Grounding Options 2

[stockdam]

(I thought I'd posted the following but it appears not so I'll try again)


I've a question about grounding as follows.

I have a data logger that uses an isolated DC power supply. Hence the power return on the input to the DC/DC converter is isolated from the 0V on the other side.

There is a differential instrumentation amplifier being supplied by the isolated supply.

The amp is inside a metal box/chassis which is floating/isolated.

I have a screened signal cable coming into the amp. The screen is connected to the metal box. The screen isn't connected at the far end (sensor end).

The whole thing is inside an aircraft and the external power supply return is connected to the aircraft structure.

The sensor is say a wheatstone bridge that is referenced to the amplifier 0V and isolated from the aircraft ground.

So in effect I have three ground connections to worry about.

Main power return/aircraft body
Data logger Chassis and the screens
Internal amplifier 0V


What should I connect the Data logger chassis to.....remember that the signal cable shields are connected to the chassis?

Should I connect the amplifier 0V to the aircraft ground or to the chassis or leave it floating?

What if the amplifier chassis is connected to the aircraft ground (it is bolted to the aircraft frame)? I now have the signal floating with respect to the cable screen.

What if now the signal source is single ended and referenced to the aircraft ground? There will be no reference between the amplifier 0V and the signal ground and so the amplifier won't work.

If the signal source is differential but referenced to the external aircraft ground then even though the amplifier is floating then there's no return path to its 0V.

If the logger has several instrumentation amplifiers and some of the sensors are referenced to the aircraft ground then what's the best way of connecting the chassis/screens and the amplifier 0V?

 

[VEBill]

"...chassis which is floating/isolated."

The aircraft installation requirements that we work with have a requirement to ensure that all boxes are firmly grounded. They even conduct bonding checks to ensure that each box is connected to aircraft structure by less than 0.5 ohms.

Also, all boxes must be mounted such that they will stay attached in the event of a crash. This usually implies that the boxes are bolted into rack. Being metal, they end up being grounded via the mounting.

This is a separate issue from your actual question...

A differential amplifier should have a differential input (isolated from ground) and thus the chassis could be grounded. Can you confirm this detail?

We sometimes use double shielding. An outer shield that is grounded to aircraft struture, and an inner shield that is terminated to the box ground.

 

[IRstuff]

"There will be no reference between the amplifier 0V and the signal ground and so the amplifier won't work."

This is not strictly correct. A differential amplifier does not need a ground reference, at least, not within its operational common mode range.

Reference grounds are often imposed to ensure that circuits do not float away from each other, potentially causing damage from the neutralizing the common mode.

TTFN

FAQ731-376

 

[stockdam]

When I said that the chassis is floating, I meant that it isn't connected electrically to any point.

Yes when the chassis is fitted in the aircraft then it will usually be bolted to the airframe and that will ground it. It would also have an earth strap fitted which would electrically bond it to the airframe. That will then mean that the signal shields (connected to the chassis) will be at the airframe potential whereas the amplifier will be floating wrt to this point.

Yes the amplifier is a true differential one but it will have a reference to its own 0V (via bias resistors).


To answer IRstuff...........if the amplifier is floating wrt to the aircraft ground then if I connect a signal that is referenced to aircraft ground then as you say it will work providing there is no common mode voltage that exceeds the amplifier specs (the amp is floating so I would say that there will be no common mode voltage).

 

[IRstuff]

That may be the intent, but that is often not the result. Case in point, I was involved with a helicopter system where we had some stuff running directly from the 400-Hz, and some stuff running from 120AC. Since digital logic ran from 5V, everything was converted, and everyone expected the circuits, differential, to talk to each other...

NOT.

The 120AC was actually inverted from the DC-converted 400-Hz, resulting massive common modes during startup, which were equalized through the differential receivers. So, after about 3 pwoer cycles, we had to replace the receivers. We wound up strapping the circuit grounds together to bleed off the common mode, which was a no-no in our design books.

Nowadays, it's common to see, in the specifications, a requirement to provide "reference" ground from box to box.

TTFN

FAQ731-376

 

[stockdam]

IRstuff - are you talking about differential RS422 receivers?

What do you mean by "reference" ground from box to box.

 

[IRstuff]

Yes, TIA/EIA-422

Typically, EMI control rules say that you cannot run internal circuit grounds outside of a particular box, since that leads to the possibility of creating ground loops. However, in the case I described, breaking that specific rule was the only way to get both circuits to behave, and not kill the 422 receivers.

TTFN

FAQ731-376

 

[stockdam]

Thanks. We've had similar problems with differential RS422 receivers and we had to connect the internal 0V together between boxes as you said.