[done] SL 1.3 pauses or disconnects mid-print

So, i built myself a shiny new LM8UU Prusa w/ Sanguinololu 1.3a Electronics, kit from JoeM. I've been using a 1.2 SL on a Sells Mendel with good results.
However, the SL1.3a is giving me a headache:

- It frequently pauses for ~30 seconds, like every ten layers or so. I can see Marlin reporting it's turning its steppers off after ~ 20s, just like it's configured, after another 10s the print resumes just fine by itself, leaving only a minor blob as evidence.

- It not-as-frequently disconnects, on average it does ~100 layers. The error i see on the terminal running pronterface is the same one i get when pulling out the USB cable.

Exception in thread Thread-8667:
Traceback (most recent call last):
  File "/usr/lib/python2.7/threading.py", line 552, in __bootstrap_inner
    self.run()
  File "/usr/lib/python2.7/threading.py", line 505, in run
    self.__target(*self.__args, **self.__kwargs)
  File "/cdrom/rep/kliment-Printrun-6a63009/printcore.py", line 82, in _listen
    line=self.printer.readline()
  File "/usr/lib/python2.7/dist-packages/serial/serialposix.py", line 456, in read
    raise SerialException('device reports readiness to read but returned no data (device disconnected?)')
SerialException: device reports readiness to read but returned no data (device disconnected?)

The print is of course ruined then, both heated bed and extruder stay warm though.
What i already tried:
- Different USB cable
- Different PC
- Reducing Marlin's lookahead
- Wiring 5V from the PSU to the Board instead of relying on USB (which works really fine)
- Switching PSUs

Technical details:
- SL 1.3a
- Kubuntu 11.10
- Kliment's most recent Printrun
- Marlin 1
- BeQuiet 430W 80+b / Scythe 850W 80+s (intend to run a custom-made heated bed)

I'd appreciate any help on this one.

Thanks,
Tom

Edited 2 time(s). Last edit at 01/26/2012 03:17PM by imitation.

Does your usb cable have ferrites on both sides like this one ?

I had some random disconnects in the past, could trace it to electrical interference the day I tried to switch on an hair dryer and bam - print halted.
I first started to plug my reprap power on the same UPS my computer runs on, it reduced the frequency but did not get rid of all the disconnects. Then I switched my usb cable to one like this and since then didn't have anymore problems.

Nope, my cables have no ferrites. However, some Sony cables were happy to donate some large clip-on ones, and i have just put one on each end.
The printer is doing a couple whistle now, i'll see how it goes; so far, each attempt of printing those has failed. Bad karma i guess

You might be spot-on with interference, the oil burner next door surely causes some sort of interference when igniting.

Edited 1 time(s). Last edit at 01/03/2012 10:08AM by imitation.

Oops, it did it again. It plays with my heart. Got lost in the print. Oh Prusa, Prusa...

The whistles did print with two or three pauses, but it still disconnected during the next print.
I'm starting to think this is a software bug in Marlin, i'll try with Sprinter tomorrow.

Couple other things to try. It looks like you are using ATX power supplies. Do you have a good load on 5V? On many supplies you need a 2 amp load or more to get it to settle down and work well. It depends too on the load you have on 12V. If you have a high wattage HBP drawing lots of watts off the 12V line you may need proportionally more load on 5V.

[reprap.org]

Also try other baud rates. I find on my Gen7 setup that if I slow the comms way down it helps with disconnects.

Others have advocated a 12-14 V transient voltage suppressing diode to kill spikes on the 12V line. I bought a couple but haven't installed them yet, but that may be the ultimate solution.

And still others disconnect the USB reset capacitor except when updating firmware. Adrian also swaps out a couple of the resistors to help with comms as shown below.

[reprap.org]

Yes, i am using ATX supplies. Both units are of good quality and give stable 12V, with no load it's more like 12,2V. I'm not using the 5V from the PSU, but from the host through USB. I tried using the 5V from the PSU, too, but it didn't make a difference.
But you have a good point with the reset cap - SL1.3a has a jumper that enables / disables host reset. I had it plugged in for convenience, but i'll try with it removed.
The z-diode on 12V actually makes a lot of sense, too, but i'd go for one with ~25V (or two of yours in series), since the electronics are all rated up to at least 30V. That'll keep it from draining current during regular use, while it will still protect from dangerous spikes or from overvoltage generated when moving the axes without a PSU connected (i've had that happen to the SL1.2, fried all Pololus)

How's your experience with Gen7? I'm thinking about printing one for backup and as a reference platform.

Disabling host reset solved the pauses, but not the disconnects. EDIT: nope. Didn't fix anything.
Up next: Sprinter

Edited 1 time(s). Last edit at 01/16/2012 04:43AM by imitation.

I like Gen7 mostly because it's DIY. I don't think the performance is any different from the other boards out there. If you like sprinter make sure to go with 16Mhz and 644p not standard 644.

I've tried moving the printer to the other end of my house, didn't help.
Sprinter show the same behaviour - so there must be something wrong with the SL.
Unless someone has any more ideas, i'm off to making a gen7.

