SYMPTOMS: I've had multiple prints fail with a layer shift for no reason after several hours. I've looked for all the standard layer shift root causes (loose belts, loose servo pulleys, mechanical restrictions from cables etc.) and I can't find any typical culprits. While it is possible I've overlooked something mechanical, this is my 6th 3D printer, I've tinkered with printers for years, and I am in the engineering field with some knowledge and experience on both the mechanical and electrical side.

SETUP: Prusa MK4 built a year ago (no problems with layer shifts for last year), upgraded to MK4S with retrofit kit a month or so ago. Custom input shaping values tuned via the official Prusa accelerometer kit.  The problems started after the upgrade. We have dozens of MK3S+ printers at my university and we don't have the layer shift problem with those - and they are in the same dry low humidity climate where static is the most concern.

This may not be a concern for many people in higher humidity climates... but this may become a real problem for a few of us. I've confirmed that small static discharge zap resets the board, as I was checking the tightness of the servo mounts on the X-axis and just touching the Allen wrench to the servo screws reset the board. While the printer restarted, and re-homed, there was a layer shift I saw develop in real time.

I now suspect that under certain circumstances the printer can self build static even without external interaction (e.g. user zapping it). Then it may self-discharge somewhere, resetting, ruining the print. The higher belt and extruder speeds on the MK4S only augments the problem - possibly static is building up faster than it can self-dissipate now for the faster printers.

While I understand that zapping the USB port (or something internal on a circuit board) is uncool and hard to protect against, the printer should be more immune to environmental zaps to its chassis, servo motor exterior, external bits etc. If every laptop, phone, and consumer device rebooted with every zap we would all be in a real pickle.

Sounds like we are in for a scavenger hunt on how to ground everything and manage static drain. 

I will reach out to Prusa to see if there is a way to log during a long print, to at least record a self-reset event, which would explain my layer shifts, and to see if they have any insight into a self build and static discharge event reset. If I make any additional progress on this behavior I will post here. I'm curious if anyone else sees additional issues, particularly on the newer MK4S and above.

The frame of the black power supply is connected to the ground in the wall plug via the power cable and the power supply frame is connected to the enclosure. With the various surface finishes, there's a chance that's not true in practice in all cases (but it should be). - emphasis added

To test this, I pulled out a multimeter and did some tests. The power supply case (black) is indeed connected to the ground pin on the power connector. However, my enclosure is _not_ connected electrically to the power supply. The power supply is connected to the backing plate (starting in this step), so the backing plate and the power supply are electrically connected. The backing plate is then attached to the enclosure (starting on this step), but only using the plastic rotating pieces. The top is held to the enclosure magnetically, through a printer part. There are no electrical connections from the power supply to the enclosure.

For completeness, I even tested whether there is continuity around the enclosure itself (in my case there is not). Each edge piece of the enclosure can potentially be electrically disconnected from its neighbor, as the pass-through holes for the screws are coated. The threads to connect the enclosure are on the piece connected behind, and everything looks like it is floating by itself. One way to electrically connect the entire enclosure to itself would be to insert these between the sheets when assembling.

You are essentially correct but there is no way to predict where the static discharging touch will occur, especially if printing with the enclosure open - so the route to ground may still affect sensitive electronics.

Cheerio,

There is some question about the 24V however. The output of the power supply is isolated from the case and all the printer electronics is generally happy floating from building ground -- that's how Prusa set it up and it obviously works. I think it's better to earth ground the board (otherwise what is the point of the bypass caps), but there's been plenty of debate on that.

Would putting a touch point on the enclosure with a large resistor do the trick?

You're still dumping charge through the system.  If you are certain the enclosure itself is directly grounded and not grounding via the printer's power cable, then you should be OK.

Cheerio,

https://forum.prusa3d.com/forum/english-forum-general-discussion-announcements-and-releases/static-shocks-when-touching-the-build-sheet/paged/2/#post-588901

and:

https://www.eevblog.com/forum/3d-printing/esd-and-3d-printers-from-filament/

Creality had an upgrade due to this too.

Hope that will help.

Cheers.

Other interventions (printer ground, antistatic mat ...) have not worked reliably.

wbr,

Karl

Is there somewhere you could mount a grounding point (metal contact, resistor to ground, for fun include a neon) you could touch before touching the printer?

I touch a stainless steel kitchen sink before certain electronics but without a resistor I sometimes get a shock.

Cheerio,