Firmware update soft does not see remote compass

[Varso]

I am unable to calibrate my EDC-D10A to my D10A EFIS. I have sent the EFIS and the remote compass to Dynon for syncing the firmware and testing because of communication problems. They were returned to me without any explanations about their status.
I must assume that both were working fine, the syncing was done and the issue is with my installation although it would have been nice to get a confirmation it was done.  I re-installed both and it still not working.

My setup

EFIS-D10A Sn:5745
EDC-D10A Sn:9071 part:100323-000
OAT

Symptoms:
EFIS detect remote compass and OAT. Windows firmware update software finds EFIS but does not detect EDC.
EFIS displays accurate OAT readings.
EFIS displays a wrong starting heading, heading changes when turning than slowly returns to original heading. It seems locked with a heading of about 200 to 225
Lately tried 5 calibrations, all failed but 1. Even with a successful calibration, EFIS heading is not working. See previous line.

remarks:
I asked Dynon to sync EFIS and EDC firmwares as I knew it was wrong. They never told me if they did it or not. They might still be out of sync.
I tested all EDC to EFIS pins and they are OK meaning the right EFIS pins connect to the right EDC pins based on most current documentation.
I have a concern regarding the grounds (OAT and EDC grounds) Pins 1 and 8 connected but also soldered together outside of DB-9.
It looks like an original Dynon cable but I see I was cut and reconnected halfway by the previous owner for an unknown reason.


Questions:

Pins 1-2-5-6-7-8-9 of EDC DB-9 are used. OAT used 2,7,8 and EFIS sees OAT, which leave pins 1-5-6-9 to communicate with EFIS.  I have to assume that EDC receives power(6) and the ground(1) is also OK because OAT is powered and provides a reading to EFIS.  This leaves me with data A and data B pins (pins 9 and 5) for potential error. Which data cable does the EDC uses to send OAT readings to EFIS?

I read that some DB-9 pins changed over the years/product versions.  I am testing the pins using the most up-to-date documentation.  Based on my EFIS and EDC serial numbers, Am I correct to use the most recent pins layout or should I use an older pins layout?

Am I right to assume that, as the Windows firmware update software does not detect my EDC, there is a defect either with the EFIS, the EDC or the cable. The installation is not part of the equation at this point in time?

What would be my next test?

Thanks for your help

 

[Dynon]

I've found your RMA from Dec 2014 from your email address and serial numbers. I can confirm that we got both devices on the same version and everything was calibrated and brought to factory spec when it was here.

Although, if you're not able to detect it via the support program, something may be up. The fact that your OAT works means that there is definitely EDC to EFIS communications, though.

Is there any chance that there is something near the EDC that is magnetic or has a lot of power going through it? That would explain it "pulling" back to a constant heading. Does moving a magnet around the EDC-D10A change the heading?

To answer your questions: Both data lines are used for both OAT and heading information (it's a differential serial bus, and both instruments are carried over the same serial port). If there was a communication problem, you'd lose both heading and OAT.

There was a change between the original EDC-D10 and the EDC-D10A, but that was back in 2003. All EDC-D10As are the same. We added an extra ground pin to make OAT connection easier (eliminating the need for external splicing), but that's not going to be the issue here.

Yes, the fact that you're not able to see the EDC with the support program, but are able to get data to the EFIS, may indicate an issue with either the EDC or EFIS. Are you doing the support program interrogation with the same harness as things are connected together with in the aircraft?

 

[Varso]

Thanks for your quick and VERY informative reply.

Fortunately, I had to go to the airplane earlier today and did perform the tests you suggested. 

Here are the results:
With EDC disconnected, the EFIS internal compass is working but not accurate.  Normal behavior as it was never calibrated.
With EDC connected, OAT is displayed and heading is still locked.
I turned the airplane around and a handheld compass located next to EDC was showing variation and pointing in the general right direction. Was unable to verify precision as I would have needed a mirror to perform an accurate reading.
Moving a magnet around the EDC was not showing heading changes on EFIS.

Remarks:
I have some concern about the EDC location.  The fire extinguisher and 2 seatbelt braces are located between 12" to 14" of EDC. Everything else is aluminum.
In order to keep the EDC aligned with the EFIS, the EDC is mounted on an aluminum bracket but the 4 bolts are ferrous.
The harness I used to connect to PC is the same as used in the plane.  The harness has 1 DB25 and 2 DB9. 1 for the EDC and the other for serial communications.
This you will find interesting.... I had the opportunity to try a new EDC and it also displayed a locked heading.
I only tested the pins linking the EFIS to the EDC.  I never tested all the DB25 pins.

