Ground PC boards.
Out of curiosity, I made a few resistive measurements of the grounding system in my newly constructed MK3 3D printer. These measurements are all referenced to Earth ground power input socket.
Instrumentation used: Fluke 115 DVM.
Operating conditions: Power off with the system unplugged.
DC resistance to various points on the system:
1) Earth Ground pin to Power Supply Chassis = 0.5 Ohms
2) Earth Ground pin to 3D Printer Chassis = 0.5 Ohms
3) Earth Ground pin to four screws on the right front of Y rails = 1) 104.8K, 2) 15.07K, 3) 1.1 Ohms, 4) Overload resistance measurement.
4) Earth Ground pin to four screws on the left front of Y rails = 1) 1.0, 2) 15.07K, 3) 1.0 Ohms, 4) Overload resistance measurement.
The conclusion of 3) and 4):
High resistance through the Aluminum oxide on the extruded components caused the significant variation in resistance measurements. The lower resistance measurement was due to the assembly screws breaking through the Aluminum oxide coating in the threaded screw holes, thus, contacting aluminum directly, allowing low resistance measurements. Aluminum Oxide is a nearly perfect insulator as revealed by the high resistance measurements.
5) All stepper motors are electrically "floating." It would be best if they were grounded. Normally, in other applications, stepper motors are attached to an Earth grounded metal chassis. In this case, they all attached to non-conductive plastic. Essentially each stepper motor casing is floating referred to Ground. The fact that they "float" is not a significant problem but they should be grounded.
6) Earth Ground pin to ENSY RAMBo Ground plain = open circuit, >1 Meg Ohms (Big Problem)
7) Earth Ground pin to Ground Plain of SD-card = open circuit, >1 Meg Ohms (Should be < 1.0 Ohm)
Voltage measurements referenced to Ground:
Operating conditions:
Mains Power applied, Switching Power-Supply is on and producing the correct specified DC Output Voltage:
Instrumentation used: Fluke 115 DVM.
AC Voltage (RMS) measurements at various points within the system referred to the Power-Supply chassis (Earth Ground):
1) Ground Plain AC Voltage of SD-card = 0.866V, 60Hz
2) Z Stepper motor AC body Voltage, Left = 1. 03V, ~22KHz
3) Z Stepper motor AC body Voltage, Right = 0.99V, ~19.3KHz
Apparently, the Power supply switching Voltage regulator operates at ~22KHz.
4) Y Stepper motor AC body Voltage = 0.735V, 60Hz
5) X Stepper motor AC body Voltage = 0.973V, 60Hz
6) Extruder Stepper motor AC body Voltage = 0.835V, 60Hz
6) ENSY RAMBo Ground plain AC Voltage = 1.405V, 60Hz (SIGNIFICANT)
It is interesting that there is 1.4V of AC Voltage appearing on the ENSY RAMBo Ground plane. I wonder what additional switching spikes are present on this sensitive ground plain? Furthermore, their peak Voltage amplitudes may prove to be significantly higher than the 1.4V, 60 Hz signal. A Spectrum Analysis measurement of this signal would be very reviling and could point out potential causes of operational problems in this part of the system. It is possible, if left as is may, present conditions that would damage or shorten the lifetime of the SD and ENSY RAMBo cards. Bottom line, this problem needs to be looked into and corrected ASAP.
Recommendations:
Find best ways to:
1) Connect all PCB ground plains directly to system ground, Power Supply Chassis, (Earth Ground).
2) Ground all Stepper Motor bodies directly to system ground. (less critical)
Solutions:
I grounded the ENSY RAMBo card through the hole in the back of the ENSY RAMBo card case. Fortunately, there is a threaded chassis hole located very close to the hole in the back of the plastic box. I connected the other end of the Ground wire to the nearest PCB mounting screw inside the box. After grounding, the AC Voltage at on the ENSY RAMBo ground plain was very small, virtually zero Volts.
The SD card was also grounded in the same way as the ENSY RAMBo card. The ground wire connection was between one of the PCB mounting screws and an M5x16r Front plate screw. (I used a screw that measured the lowest resist to the P.S. chasis.)
Results:
The MK3 3D printer produced outstanding results printing several test subjects, 3DBenchy_150um, Treefrog_50um, Gear Bearing_150um (The gears would turn). The machine worked flawlessly with no adhesion problems.
