Printer just stops mid print?
Colt D. Majkrzak
Hi Guys,
Need some suggestions, I’ve tried repiter which works great with my MM1.5, but with the A1 it seems to stop about mid print. Ive just tried pronterface as well, same end result, its not missing steps, its flat out stopping. See attached screen shot, its stuck on line 22743 and the ESt time just keeps increasing. LCD still appears to be responsive, but my ability to tweak the printer from the PC is toast unless I cycle the app.
Here’s the line it stopped on this time:
G1 F2700.000 E0.00775
USB cable is good, literally pulled it from my MM, plugged to the A1 and everything went. I should note this machine is Windows 8 x64, although everything except the A1, so I’m at a complete loss as to what it might be. I’d think overheat, but typically I’d be a few steps off, not stopping / pausing.
Best regards,
Colt Majkrzak F5CI, F5SE
John D
Glenn Beer
Colt D. Majkrzak
You received this message because you are subscribed to the Google Groups "trinitylabs-talk" group.
To unsubscribe from this group and stop receiving emails from it, send an email to trinitylabs-ta...@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out.
Colt D. Majkrzak
I thought that was going to be the ticket, made it to 83k lines, and then the printer just stopped dead in its tracks. I find it hard to believe it's anything but the printer at this point, literally dropped the A1 in place of a working MM, and the problems started......
The electronics literally just stop accepting commands via the serial connection, so if I attempt to send M105, I get no response, almost got the 50% print of the LCD holder this time :(
Wasting just tons of plastic here trying to get this thing to print a full job :(
From: trinityl...@googlegroups.com [mailto:trinityl...@googlegroups.com] On Behalf Of Colt D. Majkrzak
Sent: Monday, March 18, 2013 1:20 PM
To: trinityl...@googlegroups.com
Subject: RE: Printer just stops mid print?
I'm literally doing that right now, grabbed a APC 1500RS and running the print again to see. The printer should have basically been the only thing on the circuit but I cannot be 100% sure, so it's worth a try. This printer obviously draws more power with the dual PSU's, so maybe?
From: trinityl...@googlegroups.com [mailto:trinityl...@googlegroups.com] On Behalf Of Glenn Beer
Sent: Monday, March 18, 2013 1:10 PM
To: trinityl...@googlegroups.com
Subject: Re: Printer just stops mid print?
Sean Mitchell
- After slicing, edit your g-code and find/replace all EXXX.XX with nothing (so a line that looked like G1 X136.250 Y224.753 E1.006 would then have to look like G1 X136.250 Y224.753 ). This will drive your printer around without actually spitting anything out. You can also turn off your heated bed and nozzle since they won't be necessary (remove all lines with M104, M109, M140, M190 in them). If you are using linux/mac osx, this will remove all E#.######## values: %s/E[0-9\.]{1,}//g
- Enable cold extrusion in your firmware (configuration.h), and remove the temperature codes from your gcode (M104, M109, M140, M190) and then print...but first remove the filament from the extruder (or unplug the extruder). It will drive around like it should but extrude nothing.
- Enable cold extrusion in your firmware. Remove the temperatures from the g-code, and unplug the printer while it is printing. The microcontroller will be powered by USB from your computer and happily think it is printing, when really the signals aren't being converted to actual motion.
Sean Mitchell
Sean Mitchell
Colt D. Majkrzak
I’d almost have to rule out power / ground as a cause because using a power conditioning UPS, that should in theory rule out anything power wise. I do not though have a green earth ground wire anyone on my printer that I’ve seen, although I guess I didn’t look in super detail either.
Only thing I can think at this rate is put a file down onto the SD card and let it print from SD to completely rule out the PC?
Colt D. Majkrzak
Printing via pure SD now, ill report back from the other side hopefully with good news.
From: trinityl...@googlegroups.com [mailto:trinityl...@googlegroups.com]
On Behalf Of Colt D. Majkrzak
Sent: Monday, March 18, 2013 4:33 PM
To: trinityl...@googlegroups.com
Subject: RE: Printer just stops mid print?
I’d almost have to rule out power / ground as a cause because using a power conditioning UPS, that should in theory rule out anything power wise. I do not though have a green earth ground wire anyone on my printer that I’ve seen, although I guess I didn’t look in super detail either.
Only thing I can think at this rate is put a file down onto the SD card and let it print from SD to completely rule out the PC?
Best regards,
From: mitc...@gmail.com [mailto:mitc...@gmail.com] On Behalf Of Sean Mitchell
Sent: Monday, March 18, 2013 3:26 PM
To: Colt D. Majkrzak
Glenn Beer
I was using my Macbookpro to control the printer and mid print, a system update dialog popped up.
Stopped the print dead.
I have also several times seen the display of the LCD become corrupted. The print completed properly in each case.
