Thread Tools
Feb 20, 2021, 10:12 PM
Registered User
landru's Avatar
Thread OP
Discussion

FrSky R9 -- 27-second control lockout, near range, no failsafe -- What went wrong?


This afternoon, I experienced control lockout with a 2m electric glider flying on 900MHz. The model was no more than 100m away at about 70m altitude when the problem started. I was flying alone in an urban park beside a school.

When I couldn't correct a shallow arcing dive, I knew I was in trouble. Soon the model disappeared behind the 3- or 4-storey school and was lost to sight. I prepared for the worst. Against all odds, the glider emerged from behind the building a few seconds later, much lower, still uncontrollable. It then squeaked over power lines and past a telephone pole. Finally, the flaps deployed. Full flaps are a failsafe setting. Moments later, control returned.

Scary stuff. At the time, the lockout felt like an eternity. The attached log shows it was about 27 seconds. What the heck happened?

I've flown this model with the same FrSky QX7 ACCESS transmitter, R9M 2019 module, and R9MX receiver for over 70 hours without any problem. In that time, I've frequently flown beyond 400m and sometimes as far as about 800m. And I've often flown in the same location as today. The R9M 2019 is set up on the ACCESS protocol with 20dBm (100mW) power. That configuration hasn't changed.

For the past 5 hours of flying time, the model has also carried various 2.4GHz receivers piggyback as a means to log their performance relative to the R9 gear. In today's flight -- as with the previous few flights -- the piggyback receiver was a RX6R running on ACCESS and bound to the internal module in the QX7. The RX6R was set up on channels 9-16 and the R9MX on 1-8. The RX6R is connected to nothing other than power and a data logger. It should have no impact on the control of the model that I'm aware.

After I got the model back on the ground and my pulse started to drop, I did some range testing. With the model about six inches off the very wet grass, a 40m radius walk around the model showed only one problem spot, a weak signal in one very limited location, just a few paces in either direction. I hadn't noticed that dead spot in previous range tests. Regardless, control was retained until failsafe was forced by my pointing the transmitter away. The flaps lowered immediately.

Here are some possibilities that come to mind:
  • Interference at the flying site.
  • A previously unknown problem in the model set-up. For example, after coming home, I noticed that the internal and external modules have been assigned the same ID. That must have happened automatically. I hadn't noticed it before.
  • When the internal module in an ACCESS radio is active, the signal voltage to the external module is diminished. This issue has already been recognized. Perhaps the consequences can be severe?
  • I've messed something up in a way I don't recognize.
  • There's intermittent fault in the transmitter, module, or receiver.

Any thoughts much appreciated. If I've screwed up, I'd be very happy to know how. Meanwhile, I'm very hesitant to put this model back in the air. And I do emphasize, that my set-up is unusual.

Thanks,
Andrew
Last edited by landru; Feb 23, 2021 at 01:49 PM.
Sign up now
to remove ads between posts
Feb 20, 2021, 10:19 PM
Registered User
earlwb's Avatar
In the USA, 900mhz is a shared frequency band. Medical equipment and Amateur radio uses it as well as wireless telephones, public walkie talkies, FM repeaters (amateur radio) and some wireless networking too.
ref https://en.wikipedia.org/wiki/33-centimeter_band

Using 900 mhz means you have to be prepared to accept some interference though. I am not sure about what Canada does with it though. I suspect that you encountered some interference from something. Maybe the school or close by.
Last edited by earlwb; Feb 20, 2021 at 10:22 PM. Reason: typo
Feb 20, 2021, 10:24 PM
Registered User
landru's Avatar
Thread OP
Thanks. Interference could well be the culprit for all I know.

I had switched to 900MHz because the interference I was encountering on 2.4GHz was limiting reception quality and range, sometimes very severely. In fact, that was the reason behind trying to understand the relative quality of reception. Not looking good for either in my location.
Feb 21, 2021, 01:45 AM
Rick
Quote:
Originally Posted by landru
...[*]When the internal module in an ACCESS radio is active, the signal voltage to the external module is diminished. This issue has already been recognized. Perhaps the consequences can be severe?...
I've heard about that but for some reason had the impression it affected the telemetry, not the control signal. If it affects control it sounds like a potential source for your problem. One wonders what the module would do if it can't read the signals from the motherboard. Stop transmitting, that would cause failsafe? Transmit corrupt data, that would likely make your plane spasm around? Transmit last good values? If the latter, that might explain the lack of failsafe, the rx was still getting good packets but that wouldn't explain why it led to failsafe eventually.

Is this only on ACCESS? It seems like it would need to be a motherboard issue but the motherboards are different by model. It's not just a QX7 issue is it? Maybe you could bypass it by using ACCST?

Sorry for more questions than help but I would like to know more about that myself.
Last edited by choochoo22; Feb 21, 2021 at 01:59 AM.
Feb 21, 2021, 02:18 AM
Registered User
landru's Avatar
Thread OP
Hey Rick, all I know is contained in the posts following my inquiry in the OpenTx 2.3 thread here:
https://www.rcgroups.com/forums/show...postcount=4479

Kilrah indicated that the problem consisted in "telemetry not being available becasue of shared / disabled signal connections between the modules." That explanation corresponds with your understanding.

