Grounding steppers
-
I've just done quick test on my anodized frame between 2 points about a metre apart. With the probes just touching the surface the resistance is greater than 10M Ohms which is the maximum range of my meter. However, if I just lightly scratch the surface with the probe, then the resistance drops to 0.5 Ohms. From that, I'd say that it's just a matter of scraping the anodised layer away (which is only a few microns thick) in order to make good contact for grounding purposes.
-
Scratching is certainly an answer but to be real ground the frame should have unbreakable contact between all the extrusions it consists of and be connected to main power protective ground. Otherwise this has no sense from safety and static charge protection points. So you need scratch each and every single extrusion in your printer and connect them all with wires ended with appropriate crimped contact. Obvious but very tedious solution! But to my pity can't invent any better for a while. Anyway steppers SHOULD BE GROUND with or without the rest of printer.
-
@fma:
It depends of the class of your power supply. In most class I, GND of 12V/24V is connected to main earth.
Are you sure?
This would mean a 24V 400W PSU "-" has a straight path to mains earth - which would immediately trip the RCD/GFI!
AFAIK to have any meaningful DC PSU the "-" should be completely isolated from the AC side.So the question remains - grounding to WHERE?
I guess a simple bleed-off to "-" via a high impedance path should do the trick. -
@fma:
It depends of the class of your power supply. In most class I, GND of 12V/24V is connected to main earth.
Are you sure?
This would mean a 24V 400W PSU "-" has a straight path to mains earth - which would immediately trip the RCD/GFI!
AFAIK to have any meaningful DC PSU the "-" should be completely isolated from the AC side.So the question remains - grounding to WHERE?
I guess a simple bleed-off to "-" via a high impedance path should do the trick.Do you know how RCD/GFI works? Any decent designed and manufactured electronic device should have connection between protection ground of mains supply and 0V (signal ground) of low voltage circuitry except very special cases. There is no problem for RCD/GFI in this case.
-
@zov:
Scratching is certainly an answer but to be real ground the frame should have unbreakable contact between all the extrusions it consists of and be connected to main power protective ground. Otherwise this has no sense from safety and static charge protection points. So you need scratch each and every single extrusion in your printer and connect them all with wires ended with appropriate crimped contact. Obvious but very tedious solution! …............
No not really. It depends how you build the frame. I tapped the ends of the extrusions then used button head bolts which engage in the slot of the extrusion it fixes to. A clearance hole for a hex key provides access to the screw head and when tightened, it bites through the anodising and makes good contact. I've connected earth braids to several of the extrusion cut ends which goes back to the earth of the mains plug. I did the same with the aluminium build plate. So every frame member and the build plate is connected to earth. I remember checking this when I first built the printer as I was concerned that if the 240v mains bed heater developed a fault, it could potentially make the frame live.
-
Thanks Ian, but I consider something like David's KosselXL style so no tapped holes to connect extrusions into frame. Keep trying to devise something appropriate for deltas.
-
I have a possible solution (for people building new printers).
If a cable with a braided shield is used to connect the steppers, it is typically recommended to connect only one end of the braid to AC ground, for noise-rejection.
Could we then connect the other end of that shielding to the stepper motor body? There is still only one connection to ground on the shield, but we also piggy-back off the shield to ground the stepper frame as well.
Thoughts?
-
@zov:
Do you know how RCD/GFI works? Any decent designed and manufactured electronic device should have connection between protection ground of mains supply and 0V (signal ground) of low voltage circuitry except very special cases. There is no problem for RCD/GFI in this case.
If you provide a low-resistance current path from 24V to Earth there should be an imbalance between the Neutral and Hot wire of your house, because some of the current finds its way back through the Earth wire.
And this should trip the RCD/GFI, or should it not?This means DC PSUs should have floating "+" and "-", otherwise it would trip all the RCDs/GFIs…
-
Using braided shield as stepper ground wire is good solution but a problem remain where connect it on the motor? And how to ground frame witch the stepper is mounted on? BTW Duet board has no provision for connecting braided wire to signal ground.
-
There would be no need to ground the frame (unless protecting against mains voltage flowing through the frame). Some cables with shields include a drain wire, to make it easier to ground the shield.
The drain wire would be connected to the stepper frame, using kapton tape, or a circular ground lug on the mounting screws.
The OTHER end of the cable, near the DUET, would have a drain wire that needs to be connected to AC ground. This would ground the stepper, while providing the drain for the shield.
It would be essential to make sure that there is no link between the frame, the stepper, or the shield to the Duet negative path – the grounding must take place through mains wiring to earth, in order to avoid ground loops.
-
@zov:
Thanks Ian, but I consider something like David's KosselXL style so no tapped holes to connect extrusions into frame. Keep trying to devise something appropriate for deltas.