Questions:
If the OAT is working perfectly, Can I assume that the data A and data B pins are OK and do not need to be retested(ie. not reversed)?
In your lab, are you using the same setup as provided to customers (harness and software) to install the firmware? If yes, that would suggest something is wrong with my harness.
In your lab, does your factory specs include testing if EDC was providing different headings when rotated? I know one cannot test every features. I expect a no as answer.

My conclusion:
In my mind, The magnet test points to a defective EFIS, EDC or a bad harness. It is unlikely that 2 EDC are defectives. The EFIS and the EDC were also tested in your lab.  That leaves the harness but the OAT is working perfectly so I am not sure the harness is wrong.

My future test:
I will remove the EDC from its location and hold it in my hand while rotating it. I want to verify if heading lock is caused by poor location or by a defective harness. Will this test be useful has I will not maintain the alignment between the EFIS and EDC? Maybe just reposition the EDC and move around the magnet would be better.


Any suggestions?

 

[Dynon]

A bit stumped at the moment. That test would be interesting and wouldn't caues any damage or calibration issues (you can turn the EDC and EFIS independently, and actually you'll sort of get an idea for how the heading is driven over the very short term by the rotational sensors in the EFIS but over a few seconds and longer pull to the compass.

But if waving a magnet didn't produce any wiggle at all, I'm suspicious that its interference unless it's a positively/close field. Does waving a magnet around the EFIS itself when the EDC is ostensibly working and confirmed via OAT do anything? Also, does the OAT move/respond to acute temp change? Is there a white box around the OAT number?

 

[Varso]

Early next week, I will get to the plane and perform the moving the EDC test and test the OAT by rubbing an ice cube on the probe.

In response to your questions, I was moving the magnet without heading changes while the OAT was connected and reporting a temp.  As for the OAT, I never provoked a drastic temp variation but I know that the OAT has always been reporting sensible temp when ground weather was from 0F to 90F. I am 90% sure that there is not a white box around the OAT number.  I will confirm this next week.

I have a feeling that the 4 metal bolts and nuts holding the EDC might be the source of my problems. I will rule this out next week.

Hopefully, you will still be there to help me out.  I really appreciate your input.

PS. Just to be more accurate, when we change heading when flying, the EFIS heading changes but returns to its original/locked heading after about 2 secs. Technically it responds to the magnetic field variation but it goes back.  To me, it points to interference. However, the magnet trick did not provoke any heading movement. The magnet was a 1" magnet strip with a lifting strength of a few ounces. That is puzzling.

 

[Dynon]

The magnetometer sensors themselves measure heading well, but they don't respond as quickly as you do when turning your aircraft. So the short term response for heading is biased very much on the rate sensors (gyros) in the EFIS. Then over the coming seconds, it pulls to the magnetic sensors. When everything is calibrated well, everything agrees, and heading is always right. If there is something pulling on the magnetic sensors - or if there is a problem with the magnetometer - the mag sensors always want to go back to a certain heading. But when you turn, the gyros over-ride that temporarily. In short, what you're seeing is expected for either interference or a problem with the magnetometer.

 

[Varso]

Sorry for the delay, I could not go to the plane earlier due to a knee problem.

As we agreed, I did remove the ferrous screws holding the EDC. Turning a bit the EDC or moving a magnet around it did not change the heading on the EFIS.
The OAT does not have a white box around it.
The OAT temp is accurate and does vary when applying heat or cold to the probe.

I tested the pins 3 times with a ohmmeter and everything looks OK. If I understood your explanations, If the OAT works, the remote compass should work too as both are using the same data pins.

This is a recap of the symptoms so you do not need to reread the discussion.

EFIS and remote compass sent to Dynon for testing and came back calibrated and tested.
EFIS internal compass displays changes when rotating the airplane but is not accurate. Off by almost 180 in some directions. It was never calibrated
EFIS remote compass locked to a reading and moving a magnet around or rotating the remote compass does not alter the heading.
Repositioning the remote compass did not fix the locked heading.
Cannot certify that EFIS and remote compass pitch/roll/yaw are the same but the should be close. I used an electronic level to position them
OAT Works perfectly with no white box around the OAT number.
Windows firmware update software finds EFIS but does not detect EDC.

Are you aware of any wiring error resulting in the OAT working but not the EDC itself? keep in mind that the firmware software does not see the EDC but sees the EFIS.
What would be the symptoms if the pich was off?
What would be my next step? I am thinking about building a new cable just connecting the EFIS to the EDC (4 wires). The EFIS has an Internal battery. So it should work.

 

[Dynon]

Thanks for the recap. That helped!

What happens when you move a magnet around the EFIS itself? (I'm guessing also nothing, but I want to be sure)

I don't know of any wiring issues that could cause you to have comms with the OAT (and therefore the compass is fundamentally working) but not have a compass that at least reacts somewhat.

