For reducing interference DC grounds are often connected to the metal cabinet with capacitors. The path of HF noise is often not what you think compared to DC. Sometimes also done with, as you suggest, a high value resistor in parallel with a capacitor.
But I agree. Following the rules is generally the best approach. John > -----Original Message----- > From: Chris Albertson [mailto:albertson.ch...@gmail.com] > Sent: July-14-18 12:16 PM > To: Enhanced Machine Controller (EMC) > Subject: Re: [Emc-users] Grounding > > From a safety point of view. It is best to follow rules from local safety > labs be it in the US or EU. Basically if you have a metal cabinet with AC > power inside The cabinet MUST be grounded to the earth pin on the power > socket and there are rules about using a dedicated screw (not used for > other purposes) and close to the power entrance. > > If you don't like the this then the other option is "double insulated" > design which mostly means a non-conductive cabinet made from perhaps > ABS > plastic. > > There is good reason to tie you DC ground to safety ground at one point, > even if you use a 1M resister to do this. Many times (especially with > vacuum tube circuits but in other cases too) there are leakage paths and a > ground left floating can float up to a high voltage. So many times we find > some kind of RC circuit used, but more commonly a simpl wire inside the > power supply. > > The idea is that is must remain safe even AFTER is major failure such as > insulation melting off an overloaded wire because of a fire. > > You will NEVER think of everything yourself, so just follow the rules. > They are the result of 100 years of world-wide experience. > > Most consumer products we buy today use double insulation. But if you are > using a metal external housing, double insulation is complex > > The best plan is to pretend you are actually going to SELL the machines you > are wiring up. Keep is "legal" in both the US and EU and you are OK. > There is a strong temptation to say "It is just my own toys in my own > house and I never plan to apply for a UL Lab sticker so way bother with > their rules?" > > > > On Sat, Jul 14, 2018 at 11:05 AM John Dammeyer <jo...@autoartisans.com> > wrote: > > > The power supply cabinet I'm putting together for my Mill has a terminal > > strip with ground and 105VDC power (1500W) with one pair of terminals > for > > each axis. From that strip I run one pair to each Servo Drive. From each > > Servo Drive a pair to the motor. The encoder cables enter the Servo Drives > > on a different side of the board and will be not be run in parallel with > > the > > motor cables. Stepper motor for the knee has its' own power supply and > it > > too runs separately (power and ground) to the Gecko. These high power > > supplies are all run with 220VAC through a separate contactor controlled > by > > the ESTOP. > > > > The PMDX-126, the 24VDC and the PC/Monitor are also on a 220VAC circuit > > that > > is live from the Mill Master Switch. The 24VDC supply is what I'd refer to > > as the instrument bus and it's used for a variety of controls like the > > control side of the STMBL Servo drive and 12V supply for the DRO.. It will > > share the DC ground with the PMDX-126 breakout board, PC USB and > ultimately > > a series of CAN bus based controls for tool changer etc. > > > > I wanted to also run the 24V into a 15V regulator to provide power for the > > control side of the HP-UHU servo drives but here is where that pesky > ground > > sneaks in again. Connecting a 15V supply sourced from the instrument > power > > that then runs the HP_UHU connects the Instrument power to the Servo > Power > > ground. The step/dir/error from the HP_UHU is optically isolated so I > > don't > > have to worry there. > > > > The Servo Power Transformer has a 12VAC winding which through a bridge > > gives > > me roughly 16V which is the better source for the 15V that not only is > > regulated down to 5V for the HP_UHU but also supplies the gate bias > voltage > > for the FETs. I'll only know if it works well enough once I put a scope on > > the running system. > > > > This is an example of where I'm not sure I see any value in ever connecting > > the two DC grounds together. As long as there isn't a return path for the > > high voltage DC through any part of the instrument bus DC circuit it > > doesn't > > matter if the high voltage DC ground became completely disconnected. > But > > connect both of those DC grounds to Earth (metal cabinet etc) and now the > > potential for high voltage, through component failure running through the > > instrument circuit, to instrument ground to earth back up into high voltage > > ground is more likely. And touching that high voltage now ungrounded > > section and perhaps the metal frame puts the body into the circuit. > > > > And yet if you have, say an audio amplifier or radio transmitter with both > > 48V and 5V that share a common ground a circuit failure usually smokes > > things. But those are all usually in a sealed metal cabinet. A system > > where the various devices are placed around the equipment, not all in a > box > > is different. > > > > John > > > > > > > > > -----Original Message----- > > > From: Jon Elson [mailto:el...@pico-systems.com] > > > Sent: July-14-18 8:57 AM > > > To: Enhanced Machine Controller (EMC) > > > Subject: Re: [Emc-users] Grounding > > > > > > On 07/14/2018 02:37 AM, Gene Heskett wrote: > > > > Not exactly 100% correct John. High frequencies are > > > > generally carried only by current flow at the skin of a > > > > conductor. And the large conductor, while having some rise > > > > in impedance, is still the better conductor because the > > > > smaller wire has far less surface skin area, raising both > > > > its impedance and its ohmic effects faster than the big wire. > > > There's also loops. Bundling the power supply and ground > > > return close together minimizes inductance. Having the > > > supply and return form a big loop increases inductance, and > > > can cause greater noise as well as radiation of EMI. So, in > > > general, you want the return wires to follow the supply > > > wires closely. > > > > > > Jon > > > > > > > > > > ---------------------------------------------------------------------------- > > -- > > > Check out the vibrant tech community on one of the world's most > > > engaging tech sites, Slashdot.org! http://sdm.link/slashdot > > > _______________________________________________ > > > Emc-users mailing list > > > Emc-users@lists.sourceforge.net > > > https://lists.sourceforge.net/lists/listinfo/emc-users > > > > > > > > ---------------------------------------------------------------------------- -- > > Check out the vibrant tech community on one of the world's most > > engaging tech sites, Slashdot.org! http://sdm.link/slashdot > > _______________________________________________ > > Emc-users mailing list > > Emc-users@lists.sourceforge.net > > https://lists.sourceforge.net/lists/listinfo/emc-users > > > > > -- > > Chris Albertson > Redondo Beach, California > ---------------------------------------------------------------------------- -- > Check out the vibrant tech community on one of the world's most > engaging tech sites, Slashdot.org! http://sdm.link/slashdot > _______________________________________________ > Emc-users mailing list > Emc-users@lists.sourceforge.net > https://lists.sourceforge.net/lists/listinfo/emc-users ------------------------------------------------------------------------------ Check out the vibrant tech community on one of the world's most engaging tech sites, Slashdot.org! http://sdm.link/slashdot _______________________________________________ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