Thanks again for all the help so far.

- Tom

I have a Sanguinololu 1.3 and had frequent disconnects when I had a fluorescent light (has a magnifier on it) closeby and whether the light was on or off the SL 1.3 didn't like it. I unplugged the light and now don't touch it.There were time when the light was on and I just moved toward the sl1.3 and it would disconnect. I put 3 ferrite cores maybe 4 on the USB cord, which didn't seem to do anything but make Radio Shack a little more money.

I still have disconnects, but I attribute most of them to pulling to many amps on my extruder like when the temp is not high enough.

Whatever you do, don't trash your sl1.3. I would like to try it on my system to see if it disconnects more or less frequently. Perhaps we could work up some sort of deal on it as I am building a second Prusa and plan on using a sl1.3

Just some ideas.

Roy

Another question, what resistors do you have in the r7 and r8 positions? The wiki sanguino calls for 100 k.
"R7 and R8 are now 100k pull-up resistors that are on the stepper-enable lines. This ensures the stepper motors stay disabled and don't move while uploading new firmware, rebooting, etc."

Mine just has bridges in those positions, no resistors. I bought it from a guy in Poland. I am researching the reason for no resistors at those positions.
Mine does have the problem of running the extruder motor after uploading new firmware.


Adrian at prusa notes suggests " I replace R7 and R8 (1k at the moment) with 10 ohm resistors - this makes the comms more reliable." I cannot tell which version of the Sanguinololu he has considering he doesn't mention the 100 k resistors on the 1.3a page.

Roy

never mind, mind is a version 1.2 upgraded to 1.3 and it is supposed to have bridges, not resistors.

So i got myself a cheap chinese Arduino Mega and built a RAMPS for it, which is working flawlessly under the same conditions the SL previously failed.

I'm guessing the FTDI went funky, as the Atmel was apparently running stable.

Since my second SL is acting up, too, i'll try to make a Frankenlolu from those two and see what works.

Thanks again,
- Tom


-Closed-

Don't know if it can help (this subject is an old one), but I was suffering the same problem (frequently disconnection between PC and printer) and I finally solved the problem, after trying a lot of different "solutions" :

- I bought a USB cable with ferrites on both sides -> Problem still happens (but perhaps it helps).
- I plugged the PC and the printer on the same plug with protection against electric hazard -> Problem still happens (but perhaps it helps).
- I slow the baudrate down to 57600 (instead of 115200) -> Problem still happens (set back to 115200).
- I set the same baudrate in Windows USB config -> Problem still happens (but seems "better" like that).
- Finally I unchecked an option in Windows dedicated to energy saving for laptops, enabling the "selective" USB controller power off -> It works ! (be carefull, in my windows / laptop option panel, this option appears at two different places : USB bus controller power properties AND Standby Advanced Power supply config

Seems it works now !

I've found very interesting information about USB (and these disabling power options) here to solve my problem

"One more thing" to my previous post : I check the "Use Ping-Pong Communication (send only after ok)" parameter (set XON/XOFF) in the "Connection" tab of the "Printer setting" DialogBox, to avoid the faintest USB problem (Thank's Pierre from Gi-Nova Grenoble).

I have a Prusa Sanguinololu 1.3 with a ATMega 1284P running Marlin and using a Panelolu2 LCD/SD reader and have experienced resets mid-print (stops and looks like it just turned on - extruder and bed heaters set to zero) while printing from the SD card. It has a dedicated 12V power supply (not ATX) .

Early on, I exchanged the oscillator for a crystal to get it to work properly with my Arduino IDE after trying every possible fuse combination for the bootloader.

Right now the SL's 1284P is configured FF D6 FF or:

Ext Crystal :Frequency 8.0- MHz Start-up time 16K CK + 65ms
Boot reset vector Enabled
Boot Flash section size=512 words Boot start address=$FE00
Preserve EEPROM memory through Chip Erase cycle
Serial program downloading (SPI) enabled
Brown-out detection disabled

I was wondering if the fuses could have any influence on the resets that I am experiencing while printing from the SD. Any other suggestions, given the most of the ones listed in this thread are related to USB communications, would be welcome.

Quote
Jean-Baptiste
"One more thing" to my previous post : I check the "Use Ping-Pong Communication (send only after ok)" parameter (set XON/XOFF) in the "Connection" tab of the "Printer setting" DialogBox, to avoid the faintest USB problem (Thank's Pierre from Gi-Nova Grenoble).

I know this is an old thread, but I just wanted to say THANK YOU! My printer randomly started having this issue, and since I hadn't changed anything in particular, I was worried that something was wrong with either it or my computer (still a possibility).

I've been through a ton of threads and this was the first I saw mention of that fix. Enabled it mid print and the pausing stopped. Groovy!

I've NEVER had that setting enabled, and never had a problem before. I'm often browsing/watching movies/playing games while I print without issue. Very odd that this happened, but great that it was a simple fix!

Thanks again!