Short version:
The side panels are not grounded!
The chamber heater is also not grounded!
Long version:
When installing the chamber heater, I checked if everything was grounded properly with a multimeter, because the chamber heater is AC and has a metal housing, which is in mechanical contact to the printers steel side panels. Turns out the side panels are not grounded. The aluminium profiles are grounded properly, but because of the side panels non conductive paint, the panels are not grounded at all. But the chamber heater relies on the side panel for its grounding.
In my understanding, that is a serious concern. If the AC wires get loose in the chamber heater, the side panel will probably be on the same potential/AC-voltage. If someone touches the side panel in this scenario, it would be pretty bad.
Disclaimer: I’m no electrician and not qualified for electrical installation. This is no guide.
My own “solution”:
I generously removed the non-conductive paint in several areas from the side panels and checked for proper grounding again. It worked.
Additionally I´m using a RCD because I want to be extra secure.
Yikes! Thanks for catching this. I don’t have a chamber heater yet, but since the bed itself is also powered from AC mains it is a similar hazard. I’ll also be adding a GFCI
While I have no doubt your solutions where a improvement, consider a multi-meter is for working with sensitive electronic devices, so is testing resistance at a couple of volts. It takes about 1000v to first bridge a 1mm air gap (but it can sustain FAR longer once you’ve got plasma, hence a Jacobs ladder works), so the gap required to show open circuit to is somewhere between 100th and 1000th of that 0.01 to 0.001 mm. A mains power short will be 100-250v depending on your location in the world, which is a very different story. It is not rare for devices to have paint specification as part of their grounding, most computer cases (commercial racks or home user) are an example, so we can’t assume that Sovol Zero isn’t a valid implementation.
If you want to test grounding or insulation, not component resistance values, you’ll need a megga, these output a pulse 1000v DC to give an actual idea of if there was a problem, and a couple of brands can be found under $100usd (although be prepared to pay >$1000 for the industry standard fluke), with the caveat that they ARE dangerous tools intended for trained electricians, and can hurt you or the items your working on if used improperly.
Well that is “shocking”! … Ba da bump tsss… couldn’t resist…
Seriously though, thanks for sharing! This might explain why I sometimes see my HDMI screen flicker when I’m loading filament through the PTFE tube. It’s not all the time, but if the air is somewhat dry I will see it, especially with PLA Silk filament. Perhaps enough static build up occurs to cause some weirdness, you know similar to the hair and the balloon experiment. I’m wondering if the lack of grounding for the side panel is related to this.
A multimeter is absolutely the correct tool to measure the continuity of the bonding and what OP found does certainly indicate a possible issue. Granted, I’m not sure exactly where and how the OP tested, but based on the description of findings, I’m inclined to see the possibility.
All metal components that a user could come in contact with should be electrically bonded back to the equipment grounding conductor (aka the green one) in order to provide a fault path for current to trip the breaker in the event of a hot conductor coming in contact with it.
A megohmmeter is for testing insulation resistance, that is, measuring resistance between things that shouldn’t be electrically connected. That is why they output higher voltage, in order to stress the dielectric strength of the insulating material they are testing through and get an accurate reading.
Additionally it should be noted, that measuring a low resistance with a multimeter doesn’t automatically prove, that the grounding is sufficient. It only proves, that there is electrical contact, not that it is sized adequately. E.g. if current gets high during a failure, the contact point could get damaged due to heat if it isn’t sized adequately, possibly breaking the grounding.
But if you measure a high resistance with a multimeter, it definitely proves that the grounding is not sufficient.
I measured the resistance between several spots (panels, aluminium extrusion) with my multimeter. One contact was always on the grounding in the printers power socket (printer plugged off of course).
Absolutely, that continuity exists does not prove it has the ampacity to handle the fault current.
I took resistance checks on my machine as well (no chamber heater installed).
All tests conducted with the meters common lead connected to the ground prong of the machine’s power inlet, and the measurement lead placed at the point under test. Meter leads verified before and after testing, measured 0.1 ohm.
