Let's Make Robots!

08M2 had attitude problem; discipline applied, now behaving nominally :D

So this is for the hexaspider project I'm working on. Here's a rough timeline:

  • breadboarded circuit -  bot works
  • make functioning circuit board - bot works
  • while working on gait programming - bot stops working
  • assuming my board is to blame, I etch another better one - bot still not working
  • put 08M2 on breadboard and route everything from circuit board to breadboard - bot works
  • plug 08M2 back into circuit board - bot stops working

I have a very simple "servo left, servo center, servo right, repeat" program loaded. In fact, it can be any program I have. All that happens is the servo will briefly pulse in one direction every second or so. The pauses between the servo pulses do not match the pauses in the programming, fwiw.

I just don't understand how it works when the uC is on the breadboard but not on the circuit board. And all this just happened after a program download. I've tried old programs, new chips, new boards, new batteries, and everything bulletpointed above. Anyone have advice for me? I'm stumped.

Pics of schematic and PCB if anyone is interested:

 

Comment viewing options

Select your preferred way to display the comments and click "Save settings" to activate your changes.

Looks like it's constantly resetting because of a power drop when the servos move or the serial input pin (leg 2) isn't staying low. Try measuring voltages while it's running .

 

You're advice was the key. I don't know how it's possible but there was a short or something between the left tactile feeler headers. When the lipo was plugged in the two pins showed continuity but not when unplugged. When tested while unplugged there was no continuity but I got a low reading like 80-90 on the multimeter (ohms I think?). The 10K pulldown resistor only read ~3K as well. So I applied the iron to the legs of the headers, let the solder paste boil a bit more, and plugged it in.

Huzzah eureka OMG!!!!!!!11oneonewonwonwinning! #winning! #reallywinningnotCharlieSheendruggedoutwinningha

At risk of sounding boneheaded, how is it there is no continuity when the battery is unplugged but there is continuity when the battery is hooked up? Is the low ohm reading basically an almost-continuity condition?

Rick100, you should rename yourself to Rick1million :) You save me from project-ending-discouragement, thanks m8!

that I find it remarkable that this was the same problem on the previous circuit board. I remade the board and put the same problem in it.

As I've said before: Once is chance. Twice is lightning. Three times, well, we got a problem. Let's hope it was just lightning in this case.

The serial input leg will serve double duty as a tactile sensor pin as well. When it triggers it will go high so its tied to a pulldown 10K resistor normally. I doubt that is the issue but I will give your advice a go, especially the running voltages. Thanks Rick.

I'm happy to help and glad your making progress . I haven't tried using the serial input as general IO yet , but I think you have to use the disconnect command so the processor doesn't scan for a new program download . You'll have to cycle power to download a new program . I'm not sure I understand the readings you were getting but if the circuit is under power (battery plugged in) you won't get the correct resistance reading .  Hope this helps .

Good luck

Rick

I looked at your video via YouTube and noticed this other one: http://www.youtube.com/watch?v=w5eFc_B1lHY&NR=1

That is the same robot, right? If so, then it was working at one time. Thus I wonder what changed in between? --chip plugged into soclet upside down? Yeah, I know, not likely...

...so back to what is different on the breadboard; I see a lot of extra wires on the breadboard... Are there any other components on the breadboard, like a pull-up resistor or anything?

On the other hand, it only looks like about 4(?) wires hook over from the breadboard, not all 8 wires of the picaxe chip, so is there another circuit that could be shorting something out when all 8 pins of the picaxe are hooked up to something by plugging the chip into the robot board?

If you have a volt meter, I would be interested in checking that the voltage does not change when the picaxe is moved. If it does change, then something it shorting or at least putting extra load and drawing down the batteries.

 

Yes, same robot, one day before the problem set in. Curious, no? Hehe, no upside down chips. Not this time :)

The extra wires are there to confuse and confound. There are a couple resistors for the download circuit but that's all. The mass of wiring near the PICAXE are the old connections I was using for the other two servos. The wiring downwind of the chip are all artifacts of a previous test circuit. I love spaghetti <3

Correct, there are only four wires: System 5V, ground, servo's 5V, and the servo signal line. Hmm, well, I've tested for shorts but less connections is a difference. I'll try and connect all the pins and see what happens, thanks Dan.

I didn't get to test out your suggestion but I'm sure it would have failed on the breadboard too if I'd hooked up the left tactile pin. Seems that was the fouled beasty. See celebration notes above lol

 

     It was the power cable.  A broken solder joint under shrink tube can be weird like that.  But after i watched the video i suggest getting a magnifying glass and checking the ic socket.  Wires are a lot larger then the pins and could be making contact when one of the pins is not.

 

Good luck

would be my first suggestion to check too.

Plug the chip into the unpowered board and do a continuity check between the top of the PICs leg and the bottom of the solder joint for the socket.