I had updated my RH to the latest about a month or so ago, and suddenly began having communication time-out issues.
Uninstalled that version, reverted back to the version I had been using previously and no more problems.
Currently use version 1.0.6 with Slic3r version 1.2.9 no issues what so ever.

Oh and as an additional clue, if I kill the job after it does this, it goes back to home - and runs through the endstop. So whatever the micro is doing, it's losing track of where the carriage is.

I guess this means no-one has seen this kind of behaviour?

I guess I misunderstood what you meant here.

The endstops are "soft" (in that the microprocessor needs to be paying attention to them for them to work). I understand that apart from homing, they're largely ignored - as anyone who has had a head crash and resulting belt slip would find.

I'm getting pretty close to concluding that the motherboard has a problem of some sort. I'm trying to decide if it's worth reflashing or if it's better to wait for Solidoodle support to get back to my email that hasn't been replied to yet.

I guess this means no-one has seen this kind of behaviour?

Unboxed my recently arrived Solidoodle Workbench, encountered the issues that others have been having (packaging not quite up to standard, broken glass build plate, Y axis way out of alignment, etc.) ... having spent a lot of time modifying my first-gen Solidoodle 3, I was able to solve all of these problems fairly easily and get the rig printing. I downloaded the version of Repetier Host Solidoodle that was available as of last week (V0.85b) and used all default settings (except for selecting Solidoodle Workbench where necessary).

I then ran into an odd issue. A few minutes into the print (using either extruder), the printer would stop all motors and just run the extruder for over 10 seconds, so I get a puddle of plastic under the head. It would then retract for about the same length of time, then continue on with the print. Of course, it doesn't work - the retract pulls the filament all the way out of the hot end so when it continues nothing comes out (and because of the puddle of plastic, the next time the head comes around it just knocks the incomplete print off).

To debug, I got the log to show all commands and the printer to use Ping-Pong Communications. During the time when the printer pauses and just extrudes, no commands get acknowledged (the log stops) but there are no errors.

I ran exactly the same G-code several times and each time this happened in a different place (and no single command seems to stand out) so I'm fairly certain the problem isn't with the G-code (or anything up to that point).

I've also installed a fresh copy of Repetier Host (not the Solidoodle version), manually configured everything using the Solidoodle version as a guide so as to sanity check (and in the process finding that for some odd reason Repetier Host Solidoodle hasn't got a few settings set by default that it should have - little things like filament temperature!) and tried the whole process on two completely different models of computers (both Windows 7 Pro). Same issue.

Thinking the control board was overheating, I tried taking the acrylic cover off and directing a fan into the housing. No difference.

The only clue I can think of is that it *seems* to print for a little bit longer if I up the baud rate. Repetier Host Solidoodle seems to default to 115200, setting it to 250000 (like for my Solidoodle 3) seems to result in a few more layers printing before it dies.

Here are my comms settings:

Baud rate: 250000
Transfer protocol: Autodetect
Rest on connect: DTR low-high-low
Receive cache size: 63

For those of you who have working Workbenches does this match what you've got?

I can't find any similar issues in the forums ... has anyone else encountered this issue?

Cheers!

- Raymond