All panel fasteners: 0.1 ohm
All blue plastic angle-bracket fasteners: 0.1 ohm
All frame extrusions (at bare aluminum points): 0.1 ohm
All frame extrusions (on anodized coating): open
All panels on powder coating: open
Back panel (at a filed bare-steel point): 0.1 ohm
Y-gantry extrusions (at bare aluminum points): 0.1 ohm
Y-axis linear guide rail: 0.1 ohm
X-axis linear guide rail: open**
Z-axis linear guide rail: 0.1 ohm
Z-axis stepper motor case: 0.1 ohm
**For a different topic/post, but this is also relevant to @MikeHides point on static buildup.
It pained me a bit to take a file to the panel corner after searching for any existing nicks, but the newly exposed metal was kept air-gapped from the frame during testing. Given the continuous reading, I removed the back panel and confirmed no bonding jumper or supplemental bonding method present. I also inspected the mounting holes and found two or three locations where the coating was already missing, hence the continuity during test.
As you rightly pointed out regarding electrical contact versus adequate bonding, I removed enough coating so that one of the countersunk mounting screws achieved circumferential contact with bare metal on each panel. Would this meet the standard for a UL listing? Probably not. Will it provide an effective fault path to clear a breaker? Based on my judgement, yes. For five minutes of my time, it certainly didn’t make it less safe.
@rpcyan also made a solid point about GFCI protection. A breaker’s primary role is to protect conductors; a GFCI’s role is to protect people.
I appreciate you sharing your findings and bringing this to the community’s attention. I’m certainly glad I read it.
@Samuel_Donahue Thank you for adding your exact measurements! They are in accordance with mine.
Here is a picture of my “solution”. I swapped the screws for stainless steel screws and filed down the panels paint (the other panel screws not shown in the picture too).
Given the depth of the discussion here is well beyond my basic understanding of electricity could one these GFCI Power strips be a viable solution for having a localized GFCI?
I’m in the US, and have a electrical panel with 120V 10-20Amp breakers etc., no GFCI sockets other than the bathrooms, ~1975 timeline. I’ve never seen GFCI surge bars like this before, only the GFCI electrical sockets. To top it all off, I’ve always assumed the breaker panel catches hard faults and protects people.
Thank you folks for having a great conversation about grounding, I’m about to add a chamber heater…. and I learned something today!
Looks like a quality unit. There are also extension cords with GFI boxes at the “wall end”
You need to verify your outlet is wired correctly. A shocking (pun INTENDED) number of wall outlets have the live and neutral wires reversed and/or no safety ground attached.
I might have gotten lost in this thread.. a GFI wouldn’t protect you..
this thread started that if you touch the sides of the printer after installing the chamber heater, it could kill you. Having a GFI wouldn’t protect you from being shocked.
@Lion Good point, some sort of electrical connection I would imagine needs to happen to trip the GFCI, but using one I would expect at least add some level of protection. I’m also not very informed in this area.
TL;DR (all personal opinion, not a recommendation):
Improving panel bonding is good.
Good quality GFCIs are good.
@jbsdfjbadgadf Your poor machine! I went slightly less aggressive but it appears your panels have continuity with ground now.
@MikeHides, my personal preference is to use a more well established brand and ensure it has some testing behind it. Eaton, Hubbell, GE, ABB, etc. I’ve never heard of MatchEasy and I don’t see anything to prove any sort of testing, so for me, its out.
@Lion Your right, the original post was about the chamber heater with a metal case being mounted to an isolated panel. And since I only see two power conductors, it does appear to be relying on the metal cased mounting to provide a fault path. So I concede. If the panels are completely isolated from ground, a GFCI will not trip if a fault energized it, and it would not provide protection against direct contact with that panel.
That said, a GFCI would still improve safety overall, since the panels are not the only components on the machine that could become energized under a fault condition.
In my testing on my machine though, the panels were not fully isolated. When I tested at a bare metal spot I exposed, it showed continuity to ground. My examination showed that continuity came from nicks in the panel coating around fasteners. In this scenario, or where coating has been intentionally removed to ensure better bonding, the panel is electrically connected back to the equipment grounding conductor, and as long as the premise wiring is correct, a GFCI would provide additional protection in the event of a fault.
As a Master electrician, I agree. A merger is used to test insulation on wires and for shorts in motors and such, not proper grounding. Properly grounded equipment should have extremely low resistance when measured by a standard multi-meter.
You are correct. It wouldn’t keep you from getting shocked but it will keep you from dieing, as it will limit the current going through you to less than 5ma.
