Hello all, I am very much a newbie to RC and quadricopters but I'm working with a group of fellow students on constructing an autonomous quad for a university project. We are currently performing our initial setup. We do not have an RC transmitter and whilst we could beg/borrow one, we would like to know, first, how to go about calibrating the 4-in-1 ESCs without one. I have read online that it is possible to do this (although the commenter didn't say how) but that it is dangerous since for a first flight you would have no manual override in the event of mis-calibration(s) causing erratic behaviour, however we plan to tether the quad in test flights to overcome such an issue and then fine tune it with/without an RC transmitter (depending on whether we can get our hands on one without purchase). So, what is the method for performing this calibration through the APM 2.6 in Mission Planner/APMPlanner 2? One more question: We would also like to know how to incorporate an Arduino code program into a weypoint Flight Plan in Mission Planner/APMPlanner 2. A bit more background: We have built up the frame (except for the landing gear, which will be Araldite-d on soon), mounted the motors, prop savers and props, and connected the electronics to the latest versions of Mission Planner and APMPlanner 2 on the laptop by USB cable, however we have not connected the electronics to the motors. As soon as the quadricopter has left the ground there must be no further human intervention and it must autonomously carry out a mission to cross a generic flight zone and then enter a target area. It must have a payload release system and must deliver a water payload accurately over an imaginary fire in this target area and then return to launch. Throughout the flight it must be able to stay within a geofence (or if it does exit, it must swiftly and in a stable manner return to the geofenced area) and it must be able to cope with rain and gusts of wind up to 8m/s. Our permitted budget for the entire project is £250 and we are currently JUST on budget. In terms of components, we have: 3DR (clone) APM 2.6 3DR uBlox LEA-6H GPS with compass 3DR Power Module 3DR (clone) Radio Wireless Telemetry Turnigy 3S 20C 3300 mAh LiPo battery Hobbywing 20A Skywalker 20Ax4 UBEC 4-in-1 Brushless ESC 4 Turnigy Aerodrive SK3 - 2830-1020kv Brushless Outrunner Motors 4 10x4.5 Black Rubber Propellers; 2 CW, 2 CCW 4 Prop Savers 'Thermal Knife' circuit as per Bovine Aerospace Kilner strainer funnel Custom, self-built aluminium frame Coat hanger landing gear 3D printed anti-vibration mount as per Omnimac 3D printed GPS stalk mounts as per Omnimac LED flasher unit With regards to our second question, we would like the quad to automatically fly by waypoints to the target area, hover and release the water payload. This latter process is controlled by a piece of Arduino code written by Bovine Aerospace and made open to the public. It is this piece of code that we would like to incorporate into the Flight Plan in Mission Planner/APMPlanner 2. We would be extremely grateful if the members of this community would help us out. Yours faithfully, FlyPhi |
You need to be a member of diydrones to add comments!
Replies
You state that you "could beg/borrow" a Tx. Do that, and you won't have any of these silly problems and what is much more important you will have a way to assume manual control of your aircraft WHEN (not if) any of dozens of things go wrong. This may sound harsh and is not meant personally: It is stupid, dangerous and irresponsible to fly anything weighing more than a very few ounces without being able to fly it manually. We, meaning the UAV community, are in a battle to convince the public and officials that draconian regulation of our aircraft is not necessary. Plans like yours, coming from someone with no experience, are the antithesis of what we here should support and are typical of the entire problem we face. Do this right, and you will be welcomed and you will get all kinds of help, perhaps even donations, from the community. And your aircraft will also have a much better chance of surviving its first flight. (By the way, "tethered" test flights are a bad idea for a bunch of reasons).
So, get that controller and once you have your aircraft ready, learn to fly it. It's not that hard (easier with a quad than with any other aircraft).
Just buy a controller, they are nt that expensive and can be bought for $50ish.
Thanks mP1, might do that with James' suggestion.
Hi, If you do not plan to use RC with it then you could just set the parameters manually but you would have to know the default settings for your ESC. In all likelihood I imagine that the default settings are flyable any way. You could also use a cheap servo tester such as this: http://www.ebay.com/itm/Steering-Gear-Tester-Servo-Motor-Tester-Ele.... Another option might be to knock one up yourself. Any Arduino or Picaxe along with a pot should be all you need. You need to generate is variable 1-2ms pulse at 50Hz.
Do you have a link to the code by Bovine Aerospace?
Also of interest might be the potential use of a standard computer joystick attached to the GCS in place of your RC transmitter. This includes controllers such as Xbox and Playstation - I'm sure you guys will be able to rustle up one of these for little to no cost.... In this case, the GCS (Mission Planner) uses the joystick to override the RC inputs on the APM via your telemetry link. Take a look at 'Joystick' on the 'Actions' Tab Unfortunately this option will not help with the calibration.
