Potentiometer not working?

[Stan Hurd]

I am on robot building Step 12: Test Individual Servos.  When I rotate the pot clockwise from the Stop position, nothing happens.  Also, I get no voltage reading across A1 and A2.  Is that the problem or is it the pot?  The pot is wired with the power (red) on the left, ground (black) on the right and signal (white) in the center.  Any suggestions for troubleshooting?

Another issue:  Every time I turn on the power, one of the legs moves a little bit to the side.  So, after turning the power on and off a few times, that leg is all the way over to the side.  The other legs don't move at all.  Not sure what this means.

Thanks for any help.

[Stan Hurd]

Actually, its 2 legs that move as described below.

[Steve Pendergrast]

Hi Stan,

So to clarify, when the pot is set to STOP, the robot is supposed to assume a standing position. Is this not happening?  If not, what is happening in STOP position?

I vaguely recall a robot I built once that had this kind of behavior where a couple of the servos went wacky.  Unfortunately, that was probably a year ago and I don't recall the solution. It is clearly either an electrical system issue or a bad nano firmware flash.

One thing that might cause it and is easy to check would be if the wires from the nano to the servo controller are incorrect. In that case, the servo controller is going to get garbage commands and might do just about anything.

Another possible cause that's easy to try could be that the nano has lost its program or its program is damaged. This can happen due to static electricity.  If you can flash the robot program onto it again, that might just fix everything. There are instructions in the support thread for flashing the nano (and I'm working on getting that info over to the wiki where it really belongs.)

If you are certain your wiring is correct, and you've tried flashing the nano again with no luck, I can send you a new electrical system to get you up and running.

FYI, the pot wiring is supposed to look like this:  (If it doesn't, somebody made a mistake when wiring, so please let me know).

[Stan Hurd]

After looking at this more carefully, I can tell you what is going on. When I first powered up with the pot in the stop position, some of the servos moved a little bit.  Then I discovered that these servos moved a little bit every time I turned the power on and off until they reached some end state.  So, when all of the servos stopped moving, I took the horns off and repositioned the legs and knees so they were in the standing position.  I thought the servos should have moved all the way to the end state when the robot was first powered up rather than in little increments when the power is turned on and off.  Anyway, this problem seems to be solved.

The pot is still a problem.  Nothing happens when I turn it clockwise from the stop position.  Early on, one of the wires connected to the pot came loose and I soldered it back on. It may be that the pot was damaged by this.  It might be worth trying a new potentiometer if you can send me one or tell me where I can buy one.  Other than that, I have no idea what the problem is.  I have checked all the wiring and it seems to be OK.  Thanks.

Stan

On Thursday, November 16, 2017 at 10:33:18 AM UTC-8, Stan Hurd wrote:

[Steve Pendergrast]

Ok, that's good news, but sorry about the bad connection. I will ship you a new potentiometer first thing in the morning.

[Stan Hurd]

Thanks Steve.I hope this will solve my problem.

Stan

On Thursday, November 16, 2017 at 10:33:18 AM UTC-8, Stan Hurd wrote:

[Stan Hurd]

On Thursday, November 16, 2017 at 10:33:18 AM UTC-8, Stan Hurd wrote:

[Stan Hurd]

Steve: Why did you send me a big tube of hair conditioner with the new pot??

On Thursday, November 16, 2017 at 10:33:18 AM UTC-8, Stan Hurd wrote:

[Jim Neill]

I'm having a similar issue with my robot. I got to the step in the installation guide where you move the pot slightly to do the alignment and nothing happens. I tried flashing the latest firmware from github and still no luck. When I turn the pot the red light on the nano goes from flashing occasionally to on solid, so I think the pot is functional. I've verified the batteries are fully charged, etc. Is there some form of debug mode while connected to usb that I can do to tell what's happening?

[Liam McMullan]

Hi All

I seem to be having a similar problem to Stan,

The pot does not appear to do anything. Every time I power up the legs move slightly up. Rotating the pot does not do anything except to cause the buzzer to start buzzing until I get to the RC indicator and then it goes silent.

I will try re flashing the Nano, and I will also try another Nano that I have to see if that is the issue.

-Liam

[Stan Hurd]

OK, it looks like more people are having the same problem that I had.  I started out with nothing happening except the legs moving slightly every time I turned the power switch on.  Then i reflashed the program (actually twice to be sure I had it right) .  After that, when I turn on the power there is a pause of about 3 seconds, then a beep, then all the servos quickly go to their extreme positions (all the knees full counter clockwise and the legs tucked all the way up), then a second beep (a different tone).  All this takes place with the pot set to Stop.  Turning the pot makes no difference (nothing happens).