John Stevenson
--
djoatm...@gmail.com
Colt D. Majkrzak
Well according to the LCD im 60% through the print now, so I don’t want to jinx myself but it looks like it will finish. Of course this is the one time the corners lift so its going to come up looking crappy L
This of course though brings me to question the root cause. Looking at the INF itself its literally using the same exact driver as my Ardino 2560 mega R3, and as far as I understand it, same chipset, etc. My cable is plugged into the USB 2 bus on my board, so its no odd USB 3 to 2 thing going on.
So this has me down to one of the following:
-USB interface on my electronics is just jacked in some way, printing via SD gets around this, but I typically don’t print headless
-Cable is somehow shaking itself loose for a brief second, interrupting. I again say this is unlikely since I can literally take the host software, disconnect/reconnect to the printer and it will start accepting commands again, but until reconnect no commands will respond. It acts like the command buffer is full and not getting processed?
Is it recommend to reflash the board even if I was in the last few printers out the door?
From: trinityl...@googlegroups.com [mailto:trinityl...@googlegroups.com]
On Behalf Of djoatm...@gmail.com
Sent: Monday, March 18, 2013 6:31 PM
To: trinityl...@googlegroups.com
Subject: Re: Printer just stops mid print?
I hope not but it sure makes me wonder. Ive got my unit at the ready to power on stage and I'm not going to bother until EZ tells us whats is going on for fear of damaging my electronics.
--
Steamer
I've had the A1 stopping mid print three times, and then switched over to SD, not a problem since. Something seems to be too sensitive. The lcd screen also often shows garbage, but that doesn't stop the printer. Printing from SD seems the way to go right now.
Emile
Colt D. Majkrzak
Streamer, what’s your host O/S and software your driving with? Trying to nail down a pattern so I can look into this further.
From: trinityl...@googlegroups.com [mailto:trinityl...@googlegroups.com]
On Behalf Of Steamer
Sent: Monday, March 18, 2013 6:53 PM
To: trinityl...@googlegroups.com
Subject: Re: Printer just stops mid print?
--
Steamer
Colt,
Windows 7 64bit and Printrun.
Emile
Colt D. Majkrzak
Seems thus far from what people have replied to me both on and off list, most people are windows 7, 40% repitier host, 60% print run, all basically having the same issue, the work around being LCd print.
So this leads me down to one of a few possibilities here:
1) Turindo electronics have a flaw
2) Marlin flashed onto the electronics have a bug dealing with USB (Checked the GIT repo, didn’t see a bug regarding this, but I didn’t search heavily)
3) How TL is using the electronics is causing an issue?
Jean-François Talbot
To unsubscribe from this group and stop receiving emails from it, send an email to trinitylabs-talk+unsubscribe@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out.
Scott Turner
Ezra Zygmuntowicz
Alastair Seggie
--
Jon Bondy
You may be correct, but I have run my Solidoodles for months using a USB cable and have NEVER, EVER had a glitch. Not once. And I drive my printers (two at once) off of my main desktop computer, which I use for business all day long, with MANY programs open. I have printed at least a dozen 10 hour jobs. No problems (at least with communications!)
My sense is that there is something wrong if the printer cannot print reliably using USB.
Just my experience: your mileage may vary!
Jon
Scott Turner
You received this message because you are subscribed to a topic in the Google Groups "trinitylabs-talk" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/trinitylabs-talk/lcS6RY0d-nk/unsubscribe?hl=en.
To unsubscribe from this group and all its topics, send an email to trinitylabs-ta...@googlegroups.com.
Colt D. Majkrzak
Mine also says Mendel, Ezra is there an updated zip that we should maybe try flashing onto the ramps?
From: trinityl...@googlegroups.com [mailto:trinityl...@googlegroups.com]
On Behalf Of Scott Turner
Sent: Wednesday, March 20, 2013 2:49 PM
To: Alastair Seggie
Cc: trinitylabs-talk
Subject: Re: Printer just stops mid print?
I've done four hour USB prints on my Trinity One. Of course my Trinity One says Mendel in the LCD so I don't appear to have gotten the final firmware.
Sean Mitchell
BusyBotz
Cozmic Ray
Power transient causes USB reset
Electrical surge on hub port. Either a device attached to the port, or the port itself, has drawn more current than allowed, and the hub turned off the port.
Even back EMF from a cheap fluorescent light can cause it.
Your high power or transient devices can be moved to a different power source than your computer. Move your printer power source as far away from your computer.
Not easy ---- line from your main box for computer stuff another line for high power
Usually transient on powered USB hub --- draw of more than 500ma usually causes a reset
Sometimes USB device comes right back up --- other times it gets killed
Transient suppressor could be used but good ones are expensive.