Re: Ground PC boards.
and all this to say it prints better I did nothing and it prints brilliantly and the gears work lmzao 😀
Re: Ground PC boards.
Thanks Gerard for a thorough electrical test. I’m still waiting for my MK3. Would it be possible for you to post a pic of how you grounded things? It will help me and others apply the same fix.
Thanks again.
Re: Ground PC boards.
Good post. This is one reason why it's so sensitive to static discharge.
Re: Ground PC boards.
How do you measure the AC body voltage?
From what I understood, all your measurements are against the earth ground. Without seeing the schematics it's hard to understand if you're maybe creating additional issues with your approach:
1. You might create a grounding loops
2. Prusa frame is a big antenna, so you're picking up a lot of RF interference. Just connecting it to the digital ground (-) might be dangerous and creates another issues.
I'm not an expert in pcb design. But from what I understand is grounding design a pretty complex thing with many factors. Especially if you're using a switching power supply (what we have in MK3). That's why I wouldn't do any those kind of ground modifications without confirmation by PR or any other experts in this fields.
Often linked posts:
Going small with MMU2
Real Multi Material
My prints on Instagram
Re: Ground PC boards.
Good post. This is one reason why it's so sensitive to static discharge.
Yeah, tell me about it! lol
I can't even touch the PSU or it confuses the LCD for a few seconds if I have a static charge. I keep touching the screws on the wall switch plates before I go near the thing.
I'll be adding ground leads this weekend. Hope it helps!
Re: Ground PC boards.
nikolai.r,
You understand is correct, all my measurements are with reference to the Prusa frame/power supply Ground/ Earth ground. First, my measuring device FLUKE 115 incorporates high impedance probs. They are not likely to create "ground loops" that aren't already present particularly measuring DC Voltages. Second, the Prusa frame is solidly connected to Earth Ground through the Green wire in the power cord. Therefore, not a "big antenna," at least with frequencies with which we are concerned.
Bottom line:
PCB Grounds should not "float" or exhibit AC voltages with respect to Earth ground or they're DC applied voltage. Unless someone can show a valid reason why these PC board ground planes should not be at Earth ground, I will leave my system as modified. From my view, grounding these ground planes are necessary and valid.
The system performance appears to be outstanding after grounding the PC boards. I believe the system favorably agrees with the modifications.
Re: Ground PC boards.
Electrical guy here.
I'm often pondering the grounding issue as well, not just on the printers, but on grounding in general. It's complicated.
There are a few threads scattered here about the grounding issue of the printers.
Topics related to printer resets mostly, and they are gone when grounding the frame or even the einsy.
My question is, what needs to be grounded? And how do we do it correctly?
Lets say we have a power supply with "frame ground" (FG) like this one:
https://www.jameco.com/Jameco/Products/ProdDS/2101518.pdf
Now one type of grounding PSUs can have is 'earth ground', 'frame ground' and 'floating ground'.
I'm still not sure if the last two are the same, some documents and fellow electricians say the same.
The minus pol of the PSU is not connected to ground here. Capacitors are put between the minus pol and the frame ground on purpose, to avoid EMV/EMI issues. (see the PDF) Measuring the minus pol to the AC ground results in several meg ohms.
Then we have another type we have is "minus pin connected to the AC ground pin, AKA true grounding. Like really good PSUs have, like ATX PSUs.
http://www.formfactors.org/developer/specs/ATX_ATX12V_PS_1_1.pdf
(page 15)
Here the minus pol is directly connected to the AC ground pin. ATX PSUs and other types of PSUs are super complicated and filter out EMI via complicated circuits.
So the case of the run of the mill 3D Printer (like the Prusa) LED PSU has FG (frame ground), and it's connected to Earth through the 3rd pin of the AC connector.
Grounding NEMA motors is new to me, will have to look into that. The case of the NEMA is not connected in any way to the coils of the motor.
Grounding NEMAs may help shielding EMI and other high oszillations from them into other electronics, like the control board (Einsy Rambo)
Now from the schematic of the mini RAMBo (can't find one for the Einsy) i can see multiple instances of grounding, but some are all decoupled with capacitors and high ohm resistors, while others go directly to true ground, which are not the PCB mounting holes.
The schematic is to small to make everything out, i'll have to upsize it later via Waifu2x to make out what's going on, and ask a friend who is a PCB designer what to make of here.
