[AAFJAPAN]

In short my latest build which is running Afro race spec 20amp esc's and Rctimer 2206 2100kv race motors on a revo nano with oneshot enabled, arms and as soon as I arm and give the minimal throttle input (should be barely spinning), it ramps up and up and up. If I have it set normally it does it extremely quickly the first couple times, and then settles to a quick, but less inane ramp up, but will keep ramping till it reaches very high RPM's. If I attempt to set a throttle curve, even all the way down as a flat line it will still ramp up, granted not to the extreme high RPM's, but also in both scenarios it seems I have no control to reduce the throttle. I haven't tried flying it yet because it seem simply wrong to me. Any thoughts and or solution's are very appreciated. This build has been a PITA through and through, but is finally done other than this hopefully minor issue.

[jbarchuk]

...minimal throttle input (should be barely spinning), it ramps up and up and up.
I haven't tried flying it yet because it seem simply wrong to me.

That technique of 'testing' anything is 'simply wrong' from the *FC's* point of view.

I'm assuming this is on the bench with no props attached, correct?

The FC when armed does pretty much nothing till given throttle.

As soon as the throttle is touched the FC's main job is to maintain a level attitude (in attitiude mode.) Or in rate mode change the vehicle's attitude to match the rate requested.

The FC does these things by increasing and decreasing the RPM of the *MOTORS*.

However with no *PROPS* mounted the vehicle can't move.

The FC thinks to itself... "I tried to [do something.] I varied the motor RPM but that seemed to have no effect on the vehicle's attitude. What to do?? APPLY MORE RPM!!"

Aaaand no matter how much RPM the FC applies to any of the motors it never experiences and joy.

These tests sitting on the bench (or held in hand) with no props has a couple of useful purposes, but not what you're doing. Or not what you think the FC should do.

It's good for making sure motors turn in the right direction.

Held in hand and tipping to different directions it's useful to make sure the correct motors speed up or slow down to get the desired attitude/motion change.

But those tests are good for only a few seconds. The FC pretty quickly realises that it doesn't -really- have control over the vehicle's attitude or motion, and in a fit of childish pique eventually spins all the motors up to max. Which after a minute or so should burn them out.

[AAFJAPAN]

OK, so on a hunch I put one of my X8R's on this and tried it with a Taranis, and the problem is solved. The problem is these are not being built for Taranis setups. I use DJI Phantom 2 Rx/Tx combo's which I bought about 10 of on the cheap for my current builds which I sell. Never had an issue before, but tried 3 separate Tx units and it's the same issue. I will try swapping RX units next, but I suspect it is a compatibility issue perhaps with oneshot ?? Any one have any ideas. The TX and RX are use the Futaba Fasst protocol for what it's worth.  I am stumped.

[AAFJAPAN]

The FC thinks to itself... "I tried to [do something.] I varied the motor RPM but that seemed to have no effect on the vehicle's attitude. What to do?? APPLY MORE RPM!!"

Aaaand no matter how much RPM the FC applies to any of the motors it never experiences and joy.

I appreciate the response, and this had crossed my mind, but this is my 5th Open pilot type build, and the first time I have experienced this. I built an almost identical build last week just different motors and an X8R RX instead of the Phantom 2 and it behaved as I expected with the ability to give minimal throttle input and it would maintain this indefinitely until I applied more throttle. I can always control the throttle very accurately on the bench with OP builds. What you are describing only makes sense if it thinks it is already at X ammount of altitude in the air in altitude hold or GPS type modes. When you arm on the ground it does not, or should not ramp itself up to get in the air, it should rely on user input, so on the bench should be the same.

 Regardless, it seems to be a TX issue as I posted apparently the same time as your post. Any thoughts?

[jbarchuk]

What you are describing only makes sense if it thinks it is already at X ammount of altitude in the air in altitude hold or GPS type modes.

Not at all the only way. A very common question/problem is 'I arm it and turn the throttle up just a little. The props spin up faster and faster by themselves and a few seconds later it flips over.' If in attitude mode the props that spin faster try to level the vehicle, but the props that spin slower can't 'lower' any because they're on the ground. Eventually there's enough RPM to lift (not the desired action but whatever) however with zero ROMs on the opposite motors it flips over.

But let's let that go for now and figure out the problem. About this line....

When you arm on the ground it does not, or should not ramp itself up to get in the air, it should rely on user input, so on the bench should be the same.

