Forum

Welcome Guest 

Show/Hide Header

Welcome Guest, posting in this forum requires registration.





Pages: [1]
Author Topic: Trouble Troubleshooting
jnvdub
Newbie
Posts: 9
Permalink
Post Trouble Troubleshooting
on: May 18, 2013, 11:15
Quote

Hi Sandy,

First off, thanks for all your work on this really cool project. I'm new to Arduino and coding, but know basic electronics and hardware. I've been having trouble getting mine working.The machine will move up and down fine, but left and right is backwards (see wiring photo). I tried a number of configurations, but still couldn't get it working. I thought to try if it will still do to at least do some width tests, which it did only one successfully. Usually it stops, stutters or just goes off completely (see photo). I've tried both x64 and x32 versions on respective computers (PCs, Windows 7), however I was only able to get the x32 version to work on both machines. I've also tried uploading the non-compiled version with the Arduino IDE as well as the compiled with XLoader, but no difference. Any thoughts?

Thanks,

Joe

Image
Image
Image

sandy
Administrator
Posts: 1317
Permalink
sandy
Post Re: Trouble Troubleshooting
on: May 18, 2013, 11:55
Quote

WElcome to the forum! First off, that's a handsome easel. You don't mention, but I assume you're using an arduino uno with an adafruit motorshield, and loading the firmware called polargraph_server_a1. And you're using the recent version from the zipped bundle.

I think you should be able to get your back-to-frontedness sorted out by just flipping the cables. That might make the direction now in reverse, but just reverse the sequence of the wires in the screw terminals. There aren't any directional problems you can't solve by trial and error to be honest.

How are you powering it all? Voltage, wiring etc.

cheers!
sn

jnvdub
Newbie
Posts: 9
Permalink
Post Re: Trouble Troubleshooting
on: May 18, 2013, 15:49
Quote

Hi Sandy,

Thanks for the quick response! Ok, so fixed the wiring, but still having some issues.

Yes, I'm running it on an Uno r3 with adafruit motorshield, using polargraph_server_a1 and the latest uploaded bundle. It's powered by 12v 1200mh power supply with 12V, 0.4A, torque 53n.cm stepper motors. I had some gears and belts made locally (I'm in China) and just using some bearings on wood for the gondola. Once it'll work the next step is to get a solenoid on the gondola and clean it up. Here are the photos of the machine at the moment.
http://imgur.com/a/DRmG7

Still a little stumped on the issue though. I can control it for a bit using move pen to point, but then it gets snagged. It happens if doing the pen test or move to point after a couple minutes.

Thanks,
Joe

sandy
Administrator
Posts: 1317
Permalink
sandy
Post Re: Trouble Troubleshooting
on: May 18, 2013, 16:23
Quote

How is the power wired up though? You don't have the power jumper on the motorshield in the pic, so is the psu wired directly into the motorshield, or into the arduino with the jumper in place? It sounds like a power issue which is why I'm asking.

Couldn't guess on what snags it - are the motors or the drivers getting hot?

sn

jnvdub
Newbie
Posts: 9
Permalink
Post Re: Trouble Troubleshooting
on: May 19, 2013, 04:45
Quote

I tried with the power supply directly into the motorshield without the jumper, but it does get quite hot. I then tried with the jumper, but same issue. I have ready another bigger motor driver (3A with heatsink), but I'm not sure how to connect it to the arduino (see updated photos). I also tried another power supply (800mh) plugged into the Arduino with the jumper, but same problem.

Joe

kongorilla
Pro
Posts: 362
Permalink
kongorilla
Post Re: Trouble Troubleshooting
on: May 19, 2013, 05:27
Quote

Hey jnvdub, do you mean the L293D motor drivers are getting hot? You can try running at a lower voltage, or "piggybacking" the drivers (or both). Piggybacking is described on this page:
http://www.ladyada.net/make/mshield/use.html

On my old uno/motorshield polargraph setup I used both tactics to keep things running cool, but I use low voltage steppers that suck current like crazy if you let them.

