What's new
Van's Air Force

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

Autopilot coupling with APRS...

AlexPeterson

Well Known Member
On the trip to Denver this past weekend, I had the autopilot on the whole way. It is a Century 2000, with altitude hold. When the air was smooth, I noticed that periodically the airplane would pitch up and down slightly, about like hitting very light chop. The period was perhaps 1/2 second slight nose up, then back to normal. When we got to some really smooth air, I further noticed that is was always when the second hand happened to be a :43. Hmmm. Turned the autopilot off, and back on. Again every minute at :43. Put the autopilot in attitude mode, no excursions. Put the A/P back in altitude hold, back at :43 again. Turned the aprs off, and no more motion. Turned it back on and off a few times, and nailed the culprit. It seems that the baro sensor in the A/P is getting a little coupling with the aprs output.

My transmitter and (Howell) antenna are in the wingtip, so the stray RF is going a ways. I'm sure Century never had to test our ham frequencies to certify their unit.

Interesting side note - the aprs must transmit at a specific gps clock position all the time, because, even when I removed power from the aprs and restarted it at arbitrary times, it would settle back into :43 after its initial power up transmission. When I compared the analog clock on the panel (from which the :43 came) to the Garmin's gps time, :43 correlated to something like :20 on gps time.
 
I'd try looping the power wires going to the APRS transmitter through a ferrite bead (out near the wingtip). I bet the noise is getting coupled through the power lines.

I'm sure glad my autopilot isn't one of those 'certified' jobbers. :)
 
Last edited:
I'd try looping the power wires going to the APRS transmitter through a ferrite bead (out near the wingtip). I bet the noise is getting coupled through the power lines.

I'm sure glad my autopilot isn't one of those 'certified' jobbers. :)

Kevin, good advice, I'll try that, it will be easy to do and verify. Might need to add some length to the wire, but no big deal. I used an obsolete switch on my panel to power it, and somewhere it is getting into the A/P box. Any thought as to how many times to wrap it through?
 
Yours must be Anal...

Alex,

I looked at a few of my tracks and my unit is not as regular as yours - unless the time reported on the web is different than the actual pulse from the unit. That might be it.

I wonder if running shielded wire to the tracker would help. That or your antenna is goofy :eek:

Update - your pulses are irregular on the web, too - must be how it is reported.....
 
Last edited:
Kevin, good advice, I'll try that, it will be easy to do and verify. Might need to add some length to the wire, but no big deal. I used an obsolete switch on my panel to power it, and somewhere it is getting into the A/P box. Any thought as to how many times to wrap it through?

I'd say loop the wire three times through the ferrite bead.

You might try PM'ing VHS here on the forums (the APRS guy) - he might have better advice.
 
Interesting side note - the aprs must transmit at a specific gps clock position all the time, because, even when I removed power from the aprs and restarted it at arbitrary times, it would settle back into :43 after its initial power up transmission.

This is disturbing. If this is the case, it might also be true that ALL of our airborne trackers are trying to transmit at PRECISELY the same time, perfectly synchronized by perfect GPS time. You can imagine the effect if you are flying with a couple of buddies and all of you are using a Byonics APRS tracker and NONE of your signals gets out because everybody is transmitting at the same time!:eek:

Much better to have the transmit time be a function of the startup time, then you have a chance at these being more random.

What say ye, Allen?
 
43 seconds

Unless you have enabled time slotting, the timer starts from the units' start up and counts in milliseconds. The timing comes from the clock oscillator on the Micro-Trak, not the GPS time. Time slotting uses GPS time, so if he selected that, I suppose a 43 time constant is possible. The overiding issue is why his autopilot is glitching, and starting with RF getting in the wiring is a good place to begin. I suggest running the Micro-Trak and the GPS from a pack of AA batteries to determine if it is a wire-conducted RF problem, or if the AP is just sensitive about high RF fields, in which case, turn the RF power down, or reposition the antenna.

Allen
VHS
 
This is disturbing. If this is the case, it might also be true that ALL of our airborne trackers are trying to transmit at PRECISELY the same time, perfectly synchronized by perfect GPS time. You can imagine the effect if you are flying with a couple of buddies and all of you are using a Byonics APRS tracker and NONE of your signals gets out because everybody is transmitting at the same time!:eek:

Much better to have the transmit time be a function of the startup time, then you have a chance at these being more random.

What say ye, Allen?

No need to be disturbed. I'm not Allen but I can say that my tracker doesn't always transmit at the same "clock time". Whenever SmartBeaconing kicks in the "clock" gets reset and subsequent beacons will occur at the user configured interval after the last SmartBeacon.

Matter of fact, I have set the interval to 75 seconds because I've found a few iGates that seem to have a 60 second buffer and will release a stored packet about the same time a fresh one hits the net if the interval is 60 seconds. This can trigger an error in the aprs.fi parsing routine because it sees two beacons less than 5 seconds apart. Using 75 second interval reduces the chance of beacons colliding at the network server and hopefully results in fewer error packets.
 
