Grounding 0Vdc

[dasarmin]

A discussion has begun here about ?grounding? the 0Vdc. I remember this not being a good practice due to possible noise being induced by grounding the 0Vdc to what?s used as the AC ground. DC power supplies have separate outlets for + & - DC as well as a Ground Terminal.
Inside an enclosure you may find AC Power coming in, which in our case is usually 480Vac. You?ll have 3 ungrounded wires and 1 ground, which is landed on a ?ground bus? which is electrically ?earth ground or building steel?. There is usually a DC Power supply (480Vac / 24Vdc) for control voltage; the outputs +24Vdc and -24Vdc (0Vdc) are typically run to their own distribution terminals. The -24Vdc terminals are generally not jumpered to the ?ground bus?.
However, would this be the wrong thing to do?
Most of the DC power is distributed by I/O devices, which provide their own ?ground? terminals. Field devices like Proximity Switches, Transducers, etc terminate to the I/O module; if shield or ground is required the I/O module have a terminal strip for this connection. These I/O modules also have separate connections for +24Vdc and 0Vdc as well as Ground. Again, indicating the 0Vdc and Ground should remain separate.

 

[dereckbc]

Well I believe you are talking about Bipolar DC power supply where you have +24 Vdc, -24 Vdc, and reference return ground which you are calling 0 Vdc.

Think of the Reference Ground Return the same as the neutral in a single phase transformer or service where you have L1, L2, and N where N represents the 0 Vdc line. It is a current carrying load circuit conductor, or grounded circuit conductor. It is not a equipment ground to clear faults, it is a circuit conductor.

 

[dasarmin]

Grounding 0Vdc

In this case it's not bipolar, if you measured across the + & - terminals without grounding the "-" terminal you'd read 24Vdc. If you grounded the "-" terminal you'd still read 24Vdc between + & -. In the first case the 24Vdc would be "floating", in tne second case it would be referenced to same point as all other voltages. Which I believe is similar to an isolation transformer, 480V / 120V; is you measure across the output you'd get 120Vac, if you ground one leg you still get 120Vac but it's not "floating" anymore.

 

[justdavemamm]

It's a small pet peeve of mine when companies and people label the two terminals of a single voltage output as +24dc and -24dc. I would much rather see +24dc and 0V (or DC RTN) as when I see the former, I have to spend time looking around to make sure there is/isn't a 0V terminal too for a bipolar P/S.

That said, in most situations, the 0VDC terminal is to be electrically brought to the PE ground in exactly one place. Sometimes you do want to float the 0VDC, but cases for that are few and far between, like on the old SCR drives.

The 0V and Ground/PE on I/O modules should, of course, remain separate.

 

[mike_kilroy]

whew. what a diverse set of opinions here. the majority of industrial machinery we work with has not only the common of the 24vdc supplies grounded to earth ground, but also one side of any 120vac control voltages. Not sure where you guys are that you dont see this as 90% of all cases.

in no way does this defeat opto isolation of i/o in drives, controls, plcs etc.

 

[hurk27]

In most DC type circuitry we try not to parallel any AC grounding with and I/O signal grounding or shields, as this can produce AC noise into the DC, it is better to make a bond at only one point like at the source, and float the other end, many times shielding is only bonded at the input end at the equipment it serves, but if it is bi-directional then this is not always possible, but it is possible to keep it isolated from any AC grounding, if you keep in mind that any parallel path that may be shared with AC circuits such as the EGC that their may be AC current also on these paths how ever so small that can cause errors.

the easiest example is in audio systems, when they use the equipment EGC as a signal reference at both the sound board and stage at the amps or instruments, it produces a nice hum into the audio.

Some times it may not always be possible to keep it isolated as some equipment might reference the signal ground to the EGC in the supply to the equipment, this is where isolation such as optic coupling come into play.

all depends upon the design of the equipment you have to work with.

 

[petersonra]

whew. what a diverse set of opinions here. the majority of industrial machinery we work with has not only the common of the 24vdc supplies grounded to earth ground, but also one side of any 120vac control voltages. Not sure where you guys are that you dont see this as 90% of all cases.

in no way does this defeat opto isolation of i/o in drives, controls, plcs etc.

I would agree these days that most 120VAC control circuits are grounded. 24VDC is a mixed bag. I generally ground them because I personally do not think it makes any real difference and it makes troubleshooting slightly easier.