The polarshield has nice drivers with pots that allow you to adjust current to suit your motors, which is very handy.

jnvdub
Newbie
Posts: 9
Permalink
Post Re: Trouble Troubleshooting
on: May 19, 2013, 11:18
Quote

hey kongorilla,
Thanks for the help. Yes, I did mean the L293D drivers are getting hot. I bought an adjustable power supply (1000mA) and tried it at 9v, where it completes the pen test and doesn't get as hot, however now only one motor will work during the test. Not doing the test both motors work fine. I went back to 12v and same issue. I have another motor shield, would it work better?
http://tinyurl.com/d5lvpcc
A bit confused on how to connect it to the Uno though..

Thanks,
Joe

sandy
Administrator
Posts: 1317
Permalink
sandy
Post Re: Trouble Troubleshooting
on: May 19, 2013, 12:33
Quote

I'm sure your other drivers will be usable, but they'll need a bit of fixing in the code, and of course the one you linked to is only one axis - I guess you have a pair of them.

Did you have any success running the standard adafruit motorshield code examples?

I haven't ever piggy-backed the L293Ds, but then I only ever run at 6v anyway, so I can get away with a little heatsink. Whether you are putting the power in through the arduino or through the external motorshield terminal, as long as the jumper is on, then the L293Ds should be getting the full 12v. If the jumper is off, then the 12v must enter throught the motorshield power terminal.

I'm not sure what you mean when you say the pen test completes, but with only one motor running? Does it just draw a straight line then? Is the other motor loose, or locked?

When you lock the motors (with a "set home", for example), are they solid? Can they be moved by hand? Also, when moving (assuming they can do it), can the movement be stopped or diverted by holding onto the gears?

Are you sure the wiring is correct? I mean that red-green and blue-yellow correspond to the wiring of the two motor coils? It seems to me like a power issue, but I'm not sure why it would work sometimes and not others if it is.

sn

jnvdub
Newbie
Posts: 9
Permalink
Post Re: Trouble Troubleshooting
on: May 19, 2013, 14:47
Quote

Thanks for the response.

Wiring as is, with the power connected directly into the motorshield and the power jumper off, it runs the adafruit examples on both motors. I changed the example code to test both motor ports.

For the pen test yes exactly, it just draws a straight line. The problem motor (left) is locked. I mean complete as in the sequence finishes and it's ready to receive another command.

When I "set home" they are locked and can't be moved without some force. When moving they can be stopped by holding onto the gears. Then both motors work with move pen to point, but after about the 8th movement the left motor stutters, stops and the arduino isn't able to accept commands.

I think the L293D is getting hot and switching off. Only the outer L293Ds get hot, the middle does not, does that matter? The 1200mA power supply at 12v doesn't get hot at all, but the other power supplies, 1000mA and 800mA at 9v get very hot after a few minutes of use. Here is the link to the motors I'm using in case I'm missing something.
http://item.taobao.com/item.htm?spm=a230r.1.10.203.WWxfhp&id=16048743391

Joe

sandy
Administrator
Posts: 1317
Permalink
sandy
Post Re: Trouble Troubleshooting
on: May 19, 2013, 15:20
Quote

L293Ds do have a thermal shutoff, so you might be right! But I don't really know how they behave when actually driven to it, so I don't know if that's what to expect.

You say the left motor is particularly troublesome - does the right motor work ok otherwise? If you swap the motors around, does the problem migrate to the right side? I mean, is it the motor or the driver (or the wiring). IF you disconnect the motors entirely (or just one of them) does the whole thing still crash? (It should still run ok without motors.)

The two L293Ds are the chips on the outside, the one in the middle is a different thing, so it's right that it doesn't get hot.

I can't see a link to the datasheet for the motor on that page. How did you arrive at the wiring sequence? Though I guess if the adafruit sketches work then the wiring must be ok. It's a puzzler!

sn

jnvdub
Newbie
Posts: 9
Permalink
Post Re: Trouble Troubleshooting
on: May 20, 2013, 08:24
Quote

If I swap the motors, it's still the same issue, they both work for a while then it cuts out. I arrived at the wiring sequence from a sheet the supplier provided showing which colors were for which coils. I also tested the shield on a Mega2560 with the coordinating code just in case the problem was with the Uno, but same issue as before.

So I've bought more L293Ds to piggyback like kongorilla suggested, also bought some SN754410s to replace the L293Ds in the motorshield since I've read they do allow higher voltages, but it may or may not work. I've also bought another motorshield in case this one has been damaged at all by the testing. I should get these in the next few days, so we'll see if one of these solutions work.
Thanks!