Ahh yes. I know next to nothing about Deltas but aren't all the motors fairly close together? As Simon(DJDemon) pointed out, we can exclude extruder steppers as there are no belts involved. So, I'd have thought it would be a fairly easy to loop a wire from one motor to another with connection to earth from the final motor? Or have I missed something?
-
Ahh yes. I know next to nothing about Deltas but aren't all the motors fairly close together? As Simon(DJDemon) pointed out, we can exclude extruder steppers as there are no belts involved. So, I'd have thought it would be a fairly easy to loop a wire from one motor to another with connection to earth from the final motor? Or have I missed something?
You are right as always
All motors are quite near each other and even extruder one (Nimble) so no problem connect them with daizy-chain wire terminated at PG. But the problem of grounding printer's frame remains anyway.@bot: There is no ground loop concerns in this case as they are relevant mainly to sensitive small signal inputs when cable has a ground signal wire too. In our case it's no matter where and how often shield will be connected to ground. Common sense dictates that it is much simpler connect it to some grounded point on printer's case.
-
Okay if its a 2020 or similar frame and held together either by tapping the extrusions and drilling or by corners with half t-nuts (which bite into the extrusions and make a contact) then the frame should be (once you scrape through the oxide) electrically connected all over. If your corners are plastic, then youll need to use wires/screws/tnuts and ring terminals to connect the extrusions electrically to one another. I verified this when I set up my mains heated bed, the entire frame is electrically connected to the 240v Earth (and therefore the RCD should I have a mains short) to the bed/frame.
As for what to ground to, you could ground the static away to your DC negative, but then you are potentially going to create problems if any positive DC wire touches the frame when you are tinkering, or your mains earth but not the two connected to one another, I can't see that being a good idea.
If your motors are attached to the frame then you are probably already there, unless the oxide around the screw heads is preventing conduction, scrape some away and check resistance between the motor and frame. Otherwise how about a ring terminal placed under one of the motor bolts with a wire which is grounded to the frame using another ring terminal and a screw and tnut nearby, if like mine your frame is earthed.
As for RCD's they trigger in one of two ways. They have a coil inside which energises creating enough DC current to move the solenoid and unlatch the double pole contactors either if there is a discrepancy between live and neutral so current is leaking away such as into you or if current flows down the earth because the short has touched the earthed case of the device/frame. 10mA in the case of the one next to my printer which keeps it well below the level where a lethal shock is likely. But I suspect some static being generated from a belt is going to be pico-amps in magnitude even if it might be 50kv. Otherwise static "shocks" would be potentially fatal.
-
@bot:
I have a possible solution (for people building new printers).
If a cable with a braided shield is used to connect the steppers, it is typically recommended to connect only one end of the braid to AC ground, for noise-rejection.
Could we then connect the other end of that shielding to the stepper motor body? There is still only one connection to ground on the shield, but we also piggy-back off the shield to ground the stepper frame as well.
Thoughts?
That is what RepRapPro did in the Ormerod 2 design. One end of the braid was connected to the stepper motor body, the other end was connected to the metal enclosure that the Duet was mounted in.
-
@bot:
I have a possible solution (for people building new printers).
If a cable with a braided shield is used to connect the steppers, it is typically recommended to connect only one end of the braid to AC ground, for noise-rejection.
Could we then connect the other end of that shielding to the stepper motor body? There is still only one connection to ground on the shield, but we also piggy-back off the shield to ground the stepper frame as well.
Thoughts?
That is what RepRapPro did in the Ormerod 2 design. One end of the braid was connected to the stepper motor body, the other end was connected to the metal enclosure that the Duet was mounted in.
Excellent, thanks for that info.
This tidbit caught me at the exact right moment – I'm currently selecting the wiring for my printer. I'll make sure the cable has a drain wire on the shield to make it easy to ground.
-
@DjDemonD: about grounding extrusions through biting t-nut - it is a very interesting idea, should check it at once. Thanks! And I know how RCD works it's no concern for me.
-
@DjDemonD: Right now check how t-nut bites into extrusion, it works and has virtually zero resistance! You made my day, thanks!
-
This has suddenly become a lot easier for me than I first thought. All I have to do is run a short wire from the motor body to a convenient T nut or tapped bolt.
Ref the braided cable/drain wire, all I can say is bu**er me. My first printer was a RepRapPro Mendel (Tri Colour) which used the same braided cables as the Ormerod and when I built this one, I asked the question if I should wire my steppers in the same fashion. I was told that it wasn't necessary and that RepRapPro only did it for RF certification purposes.
-
Regarding shielded cable. As we have now many sensitive electronics in printers Z-sensors for example we should consider not only from RF interference to external devices point (certification) but also as a source of noise and interference for our own circuitry. So shielded cable is not a bad idea, if you have it then use it. But without it you most likely wouldn't have any problems.
