[exileneb]

Hi folks,
Just trying to get back into this, building a quad from a kit.

I've got the motors connected, CC3Drevo as the controller and I've used librepilot for the software.

It goes through the initial set up just fine, but after that nothing. I ran through the transmitter set up and it recognises the 4 analogue sticks, but none of the switches. It also does not respond to the controller after this step, so I am unable to determine what the fault could be. It's paired ok, as it seems to be otherwise happy. Motor calibration also no longer works since upgrade. It goes through an endless series of beeps and tunes, but after about 5 minutes, its still doing this. I tried to manually calibrate the motors, but even at 1500 they are silent and still.

Last time i had a go at this, it was working just fine. Any tips or advice is gratefully received.
And if you have any questions, please ask.

Thanks

[Rickolas]

Have you set up your arming procedure?

https://librepilot.atlassian.net/wiki/spaces/LPDOC/pages/2818084/Input+Configuration#InputConfiguration-ArmingSettings

[exileneb]

Hi!

Thanks for the response,

Yes,

I've set it to roll left and with 10 secs timeout. 

If the motors wont spin on calibration though? i think its lost somewhere before this point.

The controller gets a response from 4 channels on the remote control input page, only on the info bar though, not on the virtual controller. I also see no response from the other inputs. I have a FLYSKY FS-i6X controller. 2 additional analog dials, and 4 switches, but I cant see any of these on the virtual set up. Nor can i seem to configure them.

Appreciate the help!

Neb

[Rickolas]

If you can only see 4 channels, you need to enable the additional ones on the i6x controller and perform the transmitter wizard again to see the switches.

My i6x shipped with all the switches disabled, you just need to assign which switch to which channel on the i6x and try again.

This guide should help you https://blog.dronetrest.com/flysky-radio-setup/  Just remember to hold Cancel to save the settings.

Regarding the motors not spinning, did they move when you tried to perform ESC calibration?

[exileneb]

OK, So progress is progress.

Channels were as you said, default to the analogue dials. Now set for switch 0 and 3, arm and flight mode. Now responds to RC input. Exciting!

What is my next reasonable step? Don't want to do any damage. But keen to see it fly.

Still experimenting.

Neb

[exileneb]

Regarding the motors not spinning, did they move when you tried to perform ESC calibration?

How exactly do I calibrate the ESCs? Using the manual method, they all react pretty uniformly. At about 1025, they all spin consistently, but when I try to get it to fly, rotor 1, the NW one hums, and jumps back and forth.

[TheOtherCliff]

Working in one way and not in the other way sounds like you should retest the manual method to make sure they all run well there now.  Maybe it broke between the two tests.

Sputtering like that is usually caused by the motor, or a motor connector, being bad.  Swap the bad motor with a good motor to verify that the problem is the motor.

Another issue that can cause this is bad ESC firmware.  In this case the quad is slightly tipping on one direction and the lowest motor of the four is trying to start up to a higher throttle than the others.  You can try making a different quadcopter arm be the low one and see if that motor now fails.  If it does, then a work around (till you flash new firmware into all ESCs) is to see which motor sputters and tilt the quad to raise that motor slightly to get them to start evenly.  I had a quad that did this and that is how I got it working till I flashed the latest firmware and fixed the problem.

[Rickolas]

In GCS, under the flight data tab, does the craft look level in the bottom left? If not, you could try and calibrate the Gyro/accelerometer etc. THis might be why one motor is moving slower than the rest.

If not, You should probably run through the new vehicle wizard again.   

When calibrating the ESC's, once you get to the figure where the motor starts to rotate fine, add +3 to the figure and save. Do this for all your ESC's. This should ensure that you won't get a motor that stop/starts when at idle.

[TheOtherCliff]

"Expensive" ESCs have a crystal controlled CPU clock that doesn't drift much and you can probably get away with only adding 3 to the Neutral setting to get reliable Neutral "idle speed".

Less expensive ESCs have a built in CPU clock that drifts when its temperature changes.  You can see this by doing the ESC calibration and finding the very lowest number it will run at, then let it run that way, idling for 5 minutes and see that the motors will stop.  Increase by one and some will still stop after a while.  In flight, you will see them stop at idle power when they should be idling, especially if you calibrate your ESCs in the cold and then run them in the summer.  There is very little down side to adding say 15 or 20 to the Neutral number.

All of this is important if you do stunts and / or ever run your motors at idle speed during flight (e.g. CruiseControl and flips).  If they stop, they take longer to start up and spin fast than if they are idling.  There may even be cases where the motors spin backwards when stopped and the quad is falling.  It may be that they won't start up at all in that case.  All it takes is one motor that doesn't start or starts slowly.