Last edited:
Unless you have enabled time slotting, the timer starts from the units' start up and counts in milliseconds. The timing comes from the clock oscillator on the Micro-Trak, not the GPS time. Time slotting uses GPS time, so if he selected that, I suppose a 43 time constant is possible. The overiding issue is why his autopilot is glitching, and starting with RF getting in the wiring is a good place to begin. I suggest running the Micro-Trak and the GPS from a pack of AA batteries to determine if it is a wire-conducted RF problem, or if the AP is just sensitive about high RF fields, in which case, turn the RF power down, or reposition the antenna.

Allen
VHS

Hmmm. Allen, here is my config:

aprs%20screen%20shot.JPG


It does not appear that I have time slotting active, yet the :43 (remember, this is on my airplane's sweep second hand) persisted even after I turned the aprs power off and on several times.

However, as you say, getting rid of the coupling is what is important. I will try the ferrite coil first, as it is the easiest. I will also transmit with a handheld directly below the a/p controller, just for curiosity. I will be surprised if it is the raw rf coming via the air from the wingtip. The transponder antenna is much more powerful, and is only a couple feet away.

I'll report back...
 
Power glitch

Alex,

The Micro-Trak sends its longest signal when its reset after repowering. It occured to me that you need to make double-dog sure that you for some reason don't have an intermittent power loss or brownout om the ships circuit causing a reset occuring every X number of seconds.

Allen
VHS
 
Alex,

The Micro-Trak sends its longest signal when its reset after repowering. It occured to me that you need to make double-dog sure that you for some reason don't have an intermittent power loss or brownout om the ships circuit causing a reset occuring every X number of seconds.

Allen
VHS

Voltage has been rock stable for almost 1100 hours in this plane. I'm sure that is not an issue. I'll do some more power cycling next time up, with deliberate turn on points relative to the ship's clock.
 
I did a little more testing, and was indeed able to get the transmissions to regularly be at various clock seconds. Don't know why on the long flight from Denver it always settled into the :43 time, I guess it will remain a mystery.

I did try transmitting a handheld com radio directly under the autopilot, no affect (obviously, the freq range is different). I'll report back when I try the choke. There are several wires in the wingtip that could be coupling, I'll start with the power wire for the transmitter.
 
Update

I finally got around to taking the wingtip off today. I put a small snap together ferrite choke on the 12v and ground wires just next to the APRS transmitter. I believe I had enough wire for about 3 or 4 wraps.

Problem solved. I wasn't able to find completely smooth air, so there could be some residual affect on the autopilot that was too subtle for me to notice today. It appears that the errant RF energy was coming back into the panel through the power line, which is a dedicated wire from the fuse panel out to the wingtip.
 
Power wire only wrapped around ferrite bead?

I am having the same issue as the original poster (OP) with my autopilot causing a momentary burble each time the APRS transmits. Would wrapping only the power wire a few times around the ferrite bead solve the problem or should I wrap the ground wire too? Thanks!
 
Last edited:
I am having the same issue as the original poster (OP) with my autopilot causing a momentary burble each time the APRS transmits. Would wrapping only the power wire a few times around the ferrite bead solve the problem or should I wrap the ground wire too? Thanks!

Answered my own question today. I have four wire wraps through the ferrite bead only with the power wire. It did not solve the problem. Hum?
 
Answered my own question today. I have four wire wraps through the ferrite bead only with the power wire. It did not solve the problem. Hum?

Mike, you wrapped the power wire to the aprs? Did you wrap the wire right next to the aprs in the wingtip? Also, how far back into the plane does the ground wire go? It should also be wrapped if it comes back into the plane. Four wraps is not a lot, either, but it did work in my case.
 
Mike, you wrapped the power wire to the aprs? Did you wrap the wire right next to the aprs in the wingtip? Also, how far back into the plane does the ground wire go? It should also be wrapped if it comes back into the plane. Four wraps is not a lot, either, but it did work in my case.

Alex, I only wrapped the power wire to the APRS at about 8" from the APRS unit that is mounted on the outboard right wing rib. The ground wire was not wrapped around the ferrite bead and runs all the way back to the B&C ground block on the firewall. I will wrap the ground wire as well and see if that helps. Can the wraps be on separate ferrite beads? The ones I have a too small for more that two wraps with both wires in the same one.
 
Mike, not sure about separate beads, but if yours only allow 4 wraps, guess I'd try a second one. You are trying to kill any signal coming back into the panel area, so you need to wrap the ground wire as well.
 
No joy

Just tested with the additional ferrite bead with wraps for the ground wire in addition to the power wire - no change in the burble when the APRS transmits.
 
Smae Issue

I have had this issue with the Autopilot banking right every time the APRS goes off every since the APRS was installed.

It also tends to interfere with the radio on 122.?? at times.

Never have solved the problem
 
Just tested with the additional ferrite bead with wraps for the ground wire in addition to the power wire - no change in the burble when the APRS transmits.

Mike, you may have to ferrite bead all wires running near the aprs, lights, etc. You might need to move the beads more inboard, since the signals from the aprs might be coupling inboard of the beads. ??
 
Back
Top