Puzzling results from UCB test

UCB v1.2.1
Windows XP laptop with usb-serial connector

Story so far...
PowerComms v1.2.1 board went fine including all tests.

UCB first stage soldering went fine for all 5 boards, except for a bent leg on a PIC. First two tests go as expected, no problems.

Third test-Exercise Command

With all connections apparently fine, when I start up the RepRap Host Software-> Tools-> Stepper Excerciser the software locks up into a loop. The DOS window has this repeating message:
Recieve error, resending: Timeout receiving byte
0->8: 0

I have tried programming a couple of PICs with stepmotor.hex and get the same result. I have tried multiple UCBs with the same result.
I have scanned through the forums looking for answers. Under a thread called Testing a UCB, I found KDuncan having similar issues. In that thread emf supplied 3 hex files to help with the testing; led_throb.hex, tx_baud.hex, and rx_test.hex.
I have programmed each of these on to the same PIC as I have tried previously and have had success with all 3 programs doing exactly what they are supposed to do.
I have also done conductivity test from the UCB PIC socket to T and R on the PowerComms board and am convinced that I have not switched the cables.

About the only thing I can think of is that it might be possible that the stepmotor.hex file was corrupted in the zip download and I might try that while I wait for someone brilliant to help me figure out what is going wrong. I did find a mention in one thread about shorting out P3, but I didn't find any other information or instructions about that and so am not sure what specifically I should do. I appreciate any help.

It looks like you've fallen into another gap in the testing procedure. So far, I think all of your tests have been using some other terminal software. This is the first test that uses the reprap host software and the rxtx serial library to talk to the boards. There's probably some problem with the host software configuration. Maybe there's something wrong with the way the rxtx library is installed, or perhaps you've specified the wrong serial port in the preferences.

Let's see if we can find out if the java software is seeing the serial traffic at all. Set yourself up for the powercomms-only loopback test. Open up hyperterminal (or whatever) and make sure you can see characters echoing back as you type. Now close hyperterminal and load up the reprap host software. Go to the preferences dialog and change the setting "CommsDebug" from "false" to "true". Now open up the stepper exerciser. If the host software is able talk to the serial port, you'll see a few messages scoll by on the console. I forget what they look like, but you'll see the echoed packets coming back in a line that has "rx" and then a few hex bytes printed out. Then the dialog box should pop up with all of the motor scollers disabled. If the software can't talk to the serial port, you'll probably see about the same thing as you're seeing now: no dialog box.

If the above test does bring up the stepper exerciser windows, then we can remove the loopback link, hook up your UCB, and run the same test again.

I got another loop.
Following your instructions just testing the PowerComms board.
The Hyperterminal echo worked fine, but the RepRap software is giving me in the DOS window the following:
RX ffffffff Receive error, resending: Timeout receiving byte
0->8: 0
TX 0->8: 0

I did make one other change. Thinking that the issue lies in the RXTX library, I followed these instructions [rxtx.qbang.org] to add the 3 files to the Java Development Kit folder as I had previously only had them in the Runtime Environment folder.

I'm running out of ideas. Hopefully someone with a windows system will chime in.

On unix, I never get the "RX ffffffff " part of the output. If nothing is connected, I get:

[java] comms: tx 0->8: 0 [0.304s/304ms]
[java] Receive error, re-sending: Timeout receiving byte

looping forever. If the loopback link is in place, I get:

[java] comms: tx 0->8: 0 [0.000s/-1193702841775ms]
[java] comms: rx: 54 51 31 8 0 0 ba [0.023s/23ms]

followed by similar lines for addresses 2, 3, and 4.

I'm not sure what to make of the ffffffff in your output. I think it's safe to say that your problem is a host software problem and not a hardware problem, but past that I'm not sure where to look.

Question:

If we assume that the hardware is correct at this point, is there any reason why I should not proceed with the next stage of soldering and we can maybe sort out the software issue later?

Once again, I really appreciate your help.

