For years I've been using floating auxiliary power supply for A-B PLC I/O racks.

However some people prefer to use grounded PS- in this case it's -24VDC terminal is connected to the grounding bus. Some people claim that this is a "safer" type of installation (allows to detect +24VDC to ground shortage by burning a correspondent fuse, others simply talk about "good engineering practice".

I believe that both approaches are acceptable depending on application. Are there any circumstances, where "grounded" PS is preferable ?

Thx, Victor.

 

The problem with a floating system is that it may continue to run even with a primary to secondary short. However, the first person to touch a secondary conductor will provide that ground, possibly with lethal consequences.

Depending upon the circumstances and the electrical code of whatever country you are in, you may need additional protections in place to safely use a floating power supply. This many include using a shielded transformer in the power supply. A shielded transformer has a grounded "shield" conductive layer between the primary and secondary windings to ensure that a primary to secondary short will take out the primary fuse.

In other words, consult the appropriate laws, regulations, and standards in your country. Or, just ground the power supply and be done with it.

 

M.Griffin's advice is sound.

Many measurement and control systems have both floating parts and grounded parts. You can protect systems which are floating from being a potential danger to personnel in the event of faults, through the use of isolator modules. And you can protect grounded equipment in the system from ground loops and pick-up from surges with isolators too.

If you would like some detailed suggestions, contact us at Weidmuller, Richmond VA (alan.balcombe @ weidmueller.de)

 

I prefer a grounded power supply myself, but not for safety reasons, it's pretty hard to get plated on 24 DC.

If your supply is ungrounded and you get the positive grounded it would drive an analog input to around - 23 V, I doubt the analog input could handle that and give the correct reading. With a solid ground a short will blow a fuse or trip breaker making it quite easy to zero in on the problem.

I have seen systems with 2 24 V lamps in series with the center point grounded. The lamps normally just glow whereas a short will cause 1 lamp to light up.

Roy

 

In reply to Roy Matson: You said: "it's pretty hard to get plated on 24 DC".

If you have a primary to secondary short with an ungrounded secondary, you could have a voltage with respect to ground that is up to the primary voltage appearing on the secondary. The voltage across the secondary could still be 24V. The voltage from the secondary to ground though could be as high as whatever the primary is.

 

Hang on a minute. Surely the risks of a primary to secondary short are too small to be a serious concern. We all use dozens of this type of device every day: Plug pack / wall wort transformers for phone chargers, comms equipment, laptops, cordless power tools etc. Then there are the higher power devices such as car battery chargers, audio amplifiers and even arc welders. I have never even heard of a case of a primary to secondary short in any one of these devices. The risk of a primary to secondary short when using these devices is so minute that it doesn't even bear contemplation. Surely the risk of a short is even lower in a device built to industrial standards.
To summarise: I don't think that concern about primary to secondary shorts is a good reason for grounding DC power supplies.

 

That particular risk is small but it is only one of many hazards that grounding helps. And the consumer environment is very different from industrial sites. I have twice encountered issues where the control power scheme was raised to line voltage by a short or a miswire. And several more where it is suspected that an arc between the two destroyed electronics. Having one side of the control power grounded is much safer, especially in systems that have a lot of wire in pipe. Any problem that is fixed by floating your control power is not fixed. If things are done right, it should make no funtional difference whether the control power is grounded or not. Many types of devices can't be floated anyway.

Regards
cww

 

In reply to Nathan: As I mentioned in my first reply, whether or not an ungrounded power supply is considered a problem will depend on "the appropriate laws, regulations, and standards in your country".

I also mentioned that there are power supply designs which are specifically intended to deal with primary to secondary shorts. This may consists of a "shield" layer, or the core of the transformer may be designed such that the primary and secondary can't come in contact with each other and the coils have an extra insulator isolating them from the core.

As for whether this is ever a concern, I know of a large installation which failed electrical inspection over this issue, and every transformer in every power supply had to be replaced with a different type. It is possible to use ungrounded power supplies. There are however safety concerns that may have to be addressed in the system design, depending upon "the appropriate laws, regulations, and standards in your country".

You listed a number of consumer goods. However, without analysing each one of these, you couldn't say which ones may or may not have grounded power supplies and also what additional safety measures may be present in their designs.

 

Roy, could you, pls, elaborate on analog input going to (- 23 V) due to a + 24 VDC connection to the ground (instrument's case, for example) ? I experienced an issue with A-B 1771 AIN module going crazy when - 24 V (common) had been disconnected from one of input channels. This caused + 24 V potential on this channel, while A-B 1771 module cannot tolerate > 10 V difference between the channels.
Not sure I understand how + 24 VDC connection to the ground affects functioning of AIN module.

Thx,
Victor.

 

Indeed, we are talking about negligible risk.

Auxiliary DC power supply usually provides very reliable isolation between primary and secondary circuitry. Small trafos for auxiliary ~24 VAC power (widely used in HVAC commercial applications) utilize grounded "shielding" between primary and secondary coils as a standard.

Mis-wiring may happen, but in properly designed PLC cabinet with physical separation between electrical portion (~120 to 400 VAC depending on country) and control portion (usually 24 VDC) the risk is very low.

So I would not consider safety factor as a reason to prefer grounded 24 VDC power supply over floating.

 

Good reply Curt. I agree with everything you have said.

 

Bear in mind that wall mounted devices are double insulated or Class II devices which reduce the change or short circuits.

 

Victor,
The normal loop would go PS+24 --> transmitter Positive, transmitter Negative --> PLC Positive, PLC Negative --> PS -24 so the analog input sees a maximum of +5VDC in respect of Gnd.

Now if your supply is floating and you get a short +24 --> Gnd it's going to push the transmitter negative below Gnd potential so the card positive terminal is -23 to -19V and the card negative terminal will be -24V. This might be OK if the card is sharing the same power supply but normally it's not.

Assuming the analog input module has some sort of Op amp supplied with +/- 15VDC the input is way outside of that so the Op amp is likely saturated one way or another.

I know some inputs can handle that sort of thing but my guess is your AB module couldn't. I have seen something similar with other inputs. That's the main reason I like to ground the PS negative.

Regards
Roy

 

Not all power supplies are ground referenced on the output, so it is floating, like a battery. If the + terminal of your 24V supply hit ground potential, it would cause the - terminal of the supply to be at -24V potential, just like a battery.

I've had some noise issues when 24V power supplies were not earth referenced on their outputs. Probably was due to a parasitic ground loop on a device it was hooked to.

KEJR

 

Good explanation. I agree too.

 

For those that are still interested in this issue- go to ABs knowledgebase and search for 14088 - 1771-IFE "data jumping around" (Common Mode Voltage problem). it's a pretty good article relating to the non-isolated AB analog stuff for PLC5s.

 

Dear Kevin,

could you please e-mail me the subject article to kgb_5 [at] yahoo.com

Regards
Baskar