What's new
Van's Air Force

Don't miss anything! Register now for full access to the definitive RV support community.

Serial output to multiple avionics boxes without isolation.

akschu

akschu

Well Known Member
Patron
Group,

A friend of mine who is very good with electrics (ham, did electronics in the coast guard, does circuit level troubleshooting) was looking at my panel diagram and recommends against running a serial output from one device to several others without an isolator.

For example, the aviation serial output on the GNS430W is connected to the EFIS and Portable GPS.

Far and wide this is minimized by each device having multiple inputs/outputs, but in some cases there isn't enough serial ports and it doesn't seem reasonable to output the same information twice.

So here are the questions:

1. Is it acceptable to split serial outputs?
2. Is there an isolator I can buy?
3. What do certified aircraft use to get around this problem?

Thanks,
schu
 
This is one person at one company's opinion, but I've never heard of an RS-232 isolator, and we've never recommended one in any install. As far as we're concerned, RS-232 is specifically designed to be split a few times without any special work. We probably have thousands of installs with split RS-232 and have never heard of a problem.

Just to make sure, we are discussing splitting one transmit to multiple receivers. You can't combine multiple transmitters on one line.
 
I've never heard of such an isolator either, and know of hundreds of installations where a single output goes to multiple places, with products other than Dynon as well. We even do this on a certain high-flying aerospace plane I know of (on our internal laptop network), so I think it's been looked at by some pretty smart folks.

Paul
 
I was told by Garmin support that an RS-232 output on my GNS-480 can feed up to 3 devices. Not sure what (if anything) happens if you go for four...
 
Interesting subject.
We just completed a version of our COM extender (4 RS232 ports) where each port is fully isolated. Sadly, it will not go into production as it has a terribly high power consumption.

Anyway, to answer your question - your friend makes a good case and RS232 and other connections have been responsible for much trouble and even damage to systems. Here are the reasons:

a) Ground loops. Multiple ground connections with any sort of area inbetween make good loop antennas that can both radiate as well as receive RF. Isolating the grounds would elliminate any issue here.

b) Ground faults. The typical scenario here is a high power device (like a transmitter) having lost its main ground connection due to a wiring fault.
It will then use any other ground connection (including the one in a typical RS232 hookup) possibly inducing several amps of current into some other device that was never designed to handle this. Smoke is the usual result.

Easy Fix.
The typical cockpit has several items in close proximity and they are all supplied by the same source and share a common ground.
Thus it is quite possible to NOT wire any RS232 grounds at all. If the devices have true RS232 signals then the signals swing between about -10 to +10 volts and don't need an accurate ground reference.
The same goes for ARINC (differential signals) and CAN that is used on some systems. Use the ground to connect to any shields but only connect on one side.
That is about the best you can do.

Rainier
CEO MGL Avionics

akschu said:
Group,

A friend of mine who is very good with electrics (ham, did electronics in the coast guard, does circuit level troubleshooting) was looking at my panel diagram and recommends against running a serial output from one device to several others without an isolator.

For example, the aviation serial output on the GNS430W is connected to the EFIS and Portable GPS.

Far and wide this is minimized by each device having multiple inputs/outputs, but in some cases there isn't enough serial ports and it doesn't seem reasonable to output the same information twice.

So here are the questions:

1. Is it acceptable to split serial outputs?
2. Is there an isolator I can buy?
3. What do certified aircraft use to get around this problem?

Thanks,
schu
Click to expand...
 
If the devices have true RS232 signals then the signals swing between about -10 to +10 volts and don't need an accurate ground reference.
Click to expand...

Interesting...then what are the signals referenced to? To ask it a different way, +10 volts is 10 volts above what?
 
RS-232 is single ended, meaning it's referenced to the ground of the unit that is sending or receiving. Which is also why there must be some sort of ground commonality between two devices.
 
Not isolation but fault tolerant

You could put a 200ohm resistor in series with each output line. It's not isolation but it could save you from having three items lose the signal if one of them gets shorted out.
 
The Garmin 430 does have 4 separate RS232 output channels. If you're still in the design and wiring phase, and intend to drive two different devices (e.g. EFIS and an autopilot) it makes the most sense to go ahead and use two of the outputs as individual dedicated circuits and program the 430 serial config menu for whatever format is needed for each connected device.

Driving two RS232 inputs with one output shouldn't present any electrical problems however, in case the 430 is already mounted and getting access to the back panel connectors, and wiring a second RS232 connection would be too much of a hassle, and both receiving devices want the same GPS serial data format and baud rate.
 
Neal@F14 said:
The Garmin 430 does have 4 separate RS232 output channels. If you're still in the design and wiring phase, and intend to drive two different devices (e.g. EFIS and an autopilot) it makes the most sense to go ahead and use two of the outputs as individual dedicated circuits and program the 430 serial config menu for whatever format is needed for each connected device.
Click to expand...

