remote operation (or just a power bug..)

[alaris2 1]

Weird thing just happened - I have an old fashioned incandescent bulb lamp sitting next to my Ultimaker so I can check the print quality better.

It was on, was printing away just fine.. then I switched it off.

Ultimaker stopped dead and refused to print further. no errors displayed anywhere - I can only assume it introduced some sort of power glitch and the arduino hung?

has anyone else experienced this?

[dojan 0]

I have had this a couple of times now, very irritating. As far as I can figure, it's due to power-spikes, ripples or just generally un-clean mains power. I'm considering connecting it behind my UPS which has a stabilizing transformer, but I'm worried that it will introduce power-spikes to my server instead...

Anyone who's good with electronics figure this one out?

[joergen 2]

I have had this a couple of times now, very irritating. As far as I can figure, it's due to power-spikes, ripples or just generally un-clean mains power. I'm considering connecting it behind my UPS which has a stabilizing transformer, but I'm worried that it will introduce power-spikes to my server instead...

Anyone who's good with electronics figure this one out?

this happens all the time, especially in countries with poor electrical systems (i.e. US). I put the UM behind a powerstrip with surge protector, and hadn't had a case since then.

[Daid 306]

Most likely it's some high frequency EMF or ESD that is causing this. "normal" equipment is shielded against this, but I don't think the Ultimaker is tested for EMF/ESD. Which is quite difficult and requires expensive equipment. This is why your computer is inside a large metal box, and your laptop contains a lot of small metal bits and pieces to shield it.

The high frequency EMF is most likely carried by the powerlines into the power supply of the Ultimaker and transferred there to the 19V line. From the 19V line it's only a small jump to any of the Arduino pins, which do not like this.

If you have grounded wall plugs, then use those for the Ultimaker, this can help.

[rewolff 0]

I don't think a spike on the 19V is the way EMF gets into the machine.

The 19V is the "power" to the machine. It powers the motors. Power the motors with 12V they will all still work, just not as fast as they would at 19V. (If you go below 12V, some MOSFETs might stop working properly, depending on the type and way they are wired). So a powerdip on the 19V will cause you to lose a step or two at high stepping rate, but nothing more.

A power surge on the 19V will cause 24V rated mosfets to short themselves. This is permanent. Nah, you're not seeing any power surges on the 19V lines.

In any case, because of the currents involved, there is loads of passive buffering of the 19V line. Any external influence will have to have enormous "power" to influence the 19V.

The arduino works at 5V. The arduino is powered from the 5V from the computer. There is a whole lot less "power" involved in those lines. A power dip or surge on the 5V may cause the atmega2560 to miss an instruction and crash.

So:

- Get a shielded USB cable.

- Put the cable away from the powerlines to that lamp.

- get a shorter cable.

Good luck!

[Daid 306]

I think my key words are "high frequency".

The most common cause of Arduino hangs are caused by ESD glitches on the IO pins, the power of the Arduino has a pretty good regulator, and there are proper capacitors on the board to catch most problems on the power.

Note that this is a difficult subject, and I only know a tiny bit about it. Without proper testing equipment it's almost impossible to say for sure how good your design is. At my work we recently had to replace 100 boards because of a clock line which could glitch a data line. This glitch happened only once every few hours, and was hard to trace. And this was after the boards where tested for EMF already.

[rewolff 0]

Yeah. Glitches on clock lines are the nastiest....

-> The 19V SHOULD be pretty "separate" from the 5V that the electronics runs on.

it is the POWER source. Hit the switch on the side of the ultimaker and everything inside just goes on, only without any power. The motors won't turn. The heating element won't work etc etc. Turn the 19V down to 12V and things work, but maybe not as fast as normal. Turn it up to 24V and things could work a bit faster, or things break and you get to keep the pieces. It is JUST power, no logic depends on it.

On the other hand, there is the 5V that the arduino runs on. The CPU is guaranteed to work at 5.5V, and will survive 6V. It could miss an instruction or crash or "anything" between 5.5V and 6.0V. (but it should work again after you get the voltage back to around 5V). Similar things hold below 4.5V: The cpu could miss instructions or crash or ... anything when the voltage drops below 5V. Even if the "drop" is only 50ns, that is a whole instruction time.

So that's why I'm guessing that when things crash if the bulb is operated you should look at the logic supply and not at the POWER supply.