Aircraft grounding - Isolated grounds

[luissizzle]

Hello everyone, first post here.

Aircrafts use its metal chassis as Earth ground for several reasons, one of which is to provide a bleed path for ESD. I'm designing a PCB with the intention of putting it on an aircraft. This PCB uses DCDC isolators to create the rails I need, as well as providing ground isolation. However, there is no path to ground with isolated grounds to the chassis (obviously). I've heard that I can make high impedance connections between the grounds and chassis, but I am struggling to find any documentation (app notes/books) on that topic.

Can anyone provide some insight, or at least point me to some resources on this topic?

Thanks!!

[Manul]

Hello,

my question is, why your circuit is isolated/floating? Is there a need for that, or is it just accidental (because DC/DC converters happened to be isolated)?

[S. Petrukhin]

Hello everyone, first post here.

Aircrafts use its metal chassis as Earth ground for several reasons, one of which is to provide a bleed path for ESD. I'm designing a PCB with the intention of putting it on an aircraft. This PCB uses DCDC isolators to create the rails I need, as well as providing ground isolation. However, there is no path to ground with isolated grounds to the chassis (obviously). I've heard that I can make high impedance connections between the grounds and chassis, but I am struggling to find any documentation (app notes/books) on that topic.

Can anyone provide some insight, or at least point me to some resources on this topic?

Thanks!!

Usually 1-2 MOm is enough to not be strongly attached to the ground, but any charges could successfully drain into the ground.

[S. Petrukhin]

Hello,

my question is, why your circuit is isolated/floating? Is there a need for that, or is it just accidental (because DC/DC converters happened to be isolated)?

Or powered by batteries. 

[luissizzle]

Hello,

my question is, why your circuit is isolated/floating? Is there a need for that, or is it just accidental (because DC/DC converters happened to be isolated)?

For several reasons. Im using a 60V battery configuration, so I'm converting to lower voltages, there is a lot of switching the HV side I want to be isolated from, and i'm using COTS devices that specify a need for isolated supply for its serial comms.

[Manul]

For several reasons. Im using a 60V battery configuration, so I'm converting to lower voltages, there is a lot of switching the HV side I want to be isolated from, and i'm using COTS devices that specify a need for isolated supply for its serial comms.

I get the isolated comms part (that is normal and prefered), but I would consider to reference other circuitry to the common ground. I mean you can connect any point of isolated DC/DC converter output to ground. This is just my opinion, but floating the whole thing and then thinking about ESD and ESD bleed resistors seems like a wrong aproach to me. There could be problems with noise or other unexpected consequences. Don't know about avionics, but I saw quite a lot of naval tech and I don't remember seeing floating circuits often, except maybe some radar modulators.

[Manul]

Obviously it is your design, and you may know the whole picture better. Yet, floating circuit is made to allow the potential difference to exist. For example, there could be opto-isolated communication between two logic circuits, while there might be hundreds of volts potential difference between them. If your circuit is not deliberately driven to some potential above or below ground, then why not reference it to ground?

[luissizzle]

Obviously it is your design, and you may know the whole picture better. Yet, floating circuit is made to allow the potential difference to exist. For example, there could be opto-isolated communication between two logic circuits, while there might be hundreds of volts potential difference between them. If your circuit is not deliberately driven to some potential above or below ground, then why not reference it to ground?

Thanks for the feedback! If there is high current being pulled from the 60V supply, and lets say I have a solid ground plane on the bottom PCB layer, wouldn't my low voltage electronics, i.e MCU, be subject to ground bounce? This can probably be solved with clever PCB layout, but I don't see any issues with isolated grounds.

[Manul]

Thanks for the feedback! If there is high current being pulled from the 60V supply, and lets say I have a solid ground plane on the bottom PCB layer, wouldn't my low voltage electronics, i.e MCU, be subject to ground bounce? This can probably be solved with clever PCB layout, but I don't see any issues with isolated grounds.

As far as I understand your design, there should be no current going to ground. Grounding to the main external ground does not influence the possibility of ground bounce problems. I mean, you could have a truely isolated and floating circuit on a PCB and still have ground bounce as a local effect of your PCB layout.

Nothing clever, just do not run high currents and logic or analog supplies on a shared return path. You should never do it. But refercencing circuit to external potential ("ground") has nothing to do with that.

[CaptDon]

A little bit confused here, Most aircraft are either 12vdc battery buss
or 24vdc battery buss. Starter circuits drawing 400 amps at 24vdc are
somewhat common. For aircraft with bigger engines or turbines they are
still generally 24vdc with an A.P.U. to supply starter amperage or blow start
the turbines through the bleed port. All of the larger aircraft have 115/200
3 phase Y circuits with the chassis as the center point. The chassis carries
all imbalanced current from the 115/200 circuit as well as the battery return
current from landing lights and radios. I worked for a company that built
military electronics for all of the big name fixed wing and rotorcraft companies.
the secondary side of all of our supplies were 'chassis return' and not floating.
Any sensitive circuits needing noise abatement were run as differential pairs
to the various sensors with 80db or better of common mode rejection.
if you intend to make a 'marketable product' be prepared for over $100,000
worth of E.M.I. Radiated and Noise susceptibility testing in an accredited
lab!!! We spent millions getting our stuff certified for military use!!!
 

[Simon]

You don't have to have the isolation, just join the in and out grounds together. A high values resistor and a ceramic capacitor between the grounds will loosely bond your isolated side to the chassis ground to prevent it floating to some other voltage that can be a problem.

[Renate]

Since this thread is on aircraft grounding...

To ground an aircraft, disable the interlocks and raise the undercarriage.

An interesting read is Neville Shute's 1948 novel "No Highway".
It's based on the newly discovered problem of metal fatigue in aircraft, written by an aeronautical engineer.
It does involve the above scenario.

[Simon]

ooh, I vaguely remember reading that. I never realized that Neville Shute was an engineer. I really liked his books.