Interesting, thank you for the script that's really helpful. What I don't quite understand is where do we put this script in Mission Planner in order for it to work? Yes we do have to demonstrate a fully autonomous flight and any interference with an RC transmitter will mean automatic forfeit so I'm sure we'll run the mission with this script (but with an RC transmitter to hand, of course, in case something should go wrong).
Ah, no it's not rigged up in the same way that it is for Bovine Aerospace; we're using the heated Nichrome wire (the 'knife blade') to burn a small hole at the bottom of a water balloon filled with water, which sits inside the funnel. This is analogous to popping the balloon with a needle, through Sellotape stuck to the balloon wall; the air (or in our case, water) rushes out in a thin, directed stream, but the balloon essentially maintains its structure and stays in one piece. The wire only needs to be heated for about 3 seconds for this to occur. The 6V max rated reed relay switch in the circuit is activated by the 5V signal from the APM and allows the current from the 9V PP3 to flow through the Nichrome wire, heating it up. I have a circuit diagram of this, which is basically the same as the one on Bovine Aerospace's website. A doctor in Electronic Engineering at my University suggested that I use a optical couple to send the signal to the reed relay switch and thus isolate the Thermal Knife circuit from the APM. However, Bovine Aerospace connect their circuit directly to their Arduino and don't appear to have had any issues with it (although our circuit's battery does have 3 more volts than theirs). I could probably still do this with the time I have left but it depends on whether I can get the extra components in time.
Yeah that sounds about right and thank you again for the link - so is that the only way of changing the digital state of the output pin or are there other commands that I can set?
Is there no facility in Mission Planner that allows you to just tell a particular pin to go high once a waypoint has been reached?
Thanks so much for all your help!
The script is a plain text file with a .py extension. You can load and run scripts from the 'Scripts' Tab on the 'Flight Data' page. I have not done a lot of scripting but it may be possible to set auto mode, arm the copter and bump the throttle all from the script.
I found the circuit diagram for the 'knife' here:
http://bovineaerospace.files.wordpress.com/2013/06/cutdown_circuit.png
If you have this circuit built already then you can hook it up the same way they have. Note that the 5v power for the relay is SHARED by the Arduino and does not pass through it. Also the knife itself has a separate power supply (perhaps you could use a 9v battery?).
If you have not made this circuit yet then I would suggest perhaps using a single mosfet as the switching element rather than the relay. This will simplify the circuit. You do not mention the current draw of the 'knife' part of the circuit. Do you know the gauge/length of the nicrome wire you are using?
I am not sure if there are any other options for changing the state of a digital pin in AC but I do not think so. It relatively straight forward to do in code though. Is there any reason that the method I described is not suited to your task?
James, one more thing:
On the ardupilot website it states: 'Pin 54 to 62: You need to add 54 to the pin number to convert from an analog pin number to a digital pin number. So pin 54 is digital output pin on the A0 connector. Pin 58 is A4 etc.'
I don't quite understand this - where/when do I put 54? The context in which I'm asking is the circuit diagram for the Thermal Knife; one lead must be connected to the 5V rail (+), a second lead must be connected to the Ground rail (-), and a third lead must be connected to a digital output pin.
Does the '54' bit mean that I can connect the first and second leads to the + and - 63 pins on the A9 connector and the third lead to, say, the 'S' 62 pin on the A8 connector but then I have to write this down somewhere in Mission Planner? It's got me a bit confused!
Thanks in advance.
These pin numbers are not used in AC - the shutter settings in the gimbal setup page are sufficient.
You need to connect the digital input to your circuit to the signal 's' pin of A9 as seen in this image:
http://copter.ardupilot.com/wiki/common-camera-gimbal/
Your ground can be connected to the -ve pin at this same position too but you can not used the +ve pin to drive your relay as this is a low current output only. You should supply this voltage separately. How are you powering the APM?
Unfortunately I only have Powerpointviewer and not the full software. Your file appears broken into three pieces and I can not see what connects where. Looking at the Bovine Aerospace circuit, I would recommend adding a resistor between the A9 's' pin and the base of the NPN transistor to limit the current - something around 330 ohms should limit it to around 15mOhms which should be plenty to drive the transistor and within the capabilities of the APM. Without a resistor, the s pin will essentially be shorted to ground which would risk damaging the APM.
Here is an alternative circuit to power your nichrome wire triggered by the APM. Any Power MOSFET that meets the voltage/current requirements with an Rds ON of less than a an ohm when Vgs=5v will do. Something like IRLZ34 would be fine and cost only a buck or so. A FET to do the job will not be hard to find - your local electronics store should have something if you are desperate.
Here's my stripboard circuit diagram in PowerPoint if it helps for easier understanding of where the leads actually are
Stripboard.pptx
-
1
-
2
-
3
-
4
of 4 Next