Dusted off the NoodleCopter for a bit of late night flying with the KPIX news crew, all in support of http://knowbeforeyoufly.org. The reporter is Emmy award winner Joe Vazquez. I thought he was smart, asked good questions, and had a good understanding of the issues involved in flying small aircraft.
There will be quite a few people getting drones, quadcopters, RC planes, etc for Christmas and trying them out for the first time. If you're one of them, Welcome! We're glad you're here and hope you have a fun and safe time flying. Check out the rules and guidelines... they're not complicated, they can keep you and your family from getting hurt, and make life easer for all of us if we're all careful not to cause trouble.
If you know of somebody just starting out, point them to some good resources (hint: EastBay RC!) for them to learn.
Couldn't get the video to embed, but here's the direct link:
So we received an invitation to attend the big DJI product launch over on Treasure Island. It was an impressive thing... a top not presentation wrapped up by the appearance of the new "Inspire" unit over the stage.
It's an impressive unit. 4K video downlinked over DJI's Lightbridge system. The photo here is of the live video from the unit being downlinked and projected on the screen.
I met Eric Cheng a while back, when he had literally only been flying for two weeks. Even at that time I had been impressed with his natural flying ability, and of course I'm a big fan of all his photography as well. It turns out he's a pretty bodacious emcee as well!
Hey, who are those guys sitting up in the front row?
Yep, it's mythbusters Jamie and Adam. They've been using multicopters (custom built by their crew) for the last couple of years. They talked a bit about that, and how the technology has advanced in that time.
Here's a video of the actual product launch. I guess I can correctly say "literally" here! It was pretty impressive, and I liked how the legs raise up (or is it the body that lowers?) when it goes into the air? Eric and I are going to get together and do some things with his unit, and I'll collect any useful information. It looks like a serious contender for filling in the "prosumer" (i.e. not carrying a DSLR) aerial photography market segment.
Battery Chargers. These come in a couple of variations.
DC-only input. You can make these work if you (a) buy a DC power supply or (b) have an old PC power supply sitting around and are willing to do a bit of connector soldering. If that's the case, buy some 4mm banana connectors and I'll show you how to do it.
Both of these are similar. The IMAX brand has a better brand reputation than the Turnigy brand. I can't quite make out which cables come with the IMAX, but it may be that some extra cables will be necessary from the (tiny) picture.
I've got one of these. It's basically four of the Accucel-6 chargers in one case. You need to buy extra power leads, I think it only comes with one set.
these are all links to the USA warehouse. If you order from the international warehouse shipping is a lot longer. Make sure you're ordering "USA Warehouse" items if this is important to you.
APC Props
These are good props, but more expensive and only come in a single color. APC props are proudly made in California! The "P" props are the reverse orientation prop. The price is per prop, so be sure and order two (at least) of the "SF" ("slow fly") and "SFP" ("slow fly pusher") models.
Super-popular, these are sometimes hard to find. A lot of people like them because they're lighter and (the theory goes) will break and protect the motor shaft.
Here's a heads-up on an interesting new project. We may be getting some prototypes of a flying kite to test. I've always thought flying kites were beautiful and I'm excited at the change to try one out!
Preliminary specs for the two three channel models:
Not sure where I ran across these, but they've been sitting on my desktop for a while. A nice simple method of mounting LEDs with skewers and hot glue.
This is a neat idea... the LED shines forward, giving you a light for orientation at the same time that it lights up the propeller.
Jack Erbes is one of the modern day geniuses of foamy construction. Here's his RCG build log of a KFM3 wing for the Blu Baby. My BB32 wings broke in half, so I'm especially interested in this. I'll lightly edit, but the first person voice is Jack's. I'll try this later and do my own writeup if I think it might be useful. Jack writes:
I recently made a KFm2 wing with polyhedral tips for use on my KFoenix Stick and flew it in RET configuration (Rubber, Elevator, & Throttle, no ailerons) and was really impressed with it. It had excellent tracking and control and good rudder authority, which had been a problem with the stock Slow Stick wings. The wing seem to "lean into" the turns nicely as speed increased and as turns were made more aggressively. It was almost as if I were flying a wing with ailerons.
Decalage refers to the angle between a plane's main wings and the elevator, or the two pairs of wings on a biplane. There's a great discussion in the RCG UMX Radian thread. I'm capturing it here coz... well you know how searching on RCG is. Post numbers are 3431, 3434, 3435, and 3442. Big Thanks to Phil Alvarez and turboparker for sleuthing these out of the thread.
My Summary:
Porpoising is caused by bad decalage.
Don't try to solve by changing CG.
Small foamies can incur twists due to their light construction.
Standard decalage test:
Perform a power-off vertical dive from high altitude (neutralize elevator).
A. Model continues straight down: No change needed.
B. Model pulls to canopy: Increase stab incidence with respect to the wing.
C. Model pulls to belly: Reduce stab incidence with respect to the wing.
Here's the Concord Model Engineers Logo! 61 years continuously meeting, and we still have a couple of charter members that show up every month.