On Monday, March 18, 2013 12:39:43 PM UTC-4, Colt Majkrzak wrote:
Hi Guys,
Need some suggestions, I’ve tried repiter which works great with my MM1.5, but with the A1 it seems to stop about mid print. Ive just tried pronterface as well, same end result, its not missing steps, its flat out stopping. See attached screen shot, its stuck on line 22743 and the ESt time just keeps increasing. LCD still appears to be responsive, but my ability to tweak the printer from the PC is toast unless I cycle the app.
Here’s the line it stopped on this time:
G1 F2700.000 E0.00775
USB cable is good, literally pulled it from my MM, plugged to the A1 and everything went. I should note this machine is Windows 8 x64, although everything except the A1, so I’m at a complete loss as to what it might be. I’d think overheat, but typically I’d be a few steps off, not stopping / pausing.
Best regards,
Glenn Beer
Not for this reason, but I have one between my printer and cpu. ( Between all the dongles for keyboard, mouse, wifi, thumbdrives... I needed more ports. )
Or... And this question is born of some ignorance.
The printer doesn't need power from the CPU. What about a USB cable with the Positive lead clipped, leaving ground and signal lines only? The host never sees draws or spikes on the positive, so it should have no reason to kill the connection.
Cozmic Ray
A good usb hub may help --- which is probably expensive
but
usb is so cruddy up probably unavoidable
A transient on the power supply to the hub may cause problems
or
devices drawing more than 500ma on startup and an unforgiving usb hub
I think this is why the CNC guys don't use usb but use the very robust
printer (LPT) port?
djam
If I have the correct schematic for the Taurino board, then it should not even be connected (power wise) to
the USB connector, as a FET should be off, when there is local power supplied to the board via the DC in
connector.
hmm, though looking at the photos of the machine someone (don't recall who) pulled down, it looks like
there is no local power supplied. If this is the case, I would recommend connecting the Taurino to the local
power supply.
Scott Turner
--
Triffid Hunter
> I again say this is unlikely since I can literally take the host software,
> disconnect/reconnect to the printer and it will start accepting commands
> again, but until reconnect no commands will respond.
from the firmware, or the firmware isn't receiving a newline from the
host. Pronterface checksums all the commands automatically so if
firmware receives a corrupted line it'll simply ask for a resend.
However I have not seen any code in Marlin to handle the case where
the newline is dropped. The hostware will refuse to send any further
data until it has received an ok response, and there is no mechanism
by which the host can send an out-of-band request to ask what's
happening. Marlin does not implement a partial command timeout last
time I checked.
When you disconnect and reconnect your hostware, the firmware finally
gets its newline, sends an 'ok' response and the comms link becomes
unstuck.
fwiw I always alter Marlin to run at 115200 instead of 250k and have
seen this issue maybe once or twice in a whole year. At 250k, the
arduino only has 640 clock cycles between each character, and most
avr8 instructions take 2-3 clock cycles. It also generally takes
several dozen clocks to enter or exit an interrupt, mostly due to
pushing and popping registers from the stack. It only has a one byte
buffer so if an interrupt or some other operation happens to go over
that time even once, we drop bytes. It can take dozens of cycles to do
a single 32-bit add or subtract, other math ops can take more. At
115200 baud, we have a far more comfortable 1388 clocks between bytes.
The avr8 hardware serial block does report when bytes are dropped,
however I do not recall seeing any code in Marlin to handle it.
Please note that the primary limit on transfer speed is NOT the
baudrate, rather it is the latency of the USB bus itself plus the
latency of the host's kernel context-switching the gcode provider.
Even in the '60s, computer scientists understood that waiting for a
response for every packet WILL inevitably lead to poor transfer rates
due to latency, even if the line has high bandwidth. This insight is
instrumental to the current success of the TCP/IP protocol as just one
of many examples.
As a demonstration, measure the rate of downloading gcode to the SD
card using pronterface. It will be significantly less than the 25k/sec
baudrate cap. [it takes 10 "line bits" to transfer a byte- 1 start, 8
data, 1 stop, so "byte rate" = baudrate / 10]
We will be implementing an extension to the gcode protocol with
Smoothie that allows the host-firmware communication to use delayed
ACK, like a simplified version of TCP receive windows. Basically we
tell the host how big our buffer is so it can fill the buffer before
expecting a response, instead of waiting for a response for each line.
There is no technical reason preventing Marlin from implementing this
as well. It will of course depend on hostware support.
This issue can of course also be caused by EMI. A high quality, short
USB2 high speed cable with ferrite beads is strongly recommended. Also
beware fluorescent lights and large motors (washing machines!) which
are notorious for blasting your home with EMI at turn on or off. I
have one soldering iron that causes USB resets in nearby equipment
when turned on or off.. Ironically, for once it is NOT the china
special!
I haven't personally checked the A1 wiring loom but hopefully the
motor cables are kept as far as possible from the electronics and usb
cable.