Ground loop nightmare, please help.

[missgeek]

So, we are installing this 24v gauge and in order for it to be instrinsicly safe we have to tie it directly to earth ground. Unfortunately its being installed in an industrial setting and the Earth ground is not always ideal. The gauge converts a data stream to a 4-20 signal and any interferance makes it lose its mind its very sensitive. My question is, is there anyway to filter the ground running to this instrument externally or does the model have to be redesigned (our engineer is kind of a jerk). I am getting very tired of these things messing up all the time. Any suggestions would be appreciated.

 

[missgeek]

I would just like to add that the instrument is tied to ground to eliminate any floating negative voltage which is apparently very hazardous. It should also be noted that little me has tried pounding ground rods directly beside other ground rods tied them together than ran a wire directly ground back to my system but to no avail (I'm not even sure if that's up to code) but at 3 am you will try anything. Oh and the system runs perfectly fine when connected to battery power So i am certain it is the ground.

 

[JimB]

It would be helpfull if you could tell us what this "gauge" actually is.

instrinsicly safe

As is explosive atmospheres?

The gauge converts a data stream to a 4-20 signal

A "gauge" usually measures a physical quantity, pressure, temperature etc.
Are you sure that this is not a data converter rather than a gauge?

(our engineer is kind of a jerk)

Are you sure about that? What is his speciality? What is his involvement with this activity?

little me has tried pounding ground rods directly beside other ground rods tied them together than ran a wire directly ground back to my system

Sounds like something that a "jerk" would do!

Oh and the system runs perfectly fine when connected to battery power So i am certain it is the ground.

Or, it could be the power supply?

More information required before any sensible suggestions can be made.

JimB

 

[missgeek]

1. Yes it is an explosive atmosphere.
2. The Gauge measures level in oil tanks.
3. The engineer is the designer of this model. However he doesn't see how things work in the field. When brought back to the lab these gauges will work fine. However, at 3 am when they go high I have to change them.
4. Sadly I am a lowly technician and at times of desperation might become a jerk.
5. I checked the system with just our components hooked up and without the isolated ground. The data does not come up with any errors. It's only when I hook up the ground. I thought the same thing about the power supply and adding a filter to it. I've spent hours testing even with my limited knowledge I am certain it is the ground.
The engineer keeps changing the software however it is only masking the fact that this Gauge is reading these errors and making it So that takes longer or more errors to actually mess up, but it still is.
I am just wondering if there is something I can do to filter the ground coming into these gauges externally. From my ground lug to the gauge itself.

 

[missgeek]

Oh and thank you very much for the reply.

 

[ChrisP58]

A ground loop implies connection to ground at more than one location. Does your intrinsically safe standard require more than one earth ground connection? It may, if redundancy is required.

Are you, or can you, run the lines through an IS safety barrier to provide the protection without multiple direct ground connections?

google "intrinsic safety zener barriers"

 

[missgeek]

From what I can see these gauges have these "instrinsicly tener barriers one for each output (not that I've torn apart the Gauge or anything). 3 zener diodes with a fuse and a resistor??? Yes they are there. As far as I can tell everything goes through them. Now that I googled it I believe they have also developed a Gauge that has shunt diodes along the data strings that run inside the tank (I'm really sorry if I'm wording this poorly. But these still have a limited shelf life. Something isn't right with this gauge. I'm not so worried about voltage the gauge itself emits. I'm worried about the feed that im getting from ground. Oh I'm So confused

 

[missgeek]

Please excuse my spelling errors/ grammer errors I'm on a very tiny phone trying to type.

 

[ChrisP58]

Intrinsic safety is a very specialized field. Knowing what to do, and how to do it and still meet safety the requirements, can be counter intuitive.

There are probably other forums out there with members who specialize int that topic. Not saying there aren't members here who may be knowledgeable on the subject, but no guarantees either.

But it may be better to call in a specialist to solve the problem. Remember that, if you screw up in some way, and your circuit fails and blows up the oil tank, lives could be lost.

 

[dr pepper]

You can get line conditioners for the task, and ones ex rated, you can also get galvanic signal isolators for 4-20ma, not sure if you can get them for hazordous areas you'd have to look.
Its a difficult situation if you have to connect the sender end to ground.
Could you connect the ground end at the sender only, and run it back to the display end, or does the display end connect to a plc or something.
And yes you need to provide a technical file to show that you have taken adequate and correct precautions on the installation, if theres an accident then someone will come knocking asking for it.
Its unacceptable that the supplier doesnt sort this out for you, they have a duty to provide an instrument fit and safe for purpose, so long as your using a guage that is meant for the kind of material your measuring.