The logo had gone missing for a while, so it was decided to send a copy of the logo to every member in the hopes that somebody would have a copy the next time it was needed. DVCS at its finest!
We got to the end of the ArrBot class, and 6 ArrBots showed up for the rally! Sadly, a seventh ArrBot was stuck in a meeting with its driver.
I'm really pleased with how everybody did. We started off with a soldering class, raw plastic for the 3D printer, the literally cheapest parts we could find on the internet, and a partially finished prototype.
We printed the bodies, modified cheap 9 gram servos into continuous rotation units for power, and worked through some pretty tricky electrical problems. Everybody make prints and modified the servos. Most were able to run the tread test program. There's a few that just need some radio work (much easier now that we've had a couple of successful units under our belt!), and that should be ready in the next week or two.
From the class comment forms the students seemed to like everything as well. The hardest parts were the servo modification (honestly, a pretty challenging task with that size of servo) and the radio setup. There was a bit of a time crunch as well. In future classes we may schedule a class alternating with a lab so that there's no expectation of anybody having to work at home. We'll also be better on differentiating between 3.3V and 5V requirements. We needlessly burned some time having to work around that.
We'll continue to work on the ArrBot software. I think it will be a useful contribution to various two-wheel and two-tread robots. Some upcoming features:
gyro stabilized calibration
battle lights
onboard sound
Bluetooth control
ArrLang, the ArrBot robot control language
We'll also continue to study the radio hardware. There's some interference problems with the Bluetooth, and we had to add a capacitor to the nRF units.
But the end resulting units worked well and were a lot of fun to drive around. I definitely think that adding RC capabilities to educational robots is the way to go.
Notes to self for next session:
get all 3.3V pro mini's.
get switchable 3.3V/5V FTDI units instead of the FTDI cables, which were more expensive and not of particularly good quality.
get two FTDI units per student for ease of debugging radio communications. This can be done with the cost savings of eliminating the FTDI cable.
get 20mm header pins, as they held into the breadboards much more tightly.
have students provide two USB "B" cables for the FTDI connector.
get a stock of male and female crimps and covers.
have students provide their own wii nunchucks.
investigate SMD resistors for servo modification use.
Here's some video of the rally: The first item is trailer-style, the rest are are more traditional EastBay RC style videos. If anybody's an expert at embedding playlists on blogspot please give me a call!
Here's a quick overview of both the ArrBot and ArrMote.
Sigh, Youtube can be annoyingly dumb in its automated content matching. The KBS video has been disputed twice, so I'm putting the permissions here for easy reference by the Youtube staff.
BTW the KBS people were excellent to work with, this is no complaint against them. It's actually no complaint against Youtube, they're just having a problem with reflagging items they have already cleared.
Date: Wed, 15 May 2013 15:25:44 -0500
Message-id:
Subject: Electronic Copy of World Today from KBS of South Korea - UAV
From: Shin Sul
Hello, this is Kevin Shin Sul, one of US coordinators working with KBS of South Korea.
The program "World Today" with the portion on drones aired on May 11th, in South Korea.
Here
is the link to download the electronic copy of the aired program. You
are free to use your portion only in non-commercial purposes only.
Please confirm with me if you are not sure if it would be.
The portion regarding UAV will start at 33 minute mark and go on till the end.
Please click "PC" next to the link provided below:
Here's the Phase 1 ArrBot and ArrMote, completed and running.
I put the ArrMote electronics in a box to keep everything tidy. A plastic box There's a small hole for the Nunchuck wire, providing strain relief for the unit. The Nunchuck adapter is likewise twist-tied to the Nunchuck Connector. I think I'll put a switch on the outside... right now you have to open the box and attach the battery to the Pro Mini. The battery could be smaller, as low as 3.3V.
I did some fiddling with getting the nRF chip a bit tidier. I ended up using some short servo wires which were just long enough to raise the radio above the other wires. The 1000uF capacitor is way overkill but I had a bunch sitting around.
Here's the nRF24L01+. It is really neat and provides some pretty amazing capabilities considering its $1 price tag!
I think the Pro Mini 3.3V running at 8MHz is officially my favorite Arduino chip these days. Be sure and get a 3.3V FTDI cable!
The bracelet treads seem to be holding up pretty well. If you have a dud bracelet it will separate almost as soon as you start using it. If that doesn't happen it seems it will keep going for quite a while.
Here's how to set up the Arrmote. Pictures forthcoming! This page a Work in Progress!
Parts
Base
Arduino Pro Mini 3.3
Battery, 3.3V - 12V
The Pro Mini should have pins A4 and A5 on the top.
The Pro Mini should have a splitter on VCC, or a "long pin" on VCC.
The Pro Mini can optionally have a "long pin" on GND for ease of connection.
If you have two "long pins" it tidies up the assembly since the Nunchuck can be wired from the top and the nRF24L01+ can be wired from the bottom.
Nunchuck
Nunchuck
Nunchuck Adapter with header pins
4 wire F-F cable
Radio
nRF24L01+
100nF ceramic capacitor
7 wire F-F cable
(tentative) solder the capacitor across GND and 3.3V on the nRF board.
Software Setup
You can start by testing your Nunchuck and Radio as described (here and here, coming).
Edit the sketch (name)
Modify the radio channel assignment. If you are in the robotics class, your robot kit has a number. Use use channel 3*RobotNumber (e.g. robot 1 = channel 3, robot 2 = channel 6, etc).
Upload this sketch AND DISCONNECT YOUR FTDI CABLE.
Connections
Attach the Nunchuck to the Pro Mini using the 4 wire cable and Nunchuck Adapter. (details here)
"+" : VCC
"-" : GND
"d" : A4 (top, farthest from header)
"c" : A5 (top, closest to header)
Attach the nRF24L01+ to the Pro Mini (details here)
1 : GND
2 : VCC
3 : pin 9
4 : pin 10
5 : pin 13
6 : pin 11
7 : pin 12
8 : UNUSED
Attach the battery to the Pro Mini
"+" : RAW
"-" : GND
One the power is attached, your unit should be operational. (test program tba)
This project has been the first one where I've used a 3.3V Pro Mini. There's some important things to keep in mind regarding power input.
Power input on RAW can be 3.3V - 12V. It will be regulated to 3.3V. Specifically, VCC will be 3.3V
Power input from the FTDI connector is not regulated. If you are using a 5V FTDI connector VCC will be 5V!
In general you can get by with using a 5V FTDI cable on a 3.3V Pro Mini. It's out of spec but is generally accepted as working.
You can't use a 5V FTDI if you are powering a 3.3V device from VCC!
Here's a workaround if you're stuck with a 5V cable and want to use it to power a Mini that has a 3.3V device on VCC.
Remove the VCC line from the FTDI cable. Gently lift the plastic tab and the wire with attached connector will slide out. If you have an extra single-ping cable protector (take one off a female-female breadboard wire if you've got one) you can insert the VCC connector in there. You could also use some small heatshrink to cover it. You could let it go bare, but be careful not to short it out.
Attach the FTDI cable (minus the removed VCC) to the Pro Mini as usual. Attach FTDI VCC to Pro Mini RAW.
Check the voltage on Pro Mini VCC. It should be 3.3V, regulated by the Pro Mini.
Q: Are the rx/tx lines 5V tolerant? I want to know if it is possible to program this with a 5V FTDI board, without needing to get an additional 3.3V FTDI just for this board.
A: Yes, the only difference in the 3.3V and 5V Pro Minis is the crystal and the voltage regulator. Since the FTDI bypasses the regulator the only difference is the speed the run at. The ATMega328 is fine at 5V. Keep in mind if you have 3.3V on VCC and 5V on the I/O pins that is technically out of spec (you shouldn't put more than VCC on the I/O pins), but will probably work as well. Keep in mind that the 5V FTDI will put 5V on the VCC line, so make sure you don't have any 3.3V only devices connected when you do this.
blogodex = {"toc" = ["FTDI", "Arduino"], "idx" = {"ArrBot", "Electronics", "Voltage", "No Hack Too Cheap"]};
The X8R and most FrSky receivers offer 3 options for failsafe.
They default to hold last good command.
You can program a set failsafe position via the Taranis, or pushing the f/s button on the receiver while holding the channels where you would like them.
Push the F/S button on the receiver when the transmitter is turned off, and the receiver will not output anything when in failsafe. This is handy for many flight controllers to trigger a safety mode.
Courtesy of rotozuk. I didn't know about the third one.
We inserted an extra class after class 3 to give everybody a chance to catch up. Here's everybody hard at work.
The goal was to have these parts ready
3D printed body
two working continuous rotation servos
Arduino micro
and get everything assembled.
Here's a few ArrBots in various states of completion. This one is running the servo testing program to get the calibration values.
Servos in place. You can assemble the drive wheels either before or after they're attached to the servos.
More servo testing! Get that right and everything is easy.
Here's the battery case mounted on the frame. With this type of battery case you should use the wide body.
We're using 3M Dual Lock to put everything together. It snaps together and gives a better fit than Velcro, but just about anything will work if it holds the pieces together.
Downloading the test drive program. With all the pieces assembled, you can write an Arduino program to drive the ArrBot around.
We're using 5V Arduino Pro Micro clones. That's a mistake. We'll be swapping them out for 3.3V units so that we can use either a Bluetooth or nRF24L01+ unit. I think using a Mini is also better, since we're using a Mini in the remote control unit and we already have FTDI cables.
Getting ready for the next class, attaching the nRF unit.
.jpg)
.jpg)