But the screw holes on the mini RAMBo and Einsy RAMBo are directly connected to the minus pol of the board, meaning connecting the board via the screw holes to ground would also connect the PSU's insulated minus pol directly to ground. And we have a insulated frame ground here decoupled from the minus pol for a reason.
While connecting minus via the control board to FG, it can work, but will introduce some problems like high EMI spikes that swirl around shitty ground connections in the house from a microwave or a washing machine into the delicate electronics of the printer.
ATX PSUs can handle that, LED PSUs cannot.
my tl;dr and gut feeling is:
connect frame to ground: OK
connect motors to ground: hmmaybe
connect Einsy to ground: definitely not
mini RAMBo and Einsy were not made to be grounded from looking at the schematics, and connecting the Einsy to the ground works it opens a door for other problems.
If you have printer resets, you can try to ground the control board, they should be gone then. But that circumvents other protective circuits in the PSU. Also this can damage electronics faster than if they were ungrounded, because of said sudden EMI garbage spikes from microwaves or vacuum cleaners on the ground of the house.
I used a ATX PSU for my MK2, these things rock, but only have 12V. Have to suffice with the cheaper LED PSUs for now because of 24V. At least i got a high grade one from MeanWell, and i suspect the PSUs from Prusa are pretty bottom tier with all the defects going around.
edit: revised a few things here from looking at the Einsy schematics i just found.
Re: Ground PC boards.
Proper grounding is something most electrical engineers do not get. When you have multiple grounds, for example an analog ground and a digital ground, the important thing is that they only connect to each other in ONE SPOT on the PCB. This is how you avoid ground loops.
From what I can see of th3 Mk3, the PCB was not designed by someone who is an EE. Just my opinion.
Re: Ground PC boards.
gerar.r2,
By ground loops I meant not the measurement itself. Depending how many ground connection you're adding to the system, you might create ground loops while printing.
Agree. PCB ground shouldn't float, but you're measuring AC voltage. That's exactly what I'm trying to say. It's most likely an inducted voltage. Important one is DC. I just checked the schematics for EINSY 1.1, not sure if it's the MK3 board but it's a good example where the frame ground should be connected.
If you could validate and maybe post a picture where you attached your ground, it would be helpful for the others as-well.
One measurement is missing, or I missed it. Which values do you get on the power supply between earth ground and digital ground (V-)? From my understanding you should get 0 Ohm and 0V DC there.
Often linked posts:
Going small with MMU2
Real Multi Material
My prints on Instagram
Re: Ground PC boards.
I would think putting a little more money into a good PSU or just a more capable one for the US market would cost less than replacing them and paying for shipping.
Re: Ground PC boards.
Looked at the schematics a bit more, and as always, i'm weirded out how and where the engineers @ ultimachine put in the actual grounding of the boards.
Besides the USB shielding, i can't find another actual hard ground on the board. And i'm not even sure the USB shielding is going somewhere on the PCS, i would need to disassemble the printer again to check for real.
Usually, the correct way to ground this would be to simply plug in a USB cable, since the shield carries over actual hard ground from the ATX PSU of the connected PC.
Hard connecting the Einsy to ground without a PC would be to solder on a cable to the USB shielding and connect that to the frame extrusions, which should be also grounded.
This should be sufficient.
Any other takes on this?
Re: Ground PC boards.
One measurement is missing, or I missed it. Which values do you get on the power supply between earth ground and digital ground (V-)? From my understanding you should get 0 Ohm and 0V DC there.
Thanks devilhunter for the pretty good explanation. It helped me to understand depends on the power supply you will not get 0 Ohm. But 0V DC should still apply in a proper grounded home, right?
Often linked posts:
Going small with MMU2
Real Multi Material
My prints on Instagram
Re: Ground PC boards.
One measurement is missing, or I missed it. Which values do you get on the power supply between earth ground and digital ground (V-)? From my understanding you should get 0 Ohm and 0V DC there.
Thanks devilhunter for the pretty good explanation. It helped me to understand depends on the power supply you will not get 0 Ohm. But 0V DC should still apply in a proper grounded home, right?
There are a lot of things that will cause the results that OP measure.
Voltage on the case on the NEMAs? -> EMV. (you are dealing with magnetic fields here)
Voltage on the Minus pol measured from the frame ground? -> capacitors at work. (voltage will discharge slowly when measured)
Basically i get +1.4v when measuring from ground to minus pol, and it quickly discharges to 0V. Yep, that's a cap.
