when you swap the two servos, does the same servo cause the problem, or the other one?

Any issues when modifying the servo?

[soninja8]

yeah, i have tried both servos on it...both servos work fine...but not on servo input 2

and admin, i think the photovore v1.hex that you uploaded is not the right one...its the same as the photovore v1.hex as thats in your file that you released with a bunch of tools to make the real photovore v1.hex

do you have the photovore v1.hex that was produced by avrstudio?

Thanks,
Soninja8

[Admin]

What I have is what I have, and that's what I posted

And when you say 'input', do you mean 'output'?

Try changing your code to use a different output port for your servo and see what happens. Perhaps that port is whack.

You will see it in SoR_Utils.h:

PORT_ON(PORTD, 1);//change that number to something else
PORT_OFF(PORTD, 1);

[soninja8]

When I build Photovore_v1.hex in avrstudio, is the build message supposed to look like this?
Build started 9.6.2011 at 22:02:10

-------- begin --------
avr-gcc (WinAVR 20100110) 4.3.3
Copyright (C) 2008 Free Software Foundation, Inc.
This is free software; see the source for copying conditions.  There is NO
warranty; not even for MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.


Compiling: Photovore_v1.c
re_v1.o


Compiling: a2d.c

Linking: Photovore_v1.elf
put Photovore_v1.elf -Wl,-Map=Photovore_v1.map,--cref    -lm


Creating load file for Flash: Photovore_v1.hex
avr-objcopy -O ihex -R .eeprom Photovore_v1.elf Photovore_v1.hex

Creating load file for EEPROM: Photovore_v1.eep
avr-objcopy -j .eeprom --set-section-flags=.eeprom="alloc,load" \
    --change-section-lma .eeprom=0 -O ihex Photovore_v1.elf Photovore_v1.eep
c:\WinAVR-20100110\bin\avr-objcopy.exe: --change-section-lma .eeprom=0x00000000 never used

Creating Extended Listing: Photovore_v1.lss
avr-objdump -h -S Photovore_v1.elf > Photovore_v1.lss

Creating Symbol Table: Photovore_v1.sym
avr-nm -n Photovore_v1.elf > Photovore_v1.sym

Size after:
Photovore_v1.elf  :
section    size      addr
.text      1164         0
.bss          1   8388704
.stab      2820         0
.stabstr   2523         0
Total      6508



Errors: none
-------- end --------

Build succeeded with 0 Warnings...

[soninja8]

Ok, I changed the port like you suggested, to port 3....but now port 3 is very slow, and port 1 is still fast, and all the others are not programmed to work....its gotta be the code or something...

Thanks,
Soninja8

[adanvasco]

Can you post your Photovore_v1.c file content? Just to make sure there is nothing wrong with it. Also, when you switched the servo to output port 3, did you change your code to use that port? If you didn't, on your SoR_Utils.h update the line that says

Code: [Select]

PORT_ON(PORTD, 1); to

Code: [Select]

PORT_ON(PORTD, 2);.

You might as well just have a bad MCU.

[Soeren]

Hi,

I have checked all my welding millions of times.....the powers correct. Whats wrong with it?

Heres my pics of my board...

Photos of the solder side shows that you could have an error (or several) in that department and if it's not the case now, it will be when those burned wires start slapping against the solder connections - not trying to put you down, but that is by far the worst assembly I've seen in years, if not decades and I'd recommend a complete clean-up (i.e. touch up the solder and use new wires laid neatly and in a way that you don't get your soldering iron too close to them).

OTOH, I doubt that your "servo 2 output slow forward" is caused directly by this, unless it eg. loads the output, making the servo read the pulse shorter than they're meant to be, but you really need to examine the output with an oscilloscope, to see what timing and voltage the servo gets under running conditions (put the probe directly on the signal pin of the servo connector and note down the pulse width seen on forward, back and halt.
This is the only way to find out what's really going on inside the circuit.


You might be able to get access to an oscilloscope at your school, especially if you tell a science teacher about your project and the character of the problem - shouldn't take you more than 15 minutes, if you know how to set up the 'scope.

If that's not possible, a free Sound Card Oscilloscope program (and a 3.5mm stereo jack, 2 resistors and a cap) could be used to verify the approximate pulse widths.

[Admin]

Well, it's probably his first time soldering so cut him some slack
(soninja8, a lot of people had trouble with this $50 robot circuit soldering their first time, so don't feel bad)


soninja8, try this and let us know what happens . . . in the code, you have (PORTD, 1) for one servo and (PORTD, 2) for the other. Swap them around in software so that 1 is 2 and 2 is 1.

[soninja8]

Ok, I know my soldering looks back, but its pretty good. There is no solder touching whatsoever, and Ive sucked up all the excess with braid....

And admin, I have swapped the port code to 0 and 1 and 2 and 3, and it always seems that the port that is supposed to be the right servo, has all the power...I have tried changing the pulse to 25 for the left, and it runs as the same speed as the right, but it goes in reverse...

I am not in school so I have no access to an oscilloscope at the moment.

[Admin]

I have tried changing the pulse to 25 for the left, and it runs as the same speed as the right, but it goes in reverse...

So, you are saying both servos worked perfectly fine at the same time, except one was going in reverse?

[soninja8]

yes...it seems that all pulses 40 and up make it go forward, but very slow, and pulses from 30 and down make it go fast backwards....I havent done much other pulse shifting....

The pulses are located in the Photovore.c file

[Admin]

This means you didn't modify that servo properly. Change the pot position so that the funky servo isn't moving when the pulse is in the middle.

[adanvasco]

Make sure to put some crazy glue on the pot to stop it from moving once you put everything back together.

[soninja8]

This means you didn't modify that servo properly. Change the pot position so that the funky servo isn't moving when the pulse is in the middle.

your saying that both servos arent modified correctly, because they both act funky on that output...? and what is considered the middle of the pulse?

[soninja8]

omg, i'm so stupid  ....I have to upload the hold_servo.hex and move the potentiometer untill the motor stops, then glue the potentiometer down so it doesnt move....

Ok, I glued it down after uploading the hold servo, and now i'm going to sleep and wake up to a lot of work ....

Thanks,
Soninja8