Should the 24 (-) be grounded?
-
I have my entire frame grounded to earth (mains), and at my PSU I tie in earth mains with the DC - terminal. Dc42 recommended this setup. A main benefit is ensuring there is no potential between earth gnd and DC-. The main drawback is potential for ground loops with attached USB devices. Best to not have the attached USB device connected to earth ground, IE, use a laptop on a battery not plugged into the wall.
*not actually a robot
-
These things are great for avoiding USB ground loops (for low-ish bandwidth applications)
https://www.banggood.com/USB-Isolator-USB-to-USB-Optocoupler-Isolation-Module-Coupled-Protection-Board-ADUM3160-Isolation-Voltage-2500V-p-1454368.htmlor search your favourite vendor for ADUM3160 or ADUM4160 ; often around $7-10 shipped from China, usually more from US or EU.
(being fairly bulky, I tend to cut a usb cable in half and replace the usb ports on the module with the now two usb pigtails; it says optocoupled, but it is magnetic i believe)
-
@bot said in Should the 24 (-) be grounded?:
A main benefit is ensuring there is no potential between earth gnd and DC
Thanks @bot. What is the motivation for that? Safety? Signal integrity? Something else?
-
Its mostly a safety thing, but it does assume you have RCD/GFCI to catch any leakage current to mitigate additional risk of shock.
-
From what I remember reading, signal integrity. I can't personally figure out the pros and cons so I just went off of the articles and suggestions listed in the thread here: https://forum.duet3d.com/topic/7582/meanwell-power-supply-ac-ground-and-v-connection/10
-
Thanks @bot , that's a good thread. @dc42 says there that it's for safety as @bearer suggested.
@bearer, I do have a GFCI specifically for the printer. Once in a while I connect an oscilloscope to the printer (e.g. to view PWM) so this also create a ground loop. My power supply is Meanwell from a legit source and is UL listed so I trust it in general.
-
@zapta said in Should the 24 (-) be grounded?:
so this also create a ground loop.
creates the potential for a ground loop I would say, unless you use the ground clip as if it was a differential probe; otherwise it would depend on some other fault to cause a ground loop - an thats when the GFCI should trip.
-
@bearer said in Should the 24 (-) be grounded?:
unless you use the ground clip as if it was a differential probe
Yes, you are right. As long as the oscilloscope ground is connected to v(-) it should be ok.
-
@zapta said in Should the 24 (-) be grounded?:
As long as the oscilloscope ground is connected to v(-) it should be ok.
Still under the assumption of no other faults; if you accidentally disconnect the Duet's Vin ground lead the current will take the path through the scope. Thats what I was getting at with putting emphasis on potential for ground loop.
Having if the scope is on a different circuit with faulty mains that could also come up the scope ground to the Duet, but GFCI = FTW.
-
@bearer said in Should the 24 (-) be grounded?:
if you accidentally disconnect the Duet's Vin ground lead the current will take the path through the scope. Thats what I was getting at with putting emphasis on potential for ground loop.
I don't think this will trigger the CFGI. AFIK the CFGI looks for current difference between the two mains lines, regardless of ground wire current.
-
@zapta said in Should the 24 (-) be grounded?:
AFIK the CFGI looks for current difference between the two mains lines
That is indeed how they function, but if the current in live and neutral is the same then there can't be any current in the gournd wire, it has to come from somewhere, and when it does that GFCI will trip.
-
@bearer, I think that if the printer's and osciloscope's grounds are not at same potential, connecting them will create a ground loop current without involving the current through the two mains wires that go through the CFGI. I am not an expert though.
-
You're sort of right; but grounds shouldn't be at different potentials unless there is a leakage from/via live or neutral somewhere - and then the GFCI for that circuit should have tripped already. (so still a fault condition and just a potential for ground loop)
