FC / RX Redundancy

[remyd]

What are folks doing to try and improve redundancy?

I recently had an X8, 16 inch props, castle ESC's Wookong M (latest FW and upgraded IMU) go silent at a height of about 50m and flip while in GPS hold with a 20K camera on board. I am still hoping that the camera can be repaired but ...... Surprisingly i only lost three props and didn't break a boom or leg (soft ground so to speak). I had googles on at the time while framing an image so i missed the flip just heard the silence and by the time goggles were off i saw the inverted copter about 10-15m above the ground

I still haven't got to the bottom of the crash but things are pointing toward a IMU reset in mid flight (unfortunately on this machine i didn't have the OSD logging) .
From what i have been told by a DJI distributor the controller can do this if the PMU doesn't supply the correct voltage to the FC (brown out), likewise to the RX and of course if there is too much vibration (white lights) - this machine was solid (no white lights) but relatively young 10-20hrs unlike some of my others which have closer to a 100. The distributor also said that the FC takes 4-5s to reboot in GPS or Atti mode but if flicked back to manual the reboot is less than 1s nice to know now but unverified by me.

This has got me thinking about adding a bit of redundancy (yes chutes for what they may be worth) but i am also thinking more about the electronics. On my turbine running a AceOne i use a Powerbox (dual regulators and batteries to supply the FC). The power is preety clean as tested with a CRO.

What are people doing on multicopters? I know of things like Fromeco DCUP or TCAP's by Futaba and others that are essentially capacitors used to smooth voltage and supply voltage if brownouts occur. Do people supply secondary regulated 5V power to their FC/RX using secondary regulators as a back up ? or alternatives like Scorpion Backup Guard 5v battery?

I know you can never win a redundancy argument because there are so many points of failure but i am interested to see what people with high value payloads are doing - if anything?

I would also be interested if anyone has had all motors just shut off mid flight and what they narrowed the cause down too. Note i was about 3min in and batteries still had 80% charge upon recovery (running two 11Ah 6S MaxAmps).

I am now more FC wary than i was before but it seems they all have their issues. I use the GPS and waypoint functions so that makes it harder to switch to other systems with less features.

Cheers,
Remy

 

[remyd]

Anyone?

 

[Benjamin Kenobi]

It's difficult to say. There are a few parachute systems out there but not many and largely untested by the general user. An X8 system is the best redundancy you can have.

I think the main reason no one answered was because you had goggles on whilst piloting a large and heavy aircraft, a big no no! In the UK that's illegal. You must maintain line of sight at all times. I would recommend using a monitor for framing as you can just glance at it.

Brownouts are rare and usually associated with a setup that's not perfect. Bad solder joint or loose wire are often the cause.

Hope you find the cause and the 20k camera is all right! :tennis:

 

[remyd]

It's difficult to say. There are a few parachute systems out there but not many and largely untested by the general user. An X8 system is the best redundancy you can have.

I think the main reason no one answered was because you had goggles on whilst piloting a large and heavy aircraft, a big no no! In the UK that's illegal. You must maintain line of sight at all times. I would recommend using a monitor for framing as you can just glance at it.

Brownouts are rare and usually associated with a setup that's not perfect. Bad solder joint or loose wire are often the cause.

Hope you find the cause and the 20k camera is all right! :tennis:

Thanks Benjamin - Note the goggles were only placed on for framing but this also coincided with the power failing but i agree in ideal circumstances my spotter should have been doing this but..... there were reasons and we were in the middle of a large agricultural paddock in GPS hold.

According to a local DJI dealer he thinks a lot of the flips are a result of brownouts, esc failure etc. I know on high value RC planes and heli's there are lots of dual power, regulator systems to feed the RX's etc.

I am interested to know if others are doing this in the multirotor world. Reducing risk is what i am after on the power side and yes fingers crossed on the camera it is on its way back to the states for assessment.

Cheers,
Remy

 

[leofuica]

Hi, Zero UAV have a Geminis 2, a redundancy system with two FC and two GPS at the same time and interconnected, with para chute and others systems, keep in eye and mind in the future....

 

[remyd]

Hi, Zero UAV have a Geminis 2, a redundancy system with two FC and two GPS at the same time and interconnected, with para chute and others systems, keep in eye and mind in the future....

