My heli has been in this configuration for quite some while now (apart from the firmware version being updated several times) and has flown very consistently and reliably for over a year. Today when I first powered it up it took much longer than normal to be ready to arm, so I unplugged the battery, waited a minute or so, and then plugged it back in. The heli eventually then gave me the ready-to-arm warble, and arming and startup went as normal. I flew mainly in AltHld mode with a brief period of PosHld (to demonstrate to a club-mate) without any issues until I tried to RTL at about six minutes. At that point the heli simply started flying in the wrong direction so I switched back to AltHld and flew it back to land manually. Looking at the log I see some message about GPS needed for RTL, yet the log also shows that number of satellites was always around 12 or so. A link to that flight’s log is below.
I then swapped the battery for a new one but, even though the GPS module indicated GPS fix, the heli refused to arm. Back home, with the heli connected to MP, I set the log_disarmed parameter to 1 to see if I could figure what the problem was, but MP showed 0 sats even though the GPS module indicated a fix, and no log was produced. I tried a couple of times, but each time no log was produced with the flight battery connected for about 30 seconds.
So, any indications in the flight log of what might be wrong please? The param file is also attached here:
In the log:
2024-02-18 08:42:36.390 Error: Subsys EKFCHECK ECode 2
means the GPS position is bad (high variance).
Maybe the GPS module is going bad or something is interfering with the signal.
Thanks Mark. Ok, I see that. It was shortly after I’d successfully been in PosHold.
Nothing has physically changed on the heli since last week’s good flights, and it can’t be motor intererence because it now won’t even arm. I think I have another GPS module I can lash up temporarily to check if that’s it. Is it quite common for GPS modules to fail?
I’ve had one fail, but that appeared to be a loss of sensitivity (only got a few sats), and it was quite a while ago. Swapping in a known good GPS would be a good test.
every electronic can (and will) fail. But very often these a electromechanical failure like bad or broken solderings or wirings.
This information lets me believe that the communication between FC and GPS is somewhere brocken. This could be on both sides or in between. So first check very carefully all connections.
Do you have a possibilty to check the signals with an oscilloscope?
Thanks guys, I’ll start a thorough investigation this afternoon UK time. But one thing that crossed my mind is, would a bad GPS or a bad connection cause the FC to not create any log?
Hi Allan, I am not sure as I am still no real Ardupilot user. But I think you can selected somewhere if the log start either before or after successfull arming. Without GPS the FC will mostly not arm.
Yes, yesterday I temporarily selected logging to start before successful arming, and got no log.
Also, I’ve just realised that the logging shows that the number of sats was still good even after the error message noted by Mark in his post above, and when the RTL failed.
Anyway, let’s wait until I’ve had a closer look at it theis afternoon, including temporarily substituting a different GPS unit.
First, I’ve just realised that pre-arm logs are being created, but they all have the same date and time 01-01-1970 00-00-00. Could that be an indication of where the fault lies?
This afternoon I managed to get the correct plug onto a RadioLink SE100 GPS unit, and it’s flashing a solid GPS fix when connected to the R15 MiniPix. MP showed a new compass in the Install Firmware > Compass page, and also showed that the previous compass was missing. I then selected the option to remove the missing compass, and re-booted the FC. It then showed one external compass LIS3MDL and one internal compass AK8963. I then tried to calibrate the compasses, but after several cycles of twisting and turning the heli only the internal one had completed, so I cancelled that operation and re-booted the FC. When I reconnected to MP I saw that it was showing three compasses – internal LIS3MDL, internal AK8963, and external HNC5883. Going through the calibration routine again, with all three compasses selected, the calibration was successfully completed in about a minute and a half.
All the time the heli appears to respond correctly to control inputs when the flight battery is connected. But still MP shows Sat Count Blend is 0.00.
The first pre-arm log file #8 is after I’d installed the ‘new’ GPS unit and calibrated three compasses.
The second pre-arm log file # 1 is with the original GPS unit still installed
I note that in each case logging of GPS is missing altogether.
Hi Allan,
The controller has no realtime clock so without a external time information it starts always with a default value in the past.
During your tests was the RadioLink SE100 GPS instead or additional the mRo GPS connected?
Does also the RadioLink SE100 GPS itself indicates a fix but this is still not shown on MP?
Interesting is the changing number of compasses. This looked a little bit that the controller has some trouble with its ports.
Has your mRo15 more than one GPS ports. What happens if you change the ports.
Thanks Juergen. The RadioLink GPS was plugged into the original GPS slot in place of the original mRo unit. It very quickly inidcates a fix, as indicated by flashing green leds in addition to the green power led. The mRo R15 only has one GPS port. Just like with the mRo GPS, MP shows 0 satellites when the RadioLink GPS is connected, and the log file includes no category for GPS…
I’ve realised now that I can test the mRo GPS by connecting it to a RadioLink MiniPix on the bench, so I’ll do that tomorrow.
Hi Allan,
this is not good.
From my point of view, there are two reasons:
1 - the settings for the GPS port of the mRo R15 are not correct, but I am relativ sure you checked all this.
2. - your mRo R15 has a HW-failure.
Are you able to make measurements on the board with an oszilloscope. In that case you can make measurements directly on the soldered Connector pins of the GPS port.
Are you able to solder SMD components with a small iron. If yes you can try to heat and resolder the connector pins.
Also you can check the pins of the controller itself. For GPS the USART4 TX (Pin 23 PA0) and RX (Pin 24 PA1) are used.
I would discount the first possibility, unless the settings can change spontaneously mid-flight. From the initial log you can see that the heli armed and flew correctly, including a brief PosHold period, until the error message noted by Mark was triggered.
I am leaning towards the second possibility, but I don’t have an oscilloscope to do any diagnostics. I have experience of solding small SMD components, and might give it a go if my test of the GPS unit this afternoon shows that it’s working correctly.
A quick test this afternoon has shown that both GPS modules work correctly when plugged into a RadioLink Mini Pix, so the fault has to be somewhere in the mRo R15 Mini Pix.
That’s rather disappointing, for I’ve just remembered that my first mRo R15 was dead on arrival and had to be replaced. Its replacement now is going to be a Matek H743 WING, which is what I’ve got in my other two helis
This and the pre-arm logs indicate that there may be a problem with the serial port the GPS is on. It’s more likely to be a physical problem (bad connector?) than an internal failure.
You’re probably right Mark. But using a large illuminated magnifying glass I can’t see any physical problem around the GPS connector on the board and, despite many years’ soldering experience, I don’t feel qualified to trouble-shoot this multi-layered board.
Cost of postage to US is not going to be much, given the small size, so I think I’ll send it to mRo so that they can tell us what’s wrong with it.
There’s a couple of spare ports labelled for telemetry, each with Tx, Rx, CTS, RTS pins as well as + and GND. Are they serial ports? If so, how would I use them for the GPS module? Presumably I’d have to make some mods to the param file?