Perhaps I'm looking at a double whammy:
  1. Telemetry was knocked out by the simultaneous operation of the internal and external modules (which might also explain why only the red LED was illuminated on the external module).
  2. Interference knocked out the control link for the duration of the lockout but went unrecognized as a loss of valid signal event (hence no failsafe).

If the latter is true, there would be no benefit to having a redundant system as backup for the R9. Nor would there be any benefit in having an autopilot with RTH capability. Neither would kick in, or could be relied upon, without failsafe occurring predictably.

I had thought that 900MHz was a safe alternative to 2.4GHz in my location. Today's incident has me rethinking. I'm not sure where to go from here. The possibility of losing control of a model for about half a minute is not workable.

And even at the near distances in today's flight, RSSI on the 2.4GHz system was constantly tickling the very low 30s. If it had been the primary control link and been lost due to interference, the R9 would have been a dubious backup.
Last edited by landru; Feb 21, 2021 at 02:45 AM.
Feb 21, 2021, 02:48 AM
Registered User
My immediate reaction to a long loss of control with no failsafe, followed by a recovery is, it's a power problem. Brownout!

What's your power source?
Feb 21, 2021, 03:01 AM
Registered User
landru's Avatar
Thread OP
The power source is the switching BEC on the CC ESC. Receiver power was logged at 5.1-5.3V, never lower. The onboard battery was a 3S 850mAh Lipo. I'd made two launches and flown for less than 15 minutes. This battery easily yields four launches and much longer flight times. About an hour and half is the longest I've flown on a pack. Even then, it came down above storage voltage. The lockout started during a stable glide when power drain should have been as low as it gets in flight.

Of course, there may have been some power anomaly during the lockout which wasn't recorded in the logs.

A power problem was one of my first thoughts, too.

Thanks for putting thought into the matter.
Feb 21, 2021, 03:51 AM
Registered User
.2 volt drop on a switching reg would bother me. I'd carefully examine the wiring from the bec to rx and/or swap things out.
Feb 21, 2021, 03:53 AM
Registered User
Quote:
Originally Posted by landru
The RX6R is connected to nothing other than power and a data logger
How about 2.4 link?
Feb 21, 2021, 04:41 AM
Registered User
As far as understand your description:
Connection was lost, but failsafe was NOT initiated immediatly (immediatly means 1 second after loss of control)

So i would say the root cause would have nothing to do with RF problems.
(this should have triggered failsafe)
My suggestions are power supply related/ connection problems.

Quote:
Originally Posted by landru
It then squeaked over power lines and past a telephone pole. Finally, the flaps deployed. Full flaps are a failsafe setting. Moments later, control returned.
The fact that failsafe occured in the very last moments could be related to recovering of Power supply but bad signal strenght due to reflections etc..

Then the model got better receiption and control could be regained.
Feb 21, 2021, 05:19 AM
Grow old disgracefully!
dickw's Avatar
I agree with other posters. Control loss without failsafe - immediately makes me think of power supply issues.
It just sounds like an intermittent open circuit or high resistance connection in the power supply.

Dick
Feb 21, 2021, 06:21 AM
Carbo (-nator) ;-)
With a power supply issue there is no failsafe when power recovers. R9 with 100mW will not suffer a failsafe in 100m distance even if you dig the RX 1 ft into the ground
This is definitely not a power supply issue.
Feb 21, 2021, 06:30 AM
Registered User
Quote:
Originally Posted by Allerhopp
With a power supply issue there is no failsafe when power recovers. R9 with 100mW will not suffer a failsafe in 100m distance even if you dig the RX 1 ft into the ground
This is definitely not a power supply issue.
So therefore this couldn't happen.

But apparently it did.

What's your better explanation?
Feb 21, 2021, 06:50 AM
Carbo (-nator) ;-)
Quote:
Originally Posted by lexm
So therefore this couldn't happen. But apparently it did. What's your better explanation?
Well, of course there was a failsafe, but this was not due to distance or altitude. I do not know the reason for this lockout and failsafe.
I would proceed in a process of elimination, interference is also unlikely, as we would have seen it in the VFR sensor. I would continue to fly using choochoo22's standalone LQ sensor to get more data if the issue happens again.
Feb 21, 2021, 09:15 AM
Build more, websurf less
FlyingW's Avatar
I would change the setup to use a dedicated regulator for the airborne receiver, servo and electronics. Something like a Castle BEC. I trust the BECs built into ESCs only for foamies.


Quick Reply
Message:

Thread Tools

Similar Threads
Category Thread Thread Starter Forum Replies Last Post
Discussion Please help - crash footage - what went wrong? lacoa1 Multirotor Drone Talk 0 Feb 24, 2016 07:58 AM
Discussion O.M.G what went wrong with my QR X350 Pro stupeo Beginner Multirotor Drones 13 Jan 18, 2016 03:57 PM
Discussion What went wrong zooom101 Multirotor Drone Talk 14 Nov 24, 2015 04:29 PM
Discussion El Faro, what went wrong? Libelle201B Life, The Universe, and Politics 23 Oct 10, 2015 09:44 PM
Discussion Failsafe - Naza Lite - What went wrong? bachynski Multirotor Drone Talk 0 Aug 21, 2014 08:54 PM