If pitch were VERY off (say 90 degrees, pointing straight up), you'd have issues, but they'd probably look more random than what you're seeing. Is your pitch drastically off?

The fact that your EDC can't be seen by the support program suggests that there is fundamentally something up with the comms (and probably the EDC itself)

I think we should probably have you get directly in touch with our support team at 425-402-0433 or support at dynonavionics dot com so they can do some more personal troubleshooting with you.

 

[Varso]

Well, I forgot to run the magnet around the EFIS itself. Will do it this weekend. However, when I change direction, the EFIS (internal compass) heading changes.

The pitch could have been off by few degrees max. Nothing drastic here.

I intend to build a basic cable between the EFIS and the EDC (4 wires) just to see if they communicate. If it still does not work, I will follow your advice and contact directly the support team.

Something I assumed since I got back the EFIS from your shop. I assumed the problem was with my setup as I specifically indicated in my RMA the EFIS was not communicating with the EDC. The EFIS was not talking to the EDC before the RMA and did not perform any better after. On your first post, you indicated you checked the RMA and the EFIS has been fully tested.

Now, you are getting me concerned. Is comms testing part of the standard testing performed on a returned EFIS?

 

[Dynon]

It is, though it's hard to tell from the log whether the pair was definitively tested together. Which shouldn't matter. They both definitely were at factory spec and able to talk and work when they left though. In other words, both your EFIS and your EDC were passing all of our factory tests, which means that they should both work. But something is clearly amiss in either the connection or in one of those pieces of hardware (most likely the EDC from everything in this thread so far)

 

[Varso]

Thanks for reassuring me. I have then to assume it has something to do with my setup. i will continue working on my new cable. Unfortunately, the local electronic store was out of DB25 female connector. I did the rest of the cable but still waiting for the DB25.

I indicated in one of my first post in this thread that I had the chance to try another brand new EDC with the same EFIS/cable/location in the plane and had a locked heading too. So I will assume for now that the EDC and EFIS are OK (based on Dynon's testing) and will go with there is something wrong with the existing cable or the EDC location /orientation.

Once this is ruled out, I will then call Dynon's support directly.

Many thanks for your help. Your input is really appreciated.

 

[Varso]

As indicated in my previous post, I finally have built a temporary 4 wires cable connecting the D10A to the remote compass. I used the internal battery for power.

Well it looks like the D10A displays a heading and the heading moves when I apply a magnet around it or turn slightly the EDC. I am optimistic my problem is solved but I will be sure after a calibration.
On the original cable, I have checked and rechecked the pins connecting the D10A to the EDC at least 5 times, I doubt I missed a wiring error so many times. I am wondering if a wiring error on any other pins could have caused bad communications between D10A and EDC even if the 4 wires were OK? Anyway, I will rebuild the original cable.

I do not declare my problem as solved but I am very optimistic.

 

[Varso]

I finally got the EFIS and the EDC working.  I removed the EFIS and the EDC from the plane, fixed them on a board and connected with a temp cable (4Ft).  I managed to calibrate the compass and upgrade the firmware to V5.6.

BUT if I replace the 4ft cable by a 16 ft one, I loose the heading. Both cables were 22AWG copper 4 solid wires non shielded.

I intend to replace the cable by a 20AWG and see how it goes.  I am shocked that 16 ft is too long. At 16 ft the EDC is detected but communications with the EDC are not working but it's OK with the OAT.

Any recommendations for cable size? I need about 16ft long

 

[Dynon]

What you have there should work. One thing - although shielding isn't usually an issue with the EDC, that is the recommendation from the manual:

"The EDC cable in the harness supplied by Dynon consists of 4 conductors, surrounded by a metal shield and white insulation. These 4 wires are terminated with crimped female D-sub pins wrapped in plastic tubing. If you are building your own cable, we recommend that you use shielded cable as well."

Now whether or not that would magically cure your comms problem isn't easy to answer.

 

[Varso]

The cable in the plane is shielded and comes from Dynon and it does not work too.
I found out that any 18 ft cables either from Dynon or built by me does not work properly and a 4 ft cable built by me works perfectly.

I suspect that the EFIS sends a weak signal to the EDC and distance matters.


Is there a test I can perform on the EFIS or the end of the cable (like voltmeter reading between power and ground) to check for weak signal?

I will try to move the EDC closer to the EFIS. Could you tell me how far from steel cables should the EDC be located? I read in a previous post something like 14 inches. There is no electrical power near, just steel cables.

Thanks for your help.

 

[Dynon]

Probably the best thing to do at this point is to get in touch with our support team directly at 425-402-0433 or support at dynonavionics dot com.