I am waiting to see if Liam's second Nano works.  That could be the problem for me too. 

On Thursday, November 16, 2017 at 10:33:18 AM UTC-8, Stan Hurd wrote:

[Steve Pendergrast]

Hi Stan,

OK that is actually progress even though it seems not to be.

When you power up in STOP mode, you should indeed hear a beep after a few seconds, followed by the legs going into a STANDING position.

Because you say the legs are going to an extreme position, this indicates the servo horns are not adjusted to the correct positions.  The reason why you don't notice any change when you turn the dial is that the legs are so far out of alignment that the slight motion doesn't even bring them back to the point where they move.

Also, I see from your description that you actually attached the legs to the robot. You were not supposed to do that until *after* the servo horns were adjusted.  I will look at the wording on the instructions page to see if there is a better way for me to describe this.

So, here is what you need to do:

1) Take the legs off.  Leave the leg servos connected, and leave the horns on, but disconnect the legs from the leg hinges.

2) Take the hinges off the body servos as well. Leave the servos on, and leave the horns on the servos, just don't connect them to the hinges.

3) You will now have a robot where all the wires are dangling out the legs are connected to the servo controller, but the hinges are not connected to the servos on either side.

4) Turn the dial to STOP and then turn on the robot.

5) Adjust all the horns on the body servos so they stick straight out from the robot body (or as close to straight out as you can get).

6) Very slowly turn the dial counter clockwise. After just five or ten degrees of turn, the KNEE servos only (the ones on the legs) should twitch to a new position which is about 30 degrees different from before. This is "leg adjust mode" and it puts all the shafts at a 90 degree angle.

7) Adjust all the KNEE servo horns so the horn sticks straight out the top of the leg (or as close as possible to straight out, as shown in the diagram on the assembly page.

8) Use test mode to see if all the leg servos are functioning (test mode is a little more clockwise than leg adjust mode).

Let me know at what step the above fails if that doesn't work.

[Steve Pendergrast]

Hi Liam, see my reply to stan below. Apparently flashing the nano did get stan a little farther, but apparently the horns were just way out of wack on the leg servos.

The latest tested version of code is in the support dropbox folder here:

https://www.dropbox.com/sh/rib02u0tt8i0n0p/AAD1dWWoP9z36tnF2sCOfDbZa?dl=0

[Steve Pendergrast]

Yes if you connect the USB to the robot nano you will get a stream of debug info telling you what the robot thinks it should be doing.

But some questions:

1) If you start up the robot in STOP mode, are the servos under power? Do they move at all, even a little twitch? Does the buzzer in the robot beep a few seconds after turning it on?

2) To see if the servos are under power, see if trying to gently turn the servo horn by hand causes the robot to "fight you" and try to maintain position. Don't turn too hard or you could damage the servo. The servo will start to hum as it struggles against you to maintain position. If the servos just turn by hand with no problem when the robot is powered on in STOP mode, that means they are not being commanded to go anywhere. The most likely cause there would be lack of signal going from the nano to the servo controller. The wires may not be connected properly, the wires may be loose, the servo controller is not getting power from the BEC, etc.

Let me know about the two questions above.

[Jim Neill]

1) It seems that they are definitely getting power, and I can hear the servos humming. The piezo beeps a low tone and then few seconds later a higher tone.

2) Some of the servos fight back to maintain position, but not all.  Some can't be moved at all and some can be moved fairly easy, I've checked all the connections and there doesn't seem to be anything loose or incorrectly connected.

[Stan Hurd]

OK, so what you are telling me is that, after reflashing the Nano, I should have gone back and redone the step of setting the servo arm positions.  I will do this following your instructions and let you know what happens.

On Thursday, November 16, 2017 at 10:33:18 AM UTC-8, Stan Hurd wrote:

[Steve Pendergrast]

Hi Jim,

It's fairly easy to plug the servos in backwards.  You probably already checked for this, but the easiest way is to just look and see if the darker servo wires are all on the same side of the servo controller ports (and of course are matching the BLACK pins on the servo ports). If a few are plugged in backwards they'll stick out like a sore thumb. Having some plugged in backwards, or missing one of the pins when plugging in (which I've done myself from time to time), is the most likely cause for some servos getting commands and others not.

