RC aerial vehicles: because we don't have flying brooms

The Something Awful Forums > Discussion > Hobbies, Crafts, & Houses > RC aerial vehicles: because we don't have flying brooms

«‹›3 »

CapnBry: Jul 15, 2002; I got this goin'; Grimey Drawer

i own every Bionicle posted:

But because it has active freewheeling/braking and that can�t be turned off, it makes weird knocking sounds when you are advancing or retarding the throttle because the ESC is designed to respond far faster than the 50 hz update frequency the receiver sends out. It hasn�t affected the motor health in any way but it would probably freak some people out.

I know you posted this a while back but I am a sometimes lurker in this thread who just comes to get general information about a hobby I didn't want to afford. However, I scratch built a Flite Test dollar tree foam board wing out of $3 worth of materials one night a month ago and it has somewhat started an avalanche of cash since then. I've never actually done any RC aerial vehicle flying so of course I needed to make that thing fly. When I turned on the active braking in my ESC (an EMAX I got off eBay for $6 shipped) and it would make the most unnatural sounds in testing so I quickly turned it off, thinking I was breaking my also cheapie motor. I've just been driving it from the receiver but I have Omnibus F4 with iNAV that I've got all programmed and half wired and is just about ready to find its way into an aircraft, so it's good to know that the 50Hz servo signal is what is causing the weird sounds. I thought it sounded like it was overshooting, then braking, repeat.

You all might get a kick out of my first RC flight ever. I didn't really know where to fly, being somewhat in the city, but there was a small retention pond beyond my back yard. It is about 300ft long and 100ft wide at its widest point, full water, and surrounded by tall trees. Perfect, right? You can imagine what happened with my tail-heavy plane. If you guessed "it got stuck in the very first tree it came by, which you then tried to extract but encroached on a surprise wasp's nest, got stung in the loving face and had a severe allergic reaction and your whole body swelled up like a Martin Short character", well drat you nailed it!

Can someone tell me what "Arming" is with regards to flight controllers (iNAV) and fixed wings? Is it just without having that switch on, the flight controller won't do any of the other higher modes like ANGLE, RTH, HORIZON, WAYPOINT, etc? I assume it still does elevon mixing and rates and expo and drives the ESC using the iBUS signal coming right? It feels like there's wayyyy to many places to change things like transmitter endpoints and subtrim, then transmitter rates and expo, and transmitter mixes, then the flight controller can do all that too. Hopefully I just run the rates and stuff flat at the transmitter then let the FC do all the real work? I'm sorry I have so many questions.

# ¿ Aug 16, 2018 23:32

Adbot: ADBOT LOVES YOU

# ¿ Apr 25, 2024 21:56

CapnBry: Jul 15, 2002; I got this goin'; Grimey Drawer

Ambihelical Hexnut posted:

That thing is a mighty beast! How long does assembling one of those take?

I was just finishing up on my new FT Arrow tonight, hoping to fly it tomorrow. The last one I had split in half on a bad crash, I repaired it, then it split down the other side on the next crash, so this one I built the box spar through the whole wing. There's also an F4 flight controller and GPS because I LOVE THIS poo poo. I need to get a drat smaller receiver than this FlySky business but isn't it wrong to spend $12 on a micro receiver when I don't really want to stick with flysky long-term?

The goalposts holding the motor in I call "Reflex Sticks".

What does everyone flying fixed wing do with their opposite props they have left over because they have to buy sets of CW+CCW on small sizes like 6x and 5x? I accumulated like 20 backward props while learning how to fly.

# ¿ Sep 3, 2018 02:12

CapnBry: Jul 15, 2002; I got this goin'; Grimey Drawer

ImplicitAssembler posted:

You change the motor direction.

Wait what, how does the prop nut stay on then? I tried this once and the little spinner tip flew off along with the prop. Now I have have a nylock nut on there but haven't tried to run it reversed, do those just hang on better? Admittedly I do not have a lot of experience.

# ¿ Sep 3, 2018 12:55

CapnBry: Jul 15, 2002; I got this goin'; Grimey Drawer

Golluk posted:

Setting up servos, and I can't use Motor 2 as a servo output.

I ran into the same thing a month or so ago, wanting to use the Motor 2 output of my Omnibus F4 Pro V2 to run a servo and found it similarly impossible. You're going to be glad you didn't go down the rabbit hole of actually getting it to compile before trying to make the change. Internally in iNav motor 1 and motor 2 share the same resource for generating the actual pulses so motor 2 would actually get whatever ESC protocol you have set instead of 50Hz PWM. The fact that it isn't able to be switched over into the servo pool by any means apart from changing the source indicates just how tightly bound to this resource it is, so it is non-trivial. I also thought it would be a simple reassignment of removing the pin from one list and adding to another but after digging into it for an hour I also decided it wasn't going to be so simple and accepted a workaround.

# ¿ Oct 3, 2018 12:55

CapnBry: Jul 15, 2002; I got this goin'; Grimey Drawer

Corky Romanovsky posted:

I was testing a new 1106 motor on this fresh out of the pack emax 12A esc, it was exhibiting strange behavior: not spinning up until I moved the temporary mount, then only spinning up if I juked the throttle wildly. Thought it could have been a loose connection, so I checked my soldering and then split open the heat shrink on the esc to find it like this. I did thrust tests on this esc, and also a 20A, the latter was able to hit higher thrust on 3S than the former. It looks like the MOSFETs(?) all dump into that pad anyway, is that right? Could it be settings in the esc that are preventing spinup and limiting top end? I haven't any flight controllers and Google assumes I'm searching for how to change settings through an FC USB connection.

That's weird. I saw your photo and I thought for a second it was this one I took a couple of days ago

All the four pins in the center connect to the solder blob that goes to the motor wires. On the other side, three of them are the input power and one is the switch (the gate) which is easily identified because it has a thin trace and the other three are linked. Mine was acting sort of wonky like you describe as well, but it came after the last time I plugged it in, I saw smoke come out of the ESC and it straight up melted the JST-RCY connector going to it. That behavior is also what happens when there's bad connection to one of the motor phases. Sometimes the motor will just sit there and twitch, but I go from 0-100% as fast as I can it will spin and run for a little bit until the MOSFET starts to smoke again.

I've replaced the MOSFET and it seems to be ok again, but they are 70 cents from Mouser so it's not very economical to replace them if they are going to go bad all the time, EMAX! Although this was probably my fault because I thought that if a prop pulled 14A I could just set the throttle limit to 85% and stay under the current limit. The problem with that logic apparently being that it is still pulling 14A, just 85% of the time.

# ¿ Oct 19, 2018 17:05

CapnBry: Jul 15, 2002; I got this goin'; Grimey Drawer

i own every Bionicle posted:

Gave the go-ahead for production of my Turbo Raven. Friend got some video of me flying it at Fall Nall:

Holy poo poo, that's pretty amazing both in the engineering and the flying. I'm not even clear on what control inputs I'd need to give to perform some of those moves, much less be able to do them without smashing the ground on the first maneuver.

# ¿ Oct 29, 2018 21:23

CapnBry: Jul 15, 2002; I got this goin'; Grimey Drawer

i own every Bionicle posted:

Thanks! I wish more people knew about 3D/XA flying and airplanes.

It is true, I have heard of "3D" but I didn't know what it is, leaving me to wonder when I see a video "Is that 3D?". Now I can easily tell what 3D is because I know it knocks my socks off when I see it.

# ¿ Oct 30, 2018 12:58

CapnBry: Jul 15, 2002; I got this goin'; Grimey Drawer

bring back old gbs posted:

i love NANO GOB V and his big antenna hat

Same with love for NANO GOB V and his ~~black and white cat~~ big antenna hat.

Nano Goblins are deceptive in a way that seeing them in a photo doesn't convey just how small they are. A surprisingly good platform though and incredibly efficient. Like 15 minutes of fly time on a 2S 800mAh, and a pair of 18650 3Ah cells will let you pretty much fly the thing to the moon and back again.

evil_bunnY posted:

It's probably not great (yet) for stuff like complex missions with interactive payloads but whatevs. Kid throw plane, plane fly gud, kid happy.

How do you have this set up? I was hoping to scratch build a pair of gliders with my niece and nephew next weekend (already got all the foam cut) and I'm trying to work out how much electronics I need to bring. Just a receiver, flight controller, battery, servos? Any tips to make it braindead simple?

# ¿ Nov 16, 2018 14:02

CapnBry: Jul 15, 2002; I got this goin'; Grimey Drawer

evil_bunnY posted:

I bought a build kit from the fundies at flitetest, and shoved f405 fc in there. ...

Thanks, that sounds about right. I didn't even think about doing just one servo, that's a throwback to a model I had as a kid that just had rudder control. I think I'm more excited about building models with them next weekend than they're going to be. Hopefully it will all come together quickly enough that they won't get bored putting the parts together and the flight is easy enough that they're not immediately discouraged.

# ¿ Nov 17, 2018 14:16

CapnBry: Jul 15, 2002; I got this goin'; Grimey Drawer

Kids glider assembly thanksgiving weekend was a smashing success! I brought enough cut foam to make 4x planes and we each made one in about an hour and a half total. My 70 year old Dad even made one, which shocked me (because normally he would rather just look at his iPhone). We fitted two with flight controllers, 2x servos, receivers and batteries and had a good time throwing them around. Longest flight was only about 35 seconds though which I think the kids were a little disappointed with after the initial excitement wore off. I had also prepared the servos centered and with linkage stoppers so the pushrod bending wouldn't take any time, and I think that saved a lot. Thanks again for the insight, evil_bunnY.

Nerobro posted:

so while we're on the subject, where's the reliable place to get 18650 cells?

I have bought from imrbatteries.com a few times and every battery that I've bought from them has tested out as the proper capacity, although I can only test up to 10W draw with full accuracy. I use the Samsung 25R (20A 2500mAh, $4) in a lot of projects but I've gotten some LG, Panasonic, and some Efest and IMREN high current ones as well.

# ¿ Nov 28, 2018 21:54

CapnBry: Jul 15, 2002; I got this goin'; Grimey Drawer

evil_bunnY posted:

You have no idea how happy that makes me. What did you end up using for FCs?

I had two Omnibus F4 Pro V2 boards that I bought for $15/ea from Aliexpress. They are super-great for fixed wing folks because they're just about the cheapest F4 controller out there and it has

BME280 barometer
MPU6000 accelerometer / gyro
OSD (and the video VCC has a small LC filter)
MicroSD for blackbox logging
3x UARTS, although I think one is dedicated to PPM/SBUS. I put my IBUS receiver on one and GPS on the other (no compass).
Voltage / Current monitoring
5V/3A UBEC. I power everything from this and on these 5-9g micro servos it hasn't been an issue with brownouts
Buzzer output pins (no buzzer)
Addressable LED output pins
2x motor outputs, 4x servo outputs (but only 2x servo outputs have full sets of 3 pins)

The only part I don't like is that last bit because I always have to make soldered wire harnesses to go to the ailerons instead of just using off the shelf servo extensions. It also kinda sucks that the USB port is on the side so you either have to mount it facing up (roll=90) or cut a side hole for access. You can't beat that cost though for a first-class F4 flight controller. You also have to watch out when you buy them because the "Pro" has a linear BEC (yuk), the "V3 Pro" doesn't have current monitoring (I think this is actually just a lie that it is a Pro), and I've also seen "V4 Pro" and "V5 Pro" which are nothing like this.

# ¿ Nov 29, 2018 20:20

CapnBry: Jul 15, 2002; I got this goin'; Grimey Drawer

Is there a way to snoop RC protocol traffic from a second ground-based module somehow? I'd like to pull the telemetry from my FlySky receiver into a second device for recording and processing but don't want to buy a separate telemetry transmitter for the model itself. Surely there has to be a way to just passively listen to the telemetry data that's already in the air and being displayed on my Taranis.

I looked around RC Groups and everything I can find there seems to be full-blown ground stations with way more equipment than I need. I was hoping I could grab like another iRangeX module and hook its telemetry stream to a microcontroller that does what I want, which is point my big mirrorless camera at the model and zoom appropriately for the range. I'm tired of trying to find a hat camera that doesn't look like awful when I have a $900 mirrorless with a tripod, servos, and all the GPS data I need already floating through the air.

# ¿ Dec 20, 2018 20:10

CapnBry: Jul 15, 2002; I got this goin'; Grimey Drawer

My Rhythmic Crotch posted:

Very bad, no good, rotten scoundrel drone pilots are up to it again, this time terrorizing Heathrow with distinctive red and green lights!

It is true, I've never seen a floating plastic bag caught in the breeze with red and green flashing lights. :v:

I built my own little thrust test stand last night out of parts I had on hand for comparing some motor output. I was going to switch to OneShot125 to control the ESC so I can get serial telemetry from blheli_32 but I've had an incredibly long googling session turn up exactly zero documents about the protocol timing. There have to be several hundred sites that explain how OneShot125/42 and MultiShot use shorter pulse lengths, but nothing about any sort of refresh period. If I send a 125us OneShot pulse or 25uS multishot pulse and then never send another pulse again ever, will the ESC just run at 100% until the battery dies? Or do I need to send out refreshes at 1kHz or more to keep it running?

It is most convenient for me to use an atmega32u4 at 16MHz and would rather use MCU cycles to do fun stuff rather than trying to run balls to the wall all the time to refresh at several kilohertz. From looking around it seems that the only specifications are that there are pulses of Xus which are sent at flight controller loop speed, but nothing about dead time interval mins and maxes.

CapnBry fucked around with this message at 22:10 on Jan 8, 2019

# ¿ Jan 8, 2019 22:02

CapnBry: Jul 15, 2002; I got this goin'; Grimey Drawer

Combat Pretzel posted:

There's probably a time-out. Not sure what it is, probably just a few cycles of 50hz, which seems to be the slowest rate I've seen with regular PWM (--edit: 1-2ms pulsing).

Don't AVR have periodic