Joe

sandy
Administrator
Posts: 1317
Permalink
sandy
Post Re: Trouble Troubleshooting
on: May 20, 2013, 09:52
Quote

Cool, sorry I haven't been able to be more help, but keep us updated if you reach any conclusions!

good luck
sn

jnvdub
Newbie
Posts: 9
Permalink
Post Re: Trouble Troubleshooting
on: June 1, 2013, 06:08
Quote

Hi Sandy,

So I got some new motorshields along with L293Ds for piggybacking and SN754410s to replace the L293Ds. I tried piggybacking, but it didn't change anything. Once I changed motorshields, substituted the SN754410s and changed the power supply to 12v @800mA the machine would be happy to move about using move pen to point for as long as I wanted. However, once I started a pen width test or a test image, one motor would lock up resulting in another straight line. Any thoughts?

Joe

kongorilla
Pro
Posts: 362
Permalink
kongorilla
Post Re: Trouble Troubleshooting
on: June 1, 2013, 08:53
Quote

OK, if everything's running cool (or pleasantly warm), and the motors seem to have the proper amount of torque at the current they're being fed, we should also eliminate any software/firmware basics we may have overlooked:

After you uploaded the hex or compiled sketch, and edited the default.properties.txt, you also uploaded the machine specs via the controller, right? Your 100% confident about the values you entered?

Have you tried opening the console (by typing 'C') to see if there's useful info being returned when the machine misbehaves?

(Yes, I'm clutching at straws...but you never know).

sandy
Administrator
Posts: 1317
Permalink
sandy
Post Re: Trouble Troubleshooting
on: June 1, 2013, 11:49
Quote

The keyboard shortcuts are now ctrl+key, so ctrl-c. Looking at the output of the machine in the console is a very good idea. Paste it in here or email it to me if it doesn't make sense.

So you might find that there is a symptom of the crash in the software, some error that is reported. Does the software otherwise work fine after a crash - I mean, is it still responsive when it has finished it's current command?

Hum. I did have a good idea earlier, but I've forgotten what it was now.

sn

jnvdub
Newbie
Posts: 9
Permalink
Post Re: Trouble Troubleshooting
on: June 20, 2013, 12:26
Quote

So it's definitely a power supply issue. When the machine gets off track the console keeps going as if the error hasn't registered. I've uploaded the machine specs and the .txt file using the controller and I'm 99% sure the values are correct. I've tried a number of power supplies (maybe around 6-7?) varying from 600mh to 2A and the best so far seems to be the 600mH, but then again it still doesn't fully work. I've also attached some heatsinks to the SN754410s and I think that helps. However still having some issues.. Is there a trick to the motor steps or power supplies that I'm missing? You can check http://imgur.com/a/DRmG7 for some photos.

Joe

sandy
Administrator
Posts: 1317
Permalink
sandy
Post Re: Trouble Troubleshooting
on: June 22, 2013, 00:06
Quote

Hey Joe, you probably are correct - I don't think there are any settings in the software that would prompt the behaviour you are seeing. Bad geometry or shading yes, but not going kaput halfway through a drawing. Unfortunately there is no trick I can think of.

Is there any chance your motors are not the 12v 0.4A motors you think they are? Even both of them on continuously (which they almost never are) should be drawing 0.8A, and not melting your 800mA PSU.

sn

jnvdub
Newbie
Posts: 9
Permalink
Post Re: Trouble Troubleshooting
on: February 26, 2014, 14:07
Quote

Hi Sandy,

So we finally got the power issue solved, it's drawing quite nicely. However there is now a problem of ratio. When we load an image and render the pixels, the image that gets drawn is skewed so that the height is dramatically decreased. Have you had this problem before?

Thanks,
Joe

sandy
Administrator
Posts: 1317
Permalink
sandy
Post Re: Trouble Troubleshooting
on: March 1, 2014, 15:54
Quote

That kind of distortion is unexpected, and I don't think I understand how it could be coming about either. The machine itself has no concept of "height" as an axis separate to width, so it's not like it could be making the drawing shorter without making everything smaller (so narrower as well).

Is it only "render pixel" types of drawing that it happens on, or vector stuff too?

sn

Pages: [1]
Mingle Forum by cartpauj
Version: 1.0.34 ; Page loaded in: 0.063 seconds.