If that's not true, they're all in correctly, and you are certain the two signal wires are going from the nano to the servo controller SDA and SCL correctly, then my next guess is a bad servo controller.  Now, mind you, I've only seen about one bad servo controller for every hundred I've used, and that one was splattered with solder and obviously defective before I even plugged it in.  So this is a fairly rare issue, which is why I was trying to eliminate other possibilities first.

But from your tests, we know the nano is good (it's beeping the right way on power up), you've checked and rechecked the wiring, but the servos still aren't all getting power in STAND mode.

It might be worth 5 minutes to see if the robot thinks it is in STAND mode (using nano connected to USB and arduino IDE serial monitor). But in most modes there will be some kind of  motion or at least tension on the servos.

[Steve Pendergrast]

Yeah, I'm saying that if the nano was not actually flashed and running the program to begin with, and the flash fixed it, then your original leg positions were wrong because you *thought* the nano was holding the leg positions in the right place when it really wasn't.

[Stan Hurd]

Yeah!  That worked!   So, the secret is to reset the servo arms again after flashing the Nano.

Thanks.

On to the next step.

Stan

On Thursday, November 16, 2017 at 10:33:18 AM UTC-8, Stan Hurd wrote:

[Jim Neill]

Just connected to nano for debugging and see that when the pot is turned it is changing modes, as the serial communication goes from 

Stand, Sensors:
 A3=184 A6=269 A7=305 Dist=1000

to

AdjustMode

to 

SERVO: 0
SERVO: 1
SERVO: 2
SERVO: 3
SERVO: 4
SERVO: 5
SERVO: 6
SERVO: 7
SERVO: 8
SERVO: 9
SERVO: 10
SERVO: 11

to

 Rand

to

LOS###

I believe the servo motor is getting power as the power led indicator on the board is lighting up. I've triple checked the wiring between all the components.

On Sunday, November 26, 2017 at 2:00:26 PM UTC-5, Steve Pendergrast wrote:

[Jim Neill]

Just to verify, the servos are connected yellow from servo to yellow on controller, red to red, and brown from servo to black on controller. I do notice that when I power on, at the second higher pitched beep the power indicator on the servo controller dims and the servos start buzzing.  Could it be a bad servo causing the problem?  Going to try unplugging all servos but one and see if it will move at all.  If it does, then I'll move it to the next position on the controller.  If it does not I'll try a different servo and work my way from there.

[Steve Pendergrast]

Yes, "brown" on the servo wires goes to "black" on the servo controller, red in the middle, yellow is yellow.

It is possible for a servo to be bad in such a way that it sucks down enormous amounts of power, and this could definitely cause the servo controller to act badly.  We do test every servo, however it's possible that every now and then a servo is bad in such a way that it passes our tests then later gives up the ghost because it was borderline. 

It's also possible for the BEC to be bad and unable to provide enough power, although I find this to be fairly rare with the brand of BEC we use (less than 1% certainly).

Unplugging the servos one by one is tedious but would identify the culprit within five to ten minutes.  I suggest: unplug port 0, power back on and see if the other 11 act normally. If not, power down, unplug port 1, plug port 0 back in, wash, rinse, and repeat. Make sure you power down while plugging and unplugging wires.

[Jim Neill]

So I removed all the servos from the controller but one and attempted to power on and the led on the controller doesn't come on anymore.  I tested the voltage coming out of the BEC and it's a constant 5.21V.  Not sure what it is supposed to supply, 

[Steve Pendergrast]

Ok, that's enough to convince me you need a new servo controller. I will send one tomorrow. Please confirm your identity by messaging me on kickstarter, the message just needs to say "replace servo controller" or something like that.

The voltage of 5.2 is fine on the BEC by the way.

[Liam McMullan]

Good Morning Steve

Thanks for all the quick responses.

I have flashed my spare Nano with the latest code ( strangely the previous Nano was causing my iMac to reboot every time I plugged it in).

Now in Serial debug mode I can see the pot is indeed working as I get the following when travelling through the positions:

Stand, Sensors:

 A3=136 A6=218 A7=272 Dist=1000

AdjustMode

SERVO: 0

SERVO: 1

SERVO: 2

SERVO: 3

SERVO: 4

SERVO: 5

SERVO: 6

SERVO: 7

SERVO: 8

SERVO: 9

SERVO: 10

SERVO: 11

Rand

On power up I hear a beep followed by a second beep a few seconds later and the motors on all but 1 give a little twitch. Possibly as you told Stan, perhaps my servos are all out of whack. I will perform the steps you outlined to Stan and report back

Thanks

-Liam

[Liam McMullan]

Good Morning Steve

Just wanted to say all working now ! Really happy.

