Printing stops randomly

As of yesterday my printer started stopping randomly during a large print. I have tried this 4 times now, and it is getting frustrating. I am trying to do a print that would run for ~4 hours. On the last 4 attempts, the print stopped after about 1.5 hours, 1 hour, 30 minutes, and 1 hour. I am printing ABS @ 245 with a heated bed @100. I have done this a few times before without issue.

The first 3 times, I printed via USB. Twice I was present when it stopped and I could hear the sound effect in Windows that a device has disconnected. To eliminate a bunch of variables, I tried the last print via the SD card (I was still connected via USB to monitor.), but it still failed after an hour

The two things I've read as a possible cause are line noise and under powered PSU. I suppose noise is possible, but nothing changed that would cause it to start yesterday. The PSU I am using is very overpowered for this (Thermaltake 700W), so I doubt that is the issue.

The USB connection is via a Belkin powered USB hub. I'm wondering if it is going bad. I think I'm going to do another test by printing from SD, but completely disconnecting the USB cable after it starts to print. If that works, I'll try connecting directly to the back of the PC.

Any other suggestions?

EDIT: So much for that plan. Printing from SD with USB completely disconnected, it stopped after about 5 minutes while still drawing the initial outline.

Edited 1 time(s). Last edit at 02/01/2013 02:27PM by daichiasuka.

Unfortunately I am here for the same reason.
daichiasuka try switching USB port or use a different cable. I am now using a much shorter cable, so far no disconnects.

Edited 1 time(s). Last edit at 02/01/2013 06:09PM by demetris.

I think I may have fixed my issue. I'm trying another test print right now and it is ~2.5 hours in printing from SD with no USB connection.

I'm embarrassed to say, but I think the root of my issues was that the common from the PSU had worked itself loose from the terminal and was randomly breaking the connection during the print. I squared off the end and reconnected it and the 12V line. I have a few things to test print tonight. I will report back here with the results.

demetris,
Remove the auto reset jumper and print from SD. That way it doesn't matter if the USB disconnects.

---

[www.hydraraptor.blogspot.com]

I had a similar problem, and found the same solution - loose PSU connection!

Glad to see I'm not alone.

Unless this happens again, I'm considering my issue resolved. It's been printing for the better part of 7 hours now without any problems.

demetris - I'm not sure what you are doing differently that you have so many issues with USB connectivity. I haven't had any communication problems before this and the root cause of my issue was a loose power connection. I'd take nophead's advice if your having that many problems and print from SD.

Your jumpers are set to use the 5V from the board, right? Not the power from the USB line.

If all your jumpers are set correctly, try printing from a different PC. Maybe the USB controller on your current one is way too picky/old/over-utilized. Maybe try an external powered USB hub as well.

I always print from SD as I don't get good results printing over USB. I get random blobs appearing.

Using USB between two grounded devices is fundamentally flawed. It creates a ground loop which picks up radiated noise and any noise on the ground of the machine or the PC's PSU is also seen by the USB.

USB is a differential signal, so should be immune to ground noise but I think it isn't because in one of the signalling states both lines go the same way, crazy if true!

To minimise the ground loop always plug the PC and the machine into the same wall outlet and keep all the cables as short as possible and as parallel as possible. Loop pick up is proportional to the area of the loop.

---

[www.hydraraptor.blogspot.com]

Thanks everyone for your suggestions.
I switched to a shorter cable, and I screw the connectors again. Mine was a bit loose as well. So far today, no disconnections!

@nophead Interesting tip for the SD. I am still calibrating though! When I print something serious I will send it to SD for sure then.

Interesting.

I had been looking into using a Raspberry Pi as a simple front end to the Melzi. I could drop stl files on a network share and the Pi could just pick them up, slice, and go.

I wonder if the signal issues could be avoided if I powered the Pi from the 5V rail from the PSU.

Yes that should work fine. Ethernet connections are isolated by a transformer so don't suffer from ground noise. Ethernet was designed about a generation earlier than USB so is properly engineered.

---

[www.hydraraptor.blogspot.com]

I put some of those magnet cores around my USB and it seemed to help. I also stopped turning the lights off (during a print) that the share the same circuit as the printer. Could be a coincidence but a good print is a good print.

I'm also planning on putting a Raspberry Pi on my Mendel90, including a little VGA monitor (using printed brackets that will connect to the side of the gantry, maybe). The plan is to have an all-in-one CAD and printing station (the MendelStation?).

I'll tell you guys what I end up getting to work.

-jeff

I would love to see some pictures of your mendelstation completed.

jra Wrote:
-------------------------------------------------------
> I'm also planning on putting a Raspberry Pi on my
> Mendel90, including a little VGA monitor (using
> printed brackets that will connect to the side of
> the gantry, maybe). The plan is to have an
> all-in-one CAD and printing station (the
> MendelStation?).
>
> I'll tell you guys what I end up getting to work.
>
> -jeff

Take a look at OctoPrint (https://github.com/foosel/OctoPrint ). You can use a RPi as a print server and control it using a browser with any device on your home network. It even supports a webcam. It's still a work in progress, but so far it's been really useful to check on my Mendel90 while i'm somewhere else in the house.