[akschu]

Group,

I'm working up a wiring diagram and am a little confused on wiring ground pins in my panel.

What I'm finding is that my EFIS (advanced flight systems) uses 3 wires to a serial interface (RX, TX, and GND) which tells me that RX and TX are referenced to the GND pin, but some of the avionics that I'm connecting it to don't have a dedicated GND reference for the serial port.

For example, if I connect my GTX327 to serial 2 on the AFS then I run pin 13 on the AFS (Serial 2 TX) to pin 19 on the GTX-327 (RS-232 IN 1). That makes sense, but what does it reference to? The GTX doesn't have a dedicated GND pin for RS-232:1 so it must be referencing pin 19 to power or chassis ground.

Perhaps my understanding is limited, but doesn't this mean that the serial interface will be referenced against the aircraft ground path between the two boxes? Is that a good enough ground to make a serial interface work?

Let me break this down into specific questions:

1. If the serial interface is referenced against chassis ground then should I run a ground strap between each box?

2. If I run a ground strap between each box then wouldn't that cause a ground loop since I have both a chassis and power ground between each box?

3. Will the AFS box reference it's serial interfaces against chassis or power ground or do I need to run a wire from serial 2 GND on the AFS to aircraft GND on the GTX-327.

4. Any books or documentation that can help me understand how avionics grounding works when each box has a different way of doing this?

Thanks,
schu

 

[Radomir]

The short answer is -- it works w/o connecting that ground wire..

 

[breister]

The short answer is -- it works w/o connecting that ground wire..

... in most cases provided both instruments are grounded to the same ground bus (not necessarily the same exact point, but the ground points should not have appreciable resistance between them).

There are reports of a few odd widgets which require the second wire, but I've never found one.

 

[w1curtis]

a little bit of knowledge

Group,

...
2. If I run a ground strap between each box then wouldn't that cause a ground loop since I have both a chassis and power ground between each box?
...

The misunderstanding of the concept of ground loops is an interesting one that I've see cause people to come up with some strange designs in homebuilds. Things like DUAL #2 AWG GROUND cables run the length of a METAL airplane to --"avoid ground loops." It is an obsession that has taken on a life of it's own.

In a metal airplane, chassis ground AND power ground should be the same. There is absolutely no good reason for them to be separate. This separation can actually aggravate ground issues than just the single ground plane.

The primary time that you need to be worried about ground loops is when dealing with analog audio connections.

 

[noelf]

In the ideal world...

...chassis would connect to chassis, power grounds would connect to power grounds, and signal grounds would only connect to signal grounds. In the real world, each equipment manufacturer interprets and chooses which is chassis, power, and/or signal. These references are made within each piece of equipment.

Sometimes, a chassis ground is purposely connected to power and/or signal ground. If a manufacturer has a line of electronic equipment that is intended to be interconnected, it will be an engineering decision as to how best to interconnect these items. By knowing the internal construction and requirements of each item, it is possible to interconnect each item without a dedicated "signal ground" wire / pin.

This may not be possible when multiple manufacturers of the different equipment are interconnected. One guy's dedicated signal ground may be someone else's power / chassis ground. If this is the case, then it is possible to unknowingly wire up the equipment in a "ground loop" configuration.

In the special case of serial (RS-232-x) interface signaling, the specifications do call out for a dedicated signal ground. However, as stated above, this signal ground could be defined by the different manufacturers as either the chassis ground, the power ground, or the dedicated signal ground pin.

If you are integrating all of your "stuff" yourself, this can be a try-it-and-see effort. Using a radio shop (like SteinAir) takes a lot of the user guess work out of the effort, as he has the experience of what works and what does not.

 

[647jc]

F.Y.I.
Some of the example schematics provided in the GTX327 installation manual show using pin #25 as a signal ground for the RS232 interface. The pinout diagram also lists pin #25 as a signal ground and pin #13 as the power ground. I used pin #25 as a signal ground and it works well for me, don't really know if the seperate signal ground is necessary but since the example schematic (figure C-2) showed using it I thought it would be best.

 

[TCONROY]

F.Y.I.
Some of the example schematics provided in the GTX327 installation manual show using pin #25 as a signal ground for the RS232 interface. The pinout diagram also lists pin #25 as a signal ground and pin #13 as the power ground. I used pin #25 as a signal ground and it works well for me, don't really know if the seperate signal ground is necessary but since the example schematic (figure C-2) showed using it I thought it would be best.

This is correct. If there is no RS-232 specific ground (like the GTX-327), connect the AFS signal ground to one of the several ground pins on the GTX-327 (pin 25). Usually I'll look ahead at these wire interconnects and plan the type of wire I want to use. In this case, I'd use 3-strand shielded. TX, RX, Gnd, and the shield gets tied into the power ground/chassis ground on one side only.

This has been my experience over the last several years of wiring up avionics and conforms to AC43.13.