A related issue is that generally one ESC's BEC is supplying most of the power to run the FC and accessories even if you have all 4 BECs connected.  That ESC will be warmer and will need higher Neutral number to run at idle speed.

[exileneb]

Thanks for the advice!

So I've recalibrated the gyro, on a flat table top. The virtual horizon looks good and responds well when I am moving it around. I've noticed that the "bad motor" swaps to the SW one occasionally, so it looks like a software/firmware issue. I did manage to get it to hover at about 2 feet for a minute. It responded pretty well to inputs, then the SW motor sounded like it juddered, and the whole airframe crashed forwards into the ground. 2 broken props. So replaced and tried on desktop again. Same results as before, but this time the motor seemed ok at about 1035, so I've set them all to this.

I've checked the connections to motors, all ok. Freshly soldered the connections just to make sure. I'll try again later.

Regarding motors and ESCs, i think I'll upgrade them anyway as the ones I got with the kit are just generic stuff.

Anyone UK based have recommendation for a good source of quality equipment?

I'll try the idea of upping the idle too. As you say, its very little difference to the whole performance. A few score rpm extra.

Thanks for the ideas, i think we are on the right track.

[Rickolas]

Thanks for the advice!
Anyone UK based have recommendation for a good source of quality equipment?

try www.unmannedtechshop.co.uk, it's where i've had most of my gear from and they are pretty knowledgeable 

[TheOtherCliff]

I've noticed that the "bad motor" swaps to the SW one occasionally, so it looks like a software/firmware issue. I did manage to get it to hover at about 2 feet for a minute. It responded pretty well to inputs, then the SW motor sounded like it juddered, and the whole airframe crashed

Difficulty starting (but it does start, at least sometimes) is usually an incompatibility between motor and ESC firmware, and often gets fixed by flashing the latest BLHeli or SimonK firmware into the ESCs (all of them).

Motors loosing phase lock (what you called juddered) sounds like you threw a stick into the prop, and with about the same effect is also caused by an incompatibility between motor and ESC firmware.

Both of these issues are just ESC firmware.  Although admittedly some ESC hardware is designed and built very cheaply, it's rare for ESC hardware to be the main cause and it almost always runs just fine on the latest SimonK or BLHeli firmware.  Some time after getting your feet wet in this hobby, you should consider learning about ESCs and how to flash them.  That's a large study topic.

These also can be caused by bad or intermittent connections from ESC to motor.  I had it happen once where one of the factory soldered motor connectors was poorly soldered.

Generally these issues aren't related to FC firmware, other than the fact that the FC tells the ESC what speed to run, and some bad ESC firmware can't jump from 0 to 25% the way some bad car motors can't jump from 0 to 25% throttle when they are cold.

[exileneb]

try www.unmannedtechshop.co.uk, it's where i've had most of my gear from and they are pretty knowledgeable

Ok! I'll give them a look

[exileneb]

TheOtherCliff,

I'm quite into programming and electronics. Play around with arduino projects etc. I'd be interested in finding out more about flashing ESCs. I'm going to order replacement ESCs and motors, and more props! i think I might see of I can get some guards for them, until I can get it to behave.

Do you have a source for the flashing procedure? Software and hardware needed?

Many thanks for your help so far, its been valuable.

[TheOtherCliff]

IMO prop guards are only useful for tiny indoor quads where a bump into furniture is not a crash.  For a big one, the crash will move the guard enough that the prop sill still break.

There are 2 basic broad questions:
- does it have a (SimonK or BLHeli) bootloader
- what kind of processor does it have, SiLabs or Atmel

and
- do you prefer BLHeli since it can run on either CPU type or do you prefer SimonK (Atmel only) since a SimonK bootloader can be used to flash firmware / settings and (at least last time I checked) at least some versions of BLHeli bootloader can only flash BLHeli settings.

Then there is BLHeli_S which I haven't used and won't get into here.

If it has a bootloader, you can in some cases use a cheap Arduino with yet another set of firmware (Arduino sketch) to flash it (using the ESC bootloader).  If it doesn't, then you need to connect directly to the CPU pins and use a USBASP/USBISP or dedicated flashing dongle.  There is a $25 Atmel programming clip to make the connection, or you can do a lot of tiny soldering.

Although the result is the same, with SimonK, you set options in a source code file and compile those options into the firmware, but with BLHeli you program the options you want into EEPROM settings memory.

SimonK bootloader can flash firmware, but at least some versions of BLHeli bootloader can only flash BLHeli settings.

I suggest that you search Google for:
  flash esc firmware

and I also suggest you look into BLHeliSuite as the software to do the job.

Another thing.  If you flash the wrong firmware, you will immediately burn it up as soon as you plug it in.  For flashing and especially for testing, use a current limited power supply ... or a car light bulb (like #1156) in series with the battery power, inserted in line in the high current positive wire.