JohnFox

jonnyb Thank you very much for the reply, interesting and explains a lot. ( I'm a biochemist by background not an engineer ).

Do you know what prompted Ultimaker to pick the types they use? I'm guessing that the rollers are better able to cope with the stress of the cantilever build plate but not sure. Again - just curious.

Are there any engineers out there? What are the advantages of using the phosphor bronze slide bearings on the X & Y axes and the roller bearings on the Z? The Z axis roller bearings seem to be quite a source of trouble yet the X & Y slides are problem free, ( although I shouldn't tempt the fates ). Just curious.

You should be OK now. I give the two pillars a wipe with a kleenex every two weeks or so, this removes any build up of oil from the ball races and keeps the friction between the balls and the pillars high.

Whatever it was doesn't seem to be an issue now. Z travel is clearly slower during retraction but it's not a problem. Obviously due to the little gremlin that lives in the controller section, I disturbed him

Just changed the sensor, I became pretty sure that the wiring was the problem. So far so good. thanks

I'm pretty sure the problem is with the bearings. Unless the rods are very marked and dented, they should be OK. If you can undo the bearings, ( it's a real pain to do this ) and slide them up and down the rods, they should run smoothly with a slight 'whirring' sound. If the balls in the bearing get out of position they will suddenly lock and the whole thing jams. A slight twist and it will run again. In my experience, the more oil on the pillars, the more times the bearings lock. Not sure of the advantage of the roller bearings over the slide bearings used on the X and Y axes, any mechanical engineers out there? This banding problem seems to be a pretty common fault. Thanks. But I wonder if this has any implication for the hardware? The help tooltip says "Rapid movement speed for Z-axis when machine is not printing, should match actual Z-axis movement speed between layers for accurate print times"

Dude I did a system reset and things seemed to have improved, at least the object is now printing as it should. What I have noticed is that during the layers when retraction is occurring, the Z motor speed seems to be about half the normal step speed, it's quite obvious. What I don't know, is whether this has always been the case. When the fault was happening, the Z speed appeared to get slower and slower until it didn't work at all. The feeder retraction speed seems to be as it always was, so I don't think that has been affected. Later on, I will check all the motor connectors, they are the most obvious source of the fault. Not sure about Z hop, never been into the Gcode. It almost looks like a processor overload. thanks

I have had to replace my nozzle sensor and did a nozzle replacement at the same time. Normal printing now looks OK, a test print which did not require retraction looks perfect. However, when I print an object where retraction is required, the Z axis motor slows down and eventually stops turning, killing the print. I strongly suspect this is electronic as I had to undo the controller in order to get at the sensor connector, so worried I have upset something. The print that failed has been made several time before without any problems, so the fault does not lie with the Cura file. Z travel looks perfect, no evidence of mechanical issues. Any ideas? Suspect I have disturbed something so that either the retract or Z motor is drawing too much power and the controller can't run both at the same time. Just run another test on a larger object that needs retraction. Z axis is working now but it is clearly slower than normal while retraction is on. Just about to repeat the test with retraction switched off.

Why didn't you try the Arduino 'Due'? It's a lot faster than the Mega and you can run touch screens off it.

I would stick to a glass plate. Can't see why bits should come out of it though, do you use a blade to prise the parts off while they are still warm? With PLA and Prit-Stick, the models stay on the plate very well at 60C, just wait a bit longer for the plate to cool.

Just gone back to the iakovs post. That sounds exactly like my machine, the Z step is a bit noisy but is working perfectly OK. Don't think it is an issue at all, suspect the noise comes from the sudden start/stop action and vibration of the bed.

It looks like a dodgy connector on the board then. I'm printing at 210C Can't reproduce the error at all, suspect it is due to the little gremlin that lives underneath the machine :-) This would help, if he got a "Heater error"... Just to be sure: usually these error messages are showing "ERROR - STOPPED" in the first line... and the second text line looks like this? "Temp sensor" (without the appendix "...BED") This one is triggered, if the temperature sensor inside the heater block reads less than 5°C or more than 275°C. At which temperatures are you printing? There's no connector near the head, the wire goes all the way down to the electronics board. You can check the connection there. Or can you reproduce the error if you wiggle the cables a bit?

When I fitted the new ones they ran very smoothly up and down with hardly any friction. The old ones locked repeatedly. I removed the top of the machine, ( I think ). It's worth wiping the pillars from time to time to remove any oil that comes from the bearing themselves.

I suspect it is due to oil on the Z axis pillars. The Z axis bearings are ball bearing races and I think that oil on the pillars causes them to slide and not roll, making a jam. I had this problem for some time, new bearings and no oil fixed the problem. As an experiment, I oiled the pillars again, promptly creating the fault. Cleaned off the oil and everything was OK again.

All it says is 'temperature sensor error', nothing to say which sensor it is talking about. The fans are normally off on the first layer, 50% on layer 2, then 100% beyond that. The bed and nozzle are heating up for certain and pretty close to what they should be. A loose connector seems the obvious cause but can't explain why only on the first layer. I am assuming that if the controller sees an open circuit on a sensor it halts immediately. Does the nozzle sensor have a connector around the hot end area?

I think the oil on the pillars is the real culprit, the bearings hardly get a lot of use, they only go up and down once in a print.

Just had some 'temperature sensor error' messages on my UM2. They are fatal, i.e. only switch off and on again clears them but then the machine works perfectly again. Oddly, I have only seen these during the first layer. No obviously incorrect temperatures on either the nozzle or build-plate. Any ideas?

Wonder if your PLA is a little over size and getting caught in the head. Also cut the end off. If I do a 'filament change' I often get a small blob on the end of the filament which is quite a bit larger than 3mm.

I take it you stopped the print before the end? I think it was printing correctly but you should have either inverted the model in Cura or printed with supports, ( inversion is better ). All that mess on the rear of the print is due to filament being extruded into air without any support, ( it's the clip for the card holder ). Inverting the model would put the top of the object onto the build plate and give support all the way round.

I have printed this one:- http://www.thingiverse.com/thing:177794 Fits onto the spool holder of the UM2. Coils can be a bit of a pain but the big benefit is that the filament isn't in tight circles which can increase the friction in the Bowden tube.

Just downloaded Izzy's files. Thanks, that should sort it!

I agree with cloakfiend PLA is FAR easier to print than ABS. For my applications, the only reason to print in ABS would be its stability to high pH liquids. When I switched from ABS to PLA, the print failure rate dropped from about 40% down to zero.

Hot end fan spacer. Just printed the fan spacer but on my UM2 this spacer is far too thick. It pushes the fan back by about 5mm and thus prevents the heater and thermocouple cables from entering the black plastic bearing shell. If you force it, the cables will be damaged. I tried reducing the thickness to about 2mm but still the cables would be crushed by the bearing holder shell. The air flow problem is not as serious as it sounds. Yes, there are only two full slots where the air flows in but those slots, ( at least the left one ) are in the perfect place to cool the PTFE coupler. It's is possible that by allowing the air to flow though all the slots instead of two, it may reduce the flow of air around the PTFE coupler. My coupler doesn't seem to get too hot anyway.

Don't oil the pillars, that seems to be part of the problem. If they are oiled, the balls in the bearings tend to slide rather than roll and make them more prone to jamming.