Its an impressive little beastie. Off now to build the Gamepad.

Starting to 3D print parts for another !!

-Liam

On Thursday, 16 November 2017 18:33:18 UTC, Stan Hurd wrote:

[Steve Pendergrast]

Great news Liam!

So it seems like, when the nano has lost its program for some reason, this causes confusion and may lead to legs getting installed with horns all out of wack.  Apparently at least two of the three builders in this thread had this issue. (The third might be a bad servo controller, a new one was sent out).

I am going to update the build instructions with a simple procedure for testing to see if the nano has the robot program running.  Basically, you hit the RESET button on the nano, and there is a certain sequence of flashing lights you should see if it's running the robot program.

[Jim Neill]

So I just got around to replacing the servo controller and I think there is something else not working as expected.  What I did was plug the BEC into the replacement controller's port 12, the piezo into port 13, a servo on port 0, and the 2 wires from the nano and powered on from the switch.  The controller did not power on but I did get the low tone beep from the piezo and the Nano's leds lit up as expected. While troubleshooting I removed the wires on port 12 and plugged them back in to insure the connections were good on the servo controller, turned the switch back on and nothing.  The next steps I know I'm not supposed to do, but I'm a risk taker and while the switch was still on, carefully removed the wires from port 12 and plugged them back in and the servo controller powered on and the one servo attached twitched and I got the higher tone beep. I am able to reproduce this behavior over and over.  Taking into account that additional information, what do you think the cause of this strange servo controller behavior? 

[Jim Neill]

Additional information while troubleshooting. If I power on the switch and the servo controller doesn't get power, if I reset the NANO with the button on the NANO the servo controller powers on, and the demo mode moves the one servo I have attached as expected.

[Steve Pendergrast]

I have to say that's very odd.  

I just want to confirm that you also are giving power to the nano using the red and black single dupont connectors coming off the switch/battery assembly?   The rest of this post will assume you did that.

So you're saying that if you power up the nano and the servo controller at the same time, it doesn't work. But if you reset the nano, it starts working. And if you restart the servo controller (by cutting its power then reapplying power) then it also starts working.

I think at this point, I'm just going to send you an entire new electrical system.  Battery/switch assembl and nano.  Because I just don't know what's going on. It's something very odd. Maybe a bad nano that has bad I2C ports, maybe the BEC is marginal and powering both on at once causes it to glitch. I'm grasping at straws!

But you've had way too much trouble so let's just replace everything that could be bad.

[Jim Neill]

Thanks Steve, I'm more than willing to do this a piece at a time, and I think I've narrowed it down to the BEC/Switch/Battery harness.  I flashed another NANO I had with your code and it's doing a similar thing, except the reset button isn't waking the servo controller up this time.  However, it is showing the same behavior with unplugging the red and black wires and plugging them back in on the servo controller port 12 with the switch on.  That is 100% repeatable, so it would make sense that the BEC isn't supplying enough voltage initially when both devices are powering up at the same time. Once both devices are powered up, if I supply power to the NANO via USB connection I can power the switch off and the LED on the servo controller dims and turning the switch back on it gets brighter again. Pressing reset button on NANO in this state also gives me the low tone on bootup and high tone on "servos ready" and demo mode runs like a champ.

Virus-free. www.avast.com

--
You received this message because you are subscribed to a topic in the Google Groups "Vorpal Robotics Forum" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/vorpal-robotics-forum/ApuVYL6sD3g/unsubscribe.
To unsubscribe from this group and all its topics, send an email to vorpal-robotics-forum+unsub...@googlegroups.com.
To post to this group, send email to vorpal-robotics-forum@googlegroups.com.
Visit this group at https://groups.google.com/group/vorpal-robotics-forum.
To view this discussion on the web visit https://groups.google.com/d/msgid/vorpal-robotics-forum/527f18ea-0ad5-4325-b211-c7e8637583c3%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[Steve Pendergrast]

OK, I sent the new battery/switch/bec assembly but not a new nano.  Hopefully this gets you up and running.

[Jim Neill]

Steve, just wanted to let you know the new assembly appears to have resolved my issue.  I haven't wired everything back up, but just with the piezo, nano, and one stepper connected (same connections I had before when testing) everything powers on normally and the stepper moves correctly in demo mode, etc.  Thanks for the help, I look forward to finishing this project up now.

[Steve Pendergrast]

All right! Glad to hear it. It must have been something weird wrong with the BEC I guess.