Hotend temp cuts out in middle of print

[GarageBay9]

Greetings folks,

Last night I finally got my Rostock Max built and running, all calibrated and ready to go. Didn't even have to change the delta radius after I tuned the endstops. Told my wife to pick up a lottery ticket on her way home, just like the instructions said...

Anyhow, once I finally got the Onyx up to temp (it seems to top out at about 97c, but my ABS filament was sticking fine at 95 so I ran with that), I started printing the Skeinforge hollow_cube.stl calibration object. Things were going great until about 16 layers in, where the hotend temp just went into freefall until it hit about 30c (a little above ambient temp, it's been warm around here), and stayed there no matter what I told it to do with either Repetier or the LCD control panel.

I had to kill the job, shut down Repetier, shut down the Rostock, wait a while, and turn everything back on for the hotend to respond to heating commands again. The RAMBo fuse was not blown, like I found mention of in another thread, as I was able to start printing successfully again... until just about the same spot, where it happened again.

My gut instinct is either gcode bug or I'm tripping some kind of safety cutout to the hodend temp (I ran it as high as 235-240, but settled for 220 for the second job and it was working fine until it cut out). But gcode doesn't seem like it would keep the hotend from responding at ALL until the printer was powercycled. The hotend is getting full power according to the temperature graphs in Repetier, but it just keeps dropping.

I'm stumped. The Onyx stays hot and keeps responding, it cooks along nicely at 95c.

[kbob]

Maybe your hot end thermistor came loose or disconnected. Check whether the hot end is really cool to the touch when the thermistor reads 30℃.

N.B., if the thermistor is not working, the unregulated heater can destroy your hot end or burn your house down. So watch the printer carefully until you're sure you know what's going on.

If the thermistor is okay, then check the wiring to the heater resistors. Also look for intermittent issues like a bad crimp or uninsulated wires touching.

[GarageBay9]

Hotend is physically cool when it's reading low, and the filament behavior corroborates that (it starts going wonky and then stops extruding entirely as it drops below effective temp). I've checked the hotend power end-to-end just now and the circuit is good. I sealed it up pretty tidy with shrinktube and kapton tape, so I'm not concerned about movements induced shorts. The thermistor is snugged in there WELL and isn't going anywhere. It also reads correctly the whole time.

The part that stumps me is that it seems to be based on some relationship between the bed and hotend activity, and power cycling the printer after a brief wait usually seems to reset the problem.

[626Pilot]

Check the wiring in the terminal block you have the hot end connected to. I discovered that the block my Onyx was connected to was partially melted because the wires weren't making enough contact. (I was using two wires each for + and - to get more current, and I had soldered the leads together, thinking it would help. It actually made it harder for the wiring block to make contact with the leads, and this had caused heat, which in turn caused melting.) Make sure that only the copper part of the wire is screwed down, and that the insulation is NOT being clamped by the block. Also, make sure the connector is seated all the way.

Also, take a look at the Temperature Curve tab. Do you see the temperature curve suddenly dipping or spiking? Does it come and go? If the firmware can't get a reliable reading, it will go into "dry run" mode. It will still move the head around, but it will disable the heater and ignore any requests to turn it back on until the unit is reset. You can watch for a message to this effect in the log window.

[GarageBay9]

Success!

It appears that the effects of making temperature adjustments from the LCD panel is sensed by Repetier as uncommanded temperature fluctuations, which sends it into dry-run mode. I've gotten several good prints from it now, with the rule that once I hit "start job", I don't touch the LCD panel controls. Works like a charm so far.

Short version: my impatient grubby mucking fingers caused problems again.