I'd say go for it. Adding the extra components, assuming you don't do anything really bone-headed, won't prevent you from debugging the problem later. If you want to check, you can load up the debugging firmwares after you add the rest of the components and see that they still work like they did before. If they work, you probably did everything right. If they don't, you broke something =)

I would suggest running reprap only on linux not windows, you can run linux from a flash drive or CD on your windows laptop no need to install linux at all on your hard drive.

Kevin Wrote:
-------------------------------------------------------
> I would suggest running reprap only on linux not
> windows, you can run linux from a flash drive or
> CD on your windows laptop no need to install linux
> at all on your hard drive.

On that note, I've been thinking of producing a LiveCD, probably based on Ubuntu, that contained all the software you'd need. Is the host software anywhere stable enough to begin making plans to do this?

And probably even more importantly, has anybody got a spare 6-700MB on a web server for me to upload the .iso to?

Would it be posable to offer the live CD in the reprap store to bypass the need for a web server/storeage?

Kevin Wrote:
-------------------------------------------------------
> Would it be posable to offer the live CD in the
> reprap store to bypass the need for a web
> server/storeage?

Either way, if the software's stable enough I'll get on to it in the next couple of weeks and hopefully have something by the end of November. If the host software is close to a 1.0 release (or even a 0.9) I'll have a play so I know what I'm doing but not actually release the CD until the 0.9/1.0 host software release.

Zach, any problem if I mailed you x copies of the CD (or point me to a place to temporarily upload it for you) and you make additional copies as requests come in through the store?

Its a bit makeshift but it would be a good start. Ultimately its a waste of Zachs talents and the best option would be to either have a downloadable ISO or a distribution hierarchy. i.e the order is placed through the store and the nominated distributor physically close to the buyer writes the CD and posts it to them.

Of course, this is also the way I envision the whole Reprap parts ordering process; the website is the central ordering point but emails are sent out to other reprappers to print the parts and mail them to a local address.

Edited 1 time(s). Last edit at 10/30/2007 08:13PM by reece.arnott.

I'd been thinking a vmware image or bootable CD would be a good thing to have. Even a minimal one without devel tools would be good for debugging software problems like this.

You might be able to get the images hosted at sourceforge. I don't know anything about their file size policies, but I've seen projects that had gigabyte-sized downloads.

I'm not quite there yet, next few weeks also, but I will have a web site supplying to UK and European RepRapers (moulded parts etc) and I will have space for large file downloads. Reece I will PM you nearer the time if this is of help.

Ian

I found I had to install R10 and R11 before test 3 would pass.

Hi,

I've just began working on a Reprap machine as a college project in Ireland. The Reprap has been fully assembled but isnt working properly.

The reprap is interfacing with a computer using Ubuntu. Im currently experiencing a similar problem to Glow Guy, receiving an error message as follows.

"ffffffff receive error, re-sending: Timeout receiving byte
0->3: 4
Tx 0->3: 4
Rx 54 51 31 3 0 4 la
Device at 3 not present
Ignored IO exception in position update: java.io.IOException: Device at
address 3 not present"

I am not sure how to use the poke utility. It is my first time working with linux and the computer connected to the reprap has no internet connection. I downloaded the poke.cc file to my laptop and transfered it to the reprap computer but i am not sure how to use it.

I did run the test mentioned by emf whereby I opened up the stepper exerciser with only the looped powercomm. Rx and tx messages did appear in the console and the x y and z scrollers were disabled.

Usually within the stepper exerciser I can acheive full motion in the Y axis, motion in the Z axis at lower speeds (need to replace belt) but in the x axis the motor just appears to vibrate in the same position. The x axis moves fairly smoothly by hand with the motors off.

The stepper exerciser always crashes showing an error message similar to above but with any of the devices x y or z shown as not been found. I then have to turn the reprap and program off and on before i can move the exruder again.

I have the rxtx program installed.

Sorry if my problem is vague. Just in case Glow guy has found a solution and to show that the problem is not limited to windows.

Thanks

Shane

Sorry I haven't posted back to this thread for a while. So many other projects going on besides this one (researching getting a Piaggio MP3 scooter and developing a new wind turbine design, which is to the point now of getting serious financial backing).