Except....you can't really do that. Even though there are 4 outputs, you can only configure one of them to provide GPS serial aviation data out at a time. Don't ask me why, but that's the deal...you can use any one of those ports, just not 2 of them for the same thing. Most all of these devices want the same format, and the 430 only lets you use one of those ports to drive all of those devices who crave the same data!

Cheers,
Stein
 
Last edited:
No NMEA from a GNS530 / 430

Only "aviation" RS-232 out of the garmins, and yes, only one of the output can be selected to the same format at a time.
 
I think that the solution is contained within Ranier's comments.

Do NOT use a Serial "ground connect" between devices - instead connect ONLY the "+" wire from the transmitting device to the appropriate wire on the receiving devices, and then make sure all interconnected devices are securely grounded to the same ground bus. Then only if nothing else makes it work you can try connecting the second (dedicated ground) wire to the common ground bus - but I would still only try one at a time until it works and remove any unnecessary wires.

The amount of current generated by serial lines is small, and should not cause damage to any other unit provided they all share a common ground. Yes, Ranier and you other engineers - it is technically possible - but unless I hear of a rash of such occurrences I suspect that it is indeed a rare event.
 
SteinAir said:
Except....you can't really do that. Even though there are 4 outputs, you can only configure one of them to provide GPS serial NMEA or Aviation data out at a time. Don't ask me why, but that's the deal...you can use any one of those ports, just not 2 of them for the same thing. Most all of these devices want the same format, and the 430 only lets you use one of those ports to drive all of those devices who crave the same data!

Cheers,
Stein
Click to expand...

Wow, that's a bummer. I'm looking at the Garmin 400 series installation manual right now and it does not mention that it's crippled that way at all, in fact it clearly states "change the selectable RS-232 inputs and/of outputs to match that of the equipment installed in the aircraft", then goes on to list a table of supported input and output format types.

Kinda makes you wonder how they passed their FAA approval with inaccurate/misleading data in the installation manual.
 
Indeed...there are multiple outputs selectable in multiple formats, but like almost everything aviation the install manuals and related regs hardly tell the whole story. It's just one of those things you learn to deal with. As others have mentioned, just split off the single wire and you'll be good.

I only mentioned this specific issue because I noticed in your posts that you had/are working with various people on their electrical installations and I hoped you hadn't wired things assuming the multiple outputs would work. You'd have a lot of head scratchin later on!

Cheers,
Stein
 
Stein,
What's a good (ie recommended) way to split the signal up other than than direct splices? Would it be OK to fashion a terminal strip or mini-bus of some sort for this data stream so that you can add or subtract components that need the data?
 
SteinAir said:
I only mentioned this specific issue because I noticed in your posts that you had/are working with various people on their electrical installations and I hoped you hadn't wired things assuming the multiple outputs would work. You'd have a lot of head scratchin later on!
Click to expand...

Actually I had anticipated using multiple RS232 outputs in my current project, an RV-8 panel with a 430, Dynon HS34 D100 and D120 and a TruTrack DigiFlight-II autopilot. I had already wired up the harness to make use of the 430's RS232 OUT 1 & 2, but since all the harness wiring I've made up so far is still sitting all coiled up in a cardboard box, waiting on the panel itself to get painted before any of the wires actually get fixed into place, the timing is perfect to correct this. :D

On the RV-10 we just finished this summer, the entire Garmin radio stack was prewired before I got involved with the project, and the serial data line incorporated into its harness only connected to the TruTrak ADI Pilot. The radio stack was purchased and assembled before Dynon ever had all the on-screen HSI programming released and working well, I think the HSI display on the D100 was still in its infancy back in that timeframe mid-2006, so the radio harness wasn't even made to accommodate feeding the Dynon. Now that the newest D100 software has a really good working HSI in it, and the RV-10's owners would really like to have it connected to the GPS data, I was really dreading trying to gain access to the back end of the Garmin stack in the RV-10 because it's practically impossible now that everything is bolted into place. I do have much better physical access to the DB connector at the back of the autopilot, so it won't be too terrible to tap into the serial data line back there and run jumper over to the D100 now.
 
Neal@F14 said:
Actually I had anticipated using multiple RS232 outputs in my current project, an RV-8 panel with a 430, Dynon HS34 D100 and D120 and a TruTrack DigiFlight-II autopilot. I had already wired up the harness to make use of the 430's RS232 OUT 1 & 2, but since all the harness wiring I've made up so far is still sitting all coiled up in a cardboard box, waiting on the panel itself to get painted before any of the wires actually get fixed into place, the timing is perfect to correct this. :D
Click to expand...

The D100 & 120 are actually somewhat superfluous in this example. Wires will go FROM the 430 TO the HS34 and to the TruTrak; and one ARINC or Serial connection may (I'm not familiar with the 430) go from the HS34 to the 430 (to transmit Course information). All other information will be transmitted from the HS34 to both D100 & D120 via the DSAB twisted pairs.

So, your wiring will actually be quite simple. Check out the Dynon Wiki - 430/530 wiring page.
 
You must log in or register to reply here.
Back
Top