That implies that you did -not- give throttle input yet when you armed it the motors started up. But in your first post you said...

...as soon as I arm and give the minimal throttle input...

...where you did give throttle input. So the exact situation is unclear.

But here's another idea. This didn't occur to me before because I've never used it.

In Output do you have 'Motors spin at neutral output when armed and throttle below zero' ?

If yes then that's exactly what you're experiencing in 'When you arm on the ground it does not, or should not ramp itself up....'

[AAFJAPAN]

No, sorry let me clarify, when I arm the motors do not spin. Phantom Transmitters are a bit funny in that they have a sorta click mechanism to lock the at zero for arming. When you are ready to give throttle it pops up a bit and if you let go of the stick it will center at mid throttle. So the minimal throttle allowed on the phantom is what I was referring to when I said zero input. Confused? I am confused just typing it. Anyways so at that minimum amount of throttle input is where it ramps up and up and up. In case you missed it in my last post, I did switch the TX and RX to my frysky gear, and it behaves appropriately. I just took it out for a maiden in the back yard and it flew like a dream. The problem is this is not for me, it is to sell, so I need to get the phantom TX/RX combo to work. I have flown them on very similar machines and didn't have this problem. The only two variables I can see is that this machine has oneshot, and it is LibrePilot not open pilot. my gut says the second one shouldn't matter, but maybe there is some tiny piece of code misplaced. I am not a coder so really haven't a clue on that. Nor can I imagine why oneshot would make a difference. I am a builder and frame designer that knows a bit about the electronics my necessity, but am not especially gifted in that area.

[hwh]

I doubt anyone in the forum has a Phantom transmitter or receiver hooked up, myself included.  I take it we're talking about the little rectangular board with just an antenna (or two) and a 3 pin connector to the flight controller?  Not much info on them on the web, almost the only thing was that people buy phantoms and immediately replace them with other receivers 

A couple of posts say they're s.bus receivers, did you configure for sbus?   In the transmitter wizard did all the sticks calibrate and work correctly?

[jbarchuk]

I'm out of my depth here so can give only impressions of what various symptoms might mean. I've seen DJI Tx/Rx/FC stuff but never touched it.

One issue is deviating from the typical 'all DJI parts' chain of data and control. Where you said you changed out Tx/Rx and it behaved perfectly implies that the DJI stuff has a hard time working and playing well with a non-DJI FC. It implies that there's nothing at all wrong with the FC/ESCs whether OP or LP inside it.

DJI, or at least some of their stuff, uses telemetry. If using a non-DJI FC causes wacky things to happen that MIGHT imply that the Rx and even the Tx have some say in the behavior of the FC. With that, if the non-DJI FC doesn't return appropriate telemetry that the Rx/Tx wnat/expect/need, then the Rx and/or Tx have fits. Maybe it tosses the FC into an RTH mode which is why the motors wind up and try to make it fly even if there're no props. I have no clue just guesses.

[AAFJAPAN]

Hey guys, in short, yes it's S-bus, After the Phantom 2 or I think something like Phantom2 V2 (forgot the actual name) they changed to 5.8ghz and it all became proprietary. This TX RX combo actually is pretty decent. I have run them on CC3D boards, and they performed really nicely. Of course it lacks the features of the Taranis, but it had no control issues. Very nice for backpacking a mini quad all over since the TX is insanely light.

 I have to head out soon, but I am going to pull another DJI RX out of the box and run through the setup wizard one more time. If that doesn't solve it, after work I will go though some other steps like doing a new CC3D setup with librepilot, if the CC3D board has the problem, it may very well be Librepilot, or perhaps FW related. I do have one of my older CC3D machines still hanging on the wall which is bound with the Phantom TX/RX combo, so I will check which FW it is running. It was setup using OP. and never had any issues during setups. Anyways not much to really say till I go through these steps.

[AAFJAPAN]

Well, I switched RX units to another phantom 2 unit, and it solved the problem. Decided to switch back just to see the difference again, and one of those magic WTF moments happened as it began responding appropriately. Still a very slight ramping, but nothing that matters in the real world. Flew it, and it was really nice. Actually felt more precise than in the taranis. The DJI sticks are just so light but but responsive. I am not a DJI fan, and other than a NAZA unit I bought in an emergency situation these TX/RX combos are the only DJI products I own.

 Anyways thanks guys. I may post pics of this build soon. I am liking it. Massive power.