 

[dougy83]

Hi, but have you thought about (or asked about) wiring the ground of the sensor back to the ground of the receiver/controller? See https://www.electro-tech-online.com/custompdfs/2012/10/avoiding-ground-loops.pdf

 

[JimB]

I would just like to add that the instrument is tied to ground to eliminate any floating negative voltage which is apparently very hazardous.

Reading between the lines, I assume that:
the gauge is powered by the 4-20mA signal loop,
the loop to the gauge is fed from the safe area through an IS Barrier,
the power to the whole instrument loop is 24v from the safe area power supply.

If my assumptions are correct, then your statement:

I would just like to add that the instrument is tied to ground to eliminate any floating negative voltage which is apparently very hazardous.

is the key to the problem.

Have a look at the link provided by Dougy83, this shows good and bad earthing practice.
Having the negative side of the loop connected to earth out at the tank sounds like a very bad idea.

From experience, I would expect the body of the gauge to be electrically connected to earth, somehow bonded to the tank or local earth bar, whatever you have.
But I would expect the electrics inside the gauge to be completely isolated from the case and earth.

JimB

 

[missgeek]

I first want to add that this gauge is newly developed and in my opinion should have remained a prototype. The are two parts to this gauge the main unit is inside the junction box and then a cable is run up to the tank where there is a bunch of[ sensors inside a string that measure magnetic signals but all outputs are from the main unit. The main device is powered by 24V the output to the string I'm not sure about I think it actually is outputing a very low current 5v, and the output to the display is 4-20ma. I wish I could quote on this phone. The hazardous side of the gauge in the junction box has to be tied directly to earth ground which is the same ground that runs up to the tank. It is grounded properly from my understanding of that schematic. It's just that your sending a dirty ground that may not have a zero potential to a delicate and now think of 5V a digital??? circuit. The idea of a power conditioner seems interesting. However, if I use the output ground of a power conditioner to tie in my components under safety standards is that still considered the same ground as the input?

 

[dr pepper]

Sounds like a bag of worms, maybe you should put the whole thing back in its box and return it, not because of the messing around but because of the uncertainty whether or not its gonna be safe.
A power conditioner has the same earth out as in, its just a link, the last one I looked at was a vdr a couple of caps and a torroid inductor, thinking more closely this probably isnt what your looking for.
You could try measuring the volatge with a meter from the tank earth and the control panel earth to see if there is a difference, if there is then you know thats the issue, but before you strap fat earth wires to everything to balance things out beware that there might be mandatory guidelines for such things on fuel tanks.
A proper survey on how things are powered, connected up and where cables run might be a good idea, the issues your having could be caused by substandard electrical installation, and its better to sort this out at source rather than work round it, there may be an underlining hazard.

 

[Bob Scott]

This problem sounds similar to that common with TV stations. You have remote broadcast video cameras in a sports stadium and the cameras get their AC power locally in the stadium. Video cables run to the TV station next door. The TV station has its own ground and this ground has a few volts of AC on it compared to the incoming video cables.

Station engineers use "humbucking coils" to get rid of the common mode AC "hum" on the incoming video. You can do the same by surrounding your multiconductor cable with a large ferrite bead of the type that you have seen (as a large lump) on computer monitor signal cables. If that fails, make your own humbucking coil by winding your cable through a large toroidal core or stack of cores several times. This makes a "common mode noise filter". If the noise is confined to low frequencies like AC power, you can use a stack of huge iron washers as the toroidal core.

This is an example of the coil TV stations use:
https://www.electro-tech-online.com/custompdfs/2012/11/HB90020Data20Sheet.pdf

 

[ChrisP58]

If you could put together a drawing of what's in the system, and how they're all connected together, it would probably help us understand what you're doing.

The drawing should be more on the order of a pictorial, rather than a schematic.

 

[JMW]

If you need an intrinsically safe gauge, then you are only permitted to "black box" the parts. No subs, no add-ons NADA. Sooo what to do? Take all of this info directly to the manufacturer. They will have the resources to sort this out. Never, ever try and second guess a licensed engineer, regardless of how wrong he may be. He has a license, you don't. If anyone ever gets hurt or something decides to go boom there will be a serious investigation. Guess who will take the blame? In fact, people on this forum could also be called out.
Ground loops are the result of current flow between multiple ground points in a system. A 4-20 ma system needs no grounds. The sensor may or may not be grounded. The control lines TWO not one should not be grounded and be of sufficient size and insulated as per local code. A 4-20 ma sensor can be read with an old Simpson 260, may not be accurate but different readings will be discernible. The actual meter is nothing more than Analog to Digital converter. 4 ma is chosen as either the high point or low point to distinguish it from 0 ma indicating a broken wire or defective meter.