Yes thank you saw those but there seems even less information about their hardware, control loops etc than DJI. I wonder which unit makes the decisions and what the indicators are for a change to the other unit.

My real question is how have people protected for power failure or brownouts on the flight controller. The rx is easy but i am looking for info about providing redundant power to the flight controller in case there is an issue. I will have to pull the unit off and do some tests.

The real issue is why would DJI ever allow the FC to reset while in flight or is this a myth. Certainly i have been told the same story about IMU resetting when they get to much garbage info from the IMU (excessive vibration) along with dirty power through their PMU (regulator) by a number of distributors. It seems crazy! Wouldn't you have it default to a control loop that would indicate an issue but still allow you to land so you could reset? Not saying this is what happened in my case.

 

[Bartman]

Remy,

I'm not an electrical engineer but I'd guess a small battery pack with a diode on the output might be able to kick in if the main battery supplying the FC were to weaken. You might already be losing power to the motors in this case but at least the FC would remain strong. You'd only need 800 or 1000 mah in the backup battery as it wouldn't normally be used.
I've also thought about using dual regulators with diodes on the output so that they can't push voltage into each other but can back each other up should one fail.

This issue of FC's browning out as the flight packs draw down is why I've been running a separate 3 cell pack on my flight controllers, completely independent from the flight batteries. There are issues with this if you are relying on an OSD that is being generated by the FC as you might want flight pack voltage instead of FC battery voltage but it depends on what your priorities are.

Good luck getting it straightened out.
Bart

 

[tahoejmfc]

Sorry to hear about your loss. I am currently building a X8, WKM, T-motors U7's, T-motors 17" top prop, 18" bottom prop, 2x10,000 6s, iosd mini, 1100mm diameter, tbs sony 69 fpv camera, immersion uhf. I too am concerned about a brown out from voltage ripple or voltage drop under heavy amp load. This is one of the reasons we are designing this copter with 2 batteries handling 4 of the esc's each with a 10 gauge interconnecting wire running through middle of frame from battery to battery. we are trying to not extend the esc input wire and keep it as close to the battery as possible to minimize voltage ripple.
Bartman is correct with the iosd mini not being able to monitor the correct voltage if you run an independent FC battery. you could also run into ground fault issues running two batteries, fpv gear, cameras, etc.

Have you had a chance to run a ecalc again on your setup and check your Amp draw as well as maybe with a actual amp meter and see what your drawing in C. you might have had a brown out due to the battery C rating being run too close to its cutoff. The reduce this chance and strain on your batterys, can you double them up to cut the Amps draw in half on each battery?
What is your Take off weight?

 

[remyd]

Yes no issues with amp draw (running 2 * 6S 11000 40C (listed) but probably only 8C after some digging - Maxamps) and motors only drawing between 5-8 Amps (castle esc logs) anyway and the ESC seemed to suggest the system still had power. In your build i would suggest going to the full OSD as it has data logging. I wish i had hooked mine up so as i had more info to investigate. Take off weight would have only been 7-8kg but i built this machine for bigger payloads. It was probably a little top heavy (lighter camera than i had intended and heavy batteries mounted on top) so the FC would have had to work to keep it stable even with the right COG settings. My next build will be different! I have purchased a backup battery for the RX (Scorpion back up guard) which cuts in if power is lost. I have also found a DC-UP by Fromeco which appears to take care of the voltage spiking issues. I need to see if the Battery guard can also be used on the FC and check how to create that redundancy with the FC eg. will it simply plug into a spare channel on the FC to solve the issue as it would on a RX or is the FC not wired to take the power input - more checking with very little info available about the FC internals. I am still trying to make decisions around a parachute but i won't fly expensive gear (on a multirotor) without one again (fruity chutes, MARS etc). There are still obvious issues if the copter flips and the chute is facing the wrong way. After a lot of reading i think DJI needs to look at their code. There are far too many incidents and they have been happening right from the start of their evolution in both their heli and multirotor products. Many seem to be related to GPS mode. That said the ones related to ESC/motor failure are nothing to do with the FC and are simple aerodynamics and gravity. There is also build errors and a lack of understanding that we are all guilty of. As i said in an earlier post there needs to be a backup control loop that if the FC decides it isn't happy (GPS poor, IMU vibration etc) it can default to something that isn't going to cause a reset or flip or ....... it may not be pretty but...... if it can get you back to the ground or at least give you warning and time to get back to the ground. The other course of action is to try and flip back to manual but in most cases you have only seconds to do this! It is probably worth practicing/simulating this in good conditions to build up your emergency procedure skills - he says now . I have looked at most of the other controllers and they all seem to have their own idiosyncrasy. I need GPS hold to frame areas as i am often looking at Nadir trying to get the best resolution. I am yet to rebuild and will continue to do more reading both on different FC's and current issues. There are many who have not had problems, myself included before this incident so i haven't given up completely but i have a couple of really heavy lifter 25kg copters planned and i need these to be SAFE! I need to gain some confidence in the controller again. Given the A2 issues and other issues recently i am happy to do more testing and then decide which way i will go.

Still happy to hear anything related to this topic!

Good news camera repair costs were less than 1K

 

[davide]

What are folks doing to try and improve redundancy?

Hello remyd,
I had exactly your problem last week (see http://www.rccanada.ca/rccforum/showthread.php?p=1863433&posted=1#post1863433)
So I started my investigation but I don't know what really happened.

However I'm also searching for redundant system and MikroKopter (http://www.mikrokopter.de/en/home) seems to be a very interesting company with redudant flycontroller products.

Best Regards

 

[remyd]

Thanks Davide

 

[screwdriver]

since 2014 in austria everything on your copter has to be redundant, also the FC. so if you want to fly and take photos or film legaly you also have to have 2 seperate flightcontrollers. there are some guys that built a kind of circuitboard that recognises if there is something abnormal and switches to the second FC.

 

[Carapau]

I think one of the reason this happened are the Maxamp batteries. Their 11Ah batteries may say that their C rating is 40C or whatever it is but in actual fact they are more like 4C and thus they can have trouble powering larger rigs. How do I know this is the case? When I first bought my Skyjib it had the Maxamp batteries yet the low voltage alarm was kicking in after less than 30 seconds of flight. So, I got in touch directly with Maxamps to be told that the 40C rating was purely its peak rating and not its constant and so I asked what the constant was and I got the reply that it was about 4C. 4C is not close to being powerful enough for a heavy multirotor. Bin the Maxamps and get something with a minimum of a 20C constant rating.

 

[Ronan]

Get some TATTU's for heavy lifting.

 

[remyd]

I think one of the reason this happened are the Maxamp batteries. Their 11Ah batteries may say that their C rating is 40C or whatever it is but in actual fact they are more like 4C and thus they can have trouble powering larger rigs. How do I know this is the case? When I first bought my Skyjib it had the Maxamp batteries yet the low voltage alarm was kicking in after less than 30 seconds of flight. So, I got in touch directly with Maxamps to be told that the 40C rating was purely its peak rating and not its constant and so I asked what the constant was and I got the reply that it was about 4C. 4C is not close to being powerful enough for a heavy multirotor. Bin the Maxamps and get something with a minimum of a 20C constant rating.

Thanks i am aware of the Max Amps rating and in this case that wasn't the issue - i also run the Tattu.
Cheers,
Remy

 

[leofuica]

since 2014 in austria everything on your copter has to be redundant, also the FC. so if you want to fly and take photos or film legaly you also have to have 2 seperate flightcontrollers. there are some guys that built a kind of circuitboard that recognises if there is something abnormal and switches to the second FC.

Hi screwdriver: ¿how do that?

 

[gtranquilla]

Redundancy offers tremendous benefits and is worthy of consideration where the size and weight of the electronic components is minimal as well as the electrical power demand.
In the Industrial Electrical and Automation Engineering world there is a system for high risk applications called SIL - Safety Integrity Level and there are four levels. As you can see from the table below the reliability benefit of redundancy is exceptional.
But keep in mind that such benefits cannot be realized where the individual components are flawed..... e.g., flakey firmware etc.
View attachment 21341

 

[gtranquilla]

Redundancy offers tremendous benefits and is worthy of consideration where the size and weight of the electronic components is minimal as well as the electrical power demand.
In the Industrial Electrical and Automation Engineering world there is a system for high risk applications called SIL - Safety Integrity Level and there are four levels.
http://en.wikipedia.org/wiki/Safety_Integrity_Level

As you can see from the table below the reliability benefit of redundancy is exceptional.
But keep in mind that such benefits cannot be realized where the individual components are flawed..... e.g., flakey firmware etc.
View attachment 26061