I figured out my problem to be with the USB-Serial connector that I was using to try and run the RepRap software from my laptop. When I moved things to run off a desktop with a serial port all of my problems went away (no problems running from a Windows box). I now have full motion on the 3 motors from the software and need to place my order for the plastic bits and the extruder stuff.

I am afraid that I am of no help to you Shane. Only thing that I could think of given your description is to make sure that you wired the x motor backwards when compared to the y & z motors.

Best,
Glow Guy

I'm actually runnning off a usb to serial cable. The computer its working with at the moment has a serial connection though so i'll connect it up with that.

Thanks,

Shane

Shane wrote:

> I've just began working on a Reprap machine as a
> college project in Ireland. The Reprap has been
> fully assembled but isnt working properly.

Did each board (UC test OK when wired up individually,
as the only one in the ring? A Reprap is a slightly complex
thing to build and then test as a complete entity! Better to
test each sub-assemby as you go.

> Device at 3 not present

This is the Y axis controller board apparently not being found.

> I am not sure how to use the poke utility. It is
> my first time working with linux and the computer
> connected to the reprap has no internet
> connection. I downloaded the poke.cc file to my
> laptop and transfered it to the reprap computer
> but i am not sure how to use it.

I suggest that you grab the firmware from subversion,
then install the development tools per the instructions
in the firmware/INSTALL file, and then build
the firmware. This will build poke as well, automatically :-)

Roughly, you will want to do:

SVNURL="https://reprap.svn.sourceforge.net/svnroot/reprap/trunk/reprap"
svn co $SVNURL/firmware firmware
cd firmware/
less INSTALL # Hit q to exit less when done reading it
# Do what INSTALL suggests to get sdcc and gpasm installed, then...
./configure && make
cd tools
# Now you should find a poke executable here in firmware/tools
ls -lF poke # Is is here?
file poke # What kind of file is it?
./poke -h # What does its help output say?

Note: firmware/tools/poke.cc is source code for a C++ program,
the executable file is just firmware/tools/poke

OK. Once you have poke... what good is it to you?

To run it as a basic "is my PIC out there running Reprap firmware"?
test, in the firmware/tools directory do:

echo 0 | ./poke -d 2 -t /dev/ttyS0 -v

Explanation: the 0 after echo is the SNAP command you are sending.

The 2 after the -d is the address of the board (2 is X axis, 3 is Y axis,
4 is Z axis, 8 is Extruder0, 9 is extruder1, unless you built
non-standard firmware).

The /dev/ttyS0 is the name of your serial port ( /dev/ttyS0 is the
equivalent of the MSDOS port called COM1 -- adjust as needed!).

The -v just makes poke slightly more verbose.

I have written a little shell script wrapper called "detect-boards.sh"
which runs poke like this for each standard SNAP address and
reports what it finds in a more "friendly" way than poke does.
I'll commit it tonight and post it here.

> Usually within the stepper exerciser I can acheive
> full motion in the Y axis, motion in the Z axis at
> lower speeds (need to replace belt) but in the x
> axis the motor just appears to vibrate in the same
> position. The x axis moves fairly smoothly by hand
> with the motors off.

That sounds like a stepper wiring issue more than a
comms problem between the PC and the electronics.
I think there is discussion of solving that by reversing
two wires, either on the wiki or here in the forums...?

> The stepper exerciser always crashes showing an
> error message similar to above but with any of the
> devices x y or z shown as not been found.

The newer firmware (built in the process outlined above)
has better comms processing (thanks Andreas!), so
quite a few errors that showed up before no longer
occur. If the board that the error appears for is
"random", your hardware may all be fine, just
update to the newer firmware (needs 16F648A PICs)
and see what it does for you.

If you get the *same* board disappearing, then
you may have a wiring or soldering issue with
that one board (address 3, Y axis controller, from
the posted error traceback).

It's not clear that you and Glowguy have the same
issue... but hopefully the above will help you a bit.

Jonathan (who doesn't have a complete Reprap!)

---

---

Jonathan