I think Prusa did the grounding OK here, provided the Frame is connected to the AC ground, for safety and EMI issues.
I don't have a vanilla MK3, so i can't measure. (mine are self built)
If the engineers at ultimachine built their boards right, true grounding comes via USB.
Re: Ground PC boards.
Any other takes on this?
Only one. RP2/3 is usually connected through USB to Prusa. Most RP2/3 are powered by a small power supplies without earth ground.
Just wanted to make this cases more complicated 😀
Often linked posts:
Going small with MMU2
Real Multi Material
My prints on Instagram
Re: Ground PC boards.
We'll have to ask the lead EE why there's a floating ground. (My best guess is so a partial short to chassis in one of the heaters doesn't cause a thermal runaway.) I added a 47 ohm fusible resistor from V- to ground, which theoretically would do even better at stopping EMI than a short yet will not completely defeat the floating ground.
Inspired to get into 3D printing by Micah Elizabeth Scott, Naomi Wu, and an anonymous Bitcoin girl I met in college.
Re: Ground PC boards.
Any other takes on this?
Only one. RP2/3 is usually connected through USB to Prusa. Most RP2/3 are powered by a small power supplies without earth ground.
Just wanted to make this cases more complicated 😀
Hah, you have a point here.
Cheap USB chargers are not grounded at all.
So a no win situation to a normal consumer then? heh.
Re: Ground PC boards.
OK, i came across this which explains belt static and grounding NEMAs:
https://www.duet3d.com/forum/thread.php?id=3506
Note: it is highly recommended that the stepper motor casings be grounded, especially in belt-driven printers. Otherwise, motion of the belts causes static charge to build up, which eventually arcs over to the windings. If the motors are screwed to a metal frame, grounding the frame is sufficient
Which makes sense.
Some trouble can come from here, from belts that charge themsleves up and transfer that over to the case into the stepper driver. That coupled with EMV from the magnetic fields can cause some issues.
3D printers and their stepper drivers are fairly robust, you can only kill a stepper via quickly moving a carriage from side to side while it's connected to the board. But little things like this can cause interrupts that make the stepper drivers hiccup.
OK, new recommendation:
- ground the frame
- ground the motors (unscrew one screw, thenput a small wire around the screw, screw it back in. connect wire to the frame)
unsure about this:
- ground the Einsy via USB cable from a PC (easy solution for now) i'm still looking how to do this without a PC or soldering, but i'd say solder on a wire to the USB shield, then connect that to the frame.
Re: Ground PC boards.
The motor bodies are floating as they have no electrical connection to anything. You are measuring induced voltages from the windings.
Earth grounding the 0 volt side of the switching supply can cause issues due to poor/improper earth grounding on the power mains side - AKA loss of the MAGIC smoke - I have fixed some of these for people who had grounded the 0 Volt side of the power supply!!!! It usually makes the grounding fuse on the USB circuit blow the component pads off the control board(Einsy) during the vaporization of said fuse along with whatever is on the other side of the USB cable. Your machine, so you should be free to decide what you do.
Some times the power supply will have a fault(manufacturing/failed isolation) that will make the 0 volt side actually be non-low voltage(aka - line voltage = nice flash) 😯
If you had a two pin non-polarized laptop you could make it electrically hot by how you plug it into the wall outlet. Then you would apply line voltage on the 0 volt common connection.
Fortunately the mains side of the supply comes pre-built.
Do not think the ground is the same on two portable AC generators! So you design to cover as many possibilities as possible.
Re: Ground PC boards.
Last reply on the grounding issue from me for now.
It seems that the grounding of USB is somewhat... up in the air.
If you measure a USB shield against AC ground, you"ll get usually 0 ohms. However, they are usually not directly connected.
USB shield serves, well, as a SHIELD against EMV into the data lines. Where the shield is connected to is left up to the designers.
Some will connect it directly to the ground, others via many other methods
https://forum.allaboutcircuits.com/threads/usb-device-cable-shield-connection-grounding-it-or-not.58811/
https://electronics.stackexchange.com/questions/290777/usb-3-0-hub-shield-connection
So i'd say, leave the Einsy alone for now.
No grounding whatsoever.
Final recommendation:
Connect the frame and the motor casings to ground if you want (these are usually connected via a grounded frame on well designed CNC machines anyway), connecting other things to ground is dangerous for human and electronics alike.