Having said that, the project I have sitting out in the shop right now has grounded 120VAC control, and grounded 24VDC control. Both started out as ungrounded, but the customer decided to ground them, despite his spec calling for them to be ungrounded.

 

[pfalcon]

A discussion has begun here about ?grounding? the 0Vdc.

Ground it.

I remember this not being a good practice due to possible noise being induced by grounding the 0Vdc to what?s used as the AC ground. DC power supplies have separate outlets for + & - DC as well as a Ground Terminal.

Noise is not an issue for most control circuits but paranoia often is.

Inside an enclosure you may find AC Power coming in, which in our case is usually 480Vac. You?ll have 3 ungrounded wires and 1 ground, which is landed on a ?ground bus? which is electrically ?earth ground or building steel?. There is usually a DC Power supply (480Vac / 24Vdc) for control voltage; the outputs +24Vdc and -24Vdc (0Vdc) are typically run to their own distribution terminals. The -24Vdc terminals are generally not jumpered to the ?ground bus?.
However, would this be the wrong thing to do?

Bond the 0VDC at the power supply to the ground bar. Run ONE supply conductor from the power supply to the distribution terminal strips. Never permit any other location on the 0VDC to be connected to ground.
If you have multiple 24VDC supplies. Bond from the ground bar to a distribution strip that feeds each of the 0VDC terminals on the power supplies with individual conductors. Then run individual conductors from each 0VDC to their own group of distribution terminals. Each power supply should be treated as an independent circuit with independent 0VDC.

Most of the DC power is distributed by I/O devices, which provide their own ?ground? terminals. Field devices like Proximity Switches, Transducers, etc terminate to the I/O module; if shield or ground is required the I/O module have a terminal strip for this connection. These I/O modules also have separate connections for +24Vdc and 0Vdc as well as Ground. Again, indicating the 0Vdc and Ground should remain separate.

Ground is not 0VDC. Keep them separate except at the bonding point. Both the 0VDC and Ground runs should have ONE and ONLY ONE path back to the source. Do not tie both ends of a shield to ground, just one end.

 

[mike_kilroy]

Ground it....
Noise is not an issue for most control circuits but paranoia often is......

seems a bit harsh, altho got truth to it.... When it comes to machines with servo drives and vfd drives and sensitive equipment like probes with mv outputs, 1Vpp sine encoders, etc, noise actually is quite often an issue. Enough so that I have been paid to travel from coast to coast to solve stubborn noise problems. Guess it has made me a bit of an expert at it (I WAS farther than 50 miles from home). My experience is having solved noise issues by grounding the common of 24vdc supplies and NEVER in 25+ years solving it by ungrounding it. Hence, ground it is good plan in my book.

 

[dereckbc]

In this case it's not bipolar, if you measured across the + & - terminals without grounding the "-" terminal you'd read 24Vdc. If you grounded the "-" terminal you'd still read 24Vdc between + & -.

Unless I am missing something here, that is impossible. Regardless if the 0 volt line is grounded or not, if you measure the dc voltage between the +24, and -24 volt terminals, you had better see 48 volts. If not fix it.

Taking it one step further regardless if the 0 volt line is floating or grounded, and you measure the voltage between the +24 volt and 0 volt terminals you had better measure + 24 Vdc, and vice versa when repeated on the -24 volt to 0 volt reference terminal you should read -24 Vdc

So what am I missing?

 

[pfalcon]

seems a bit harsh, altho got truth to it.... When it comes to machines with servo drives and vfd drives and sensitive equipment like probes with mv outputs, 1Vpp sine encoders, etc, noise actually is quite often an issue. Enough so that I have been paid to travel from coast to coast to solve stubborn noise problems. Guess it has made me a bit of an expert at it (I WAS farther than 50 miles from home). My experience is having solved noise issues by grounding the common of 24vdc supplies and NEVER in 25+ years solving it by ungrounding it. Hence, ground it is good plan in my book.

Saw a lot of people lifting grounds because they were told it would make the microphones/audio less noisy even when they didn't have a noise problem to fix. Even saw them try to bring that reasoning onto the shop floor. Sometimes paranoia makes us fix things that ain't broke to worse ends. I can see mixing VFD with mV sensors could be an issue. We run separate power supplies as well as grounding each. Though a good VFD these days shouldn't be passing noise onto the control circuit, the old ones did.