Commoning DC Grounds?

[kharrisma]

My Ender-based printer has three power supplies: the original no-name 24V, a Meanwell 12V that I added to power the Noctua fans I replaced the noisy OEM fans with, and a Meanwell 5V mainly to power the Pi.  I also want to use this 5V supply to power a servo that's driving linkage holding the probe dock for a Klicky probe setup.  I have several different material print beds, and my plans include a heated enclosure; the Klicky will work for all of these without the temperature degradation of the BLTouch or inductive probes, and it requires no conductive surface to work.  BUT, the Ender is a bed-slinger, and can't go outside of the print envelope by any significant amount.  Found a 'swinging dock' setup someone came up with for a Switchwire, that needs a servo to swing the dock out to the printhead so the probe can be retrieved and replaced.

I'm using the NeoPixel socket for the PWM and ground, and getting 5V from the 5V PS.  Problem is, when I use the ground pin in the NeoPixel socket, nothing works.  DVM shows no continuity between the negatives of either the 12 or the 5 volt power supplies, so no wonder things aren't working!

I already tied together the AC neutral and grounds for the 3 power supplies, but never did the same for the DC.  Can anyone come up with a solid reason why this would not be a good idea?  Doing that would enable me to utilize the nearest available ground for anything I wanted to power, any unutilized ground (negative) pin on the board would work.  Or am I working toward letting all the smoke out?

[Repman]

Hi there,

as you did not write what your controller board is there is some speculation. You should check the schematic of the board first.

3 hours ago, kharrisma said:

I'm using the NeoPixel socket for the PWM and ground, and getting 5V from the 5V PS.  Problem is, when I use the ground pin in the NeoPixel socket, nothing works.  DVM shows no continuity between the negatives of either the 12 or the 5 volt power supplies, so no wonder things aren't working!

If the GND of the 3 pin Neopixel socket is not connected to board GND, then it could be switched by a mosfet. Therefore the continuity to board GND is not there. 

I would try to connect the servos GND to a board GND Pin (check continuity of your choosen pin for that to board GND first) and the servos VCC to a arbitrary +5V Pin (any 5V pin which is not switched). You can measure with a DVM if there is a 5V voltage between this pin and the Servos GND pin. If this is ok, your servo gets power! Than you need a free PWM pin of your choice to switch the servo. Check if the neopixel pin of your board is suitable for that (it should be switchable by PWW as output).

Regards

[kharrisma]

Hi Repman,

Problem is, the servo requires current well in excess of what the 5V bus of the card (SKR Mini E3 -v- 3) can provide, which is why I'm trying to power from the 5V PS and use the card for PWM and ground.  The NeoPixel socket has good 5V and ground, but the ground ain't working for the external 5V PS, which is why I thought commoning the grounds would help.  ((Later)) Took the chance and tied all the negatives from the three power supplies together and tested the servo - - works fine now.  Bit of fine-tuning and macro tweaking and I should be good to go!  Thanks for the input!

[GarthSnyder]

It's generally a good idea to wire DC grounds together. But do it properly: no component should have multiple routes to ground, or you risk creating ground loop problems that can be subtle and hard to diagnose. Here "to ground" means "to the junction point of the DC grounds." 

This is Kurt Vonnegut's drawing of an anus from Breakfast of Champions, but it's also his drawing of an ideal ground network. Now you'll never forget it! (In practice, the radiating wires can branch, but keep it a tree.)

[polyneutron]

I would use the earth ground from all the supplies to the frame of the printer. Then as others have mentioned the mother or main board specs. 

[psyki]

2 hours ago, GarthSnyder said:

This is Kurt Vonnegut's drawing of an anus from Breakfast of Champions, but it's also his drawing of an ideal ground network. 

I will definitely remember now ROFL

[therm_virtual0y]

Maybe a bigger 24vdc power supply (if space allows) and use buck converters for the 12v and 5v?  If you do tie the common (0vdc) together, do it in a star where all the grounds come to a single point like GarthSnyder said.  

side note: are you messing with the servo because it’s fun to tinker?  I’m that way too.  Are looking for a solution to bed leveling or first layer issues, and the servo is the next solution?  Just curious 

[Polyneutron21]

I personally use buck boosts but I would not want one inside my printer, I prefer proper PSU if there is room. 

[Penatr8tor]

17 hours ago, GarthSnyder said:

This is Kurt Vonnegut's drawing of an anus from Breakfast of Champions

Big Kurt fan here!

Good one @GarthSnyder!

[Repman]

Hi there,

20 hours ago, kharrisma said:

Problem is, the servo requires current well in excess of what the 5V bus of the card (SKR Mini E3 -v- 3) can provide

What kind of board is that, which has not the capability to drive one servo? Or how big and current hungry is your servo?

I think your solution with direct powering from the 5V PS is the better one anyway.

Regards

[kharrisma]

BTT SKR Mini E3 version 3.  It's not a big servo, an SG90 micro servo.  Not sure what the current requirements are, but I know it moves both abruptly and quickly, moves from point A to point B very quickly, which tells me that there's likely a relatively ferocious inrush currrent.  MANY cautions about trying to power a servo from the control card, so since I have an external 5V PS anyway, I figured why not use that and not run the risk of overloading the card?  

Didn't "star" the negative connections, more of a daisy chain setup; 5V to 12V, 12V to 24V.  Not sure about grounding to the frame as well.  Feels like that's more likely to introduce problems than prevent them.  I can see why one might want to ground the printer frame if using an AC powered bed heater, but the rest is low-voltage DC, little to worry about (from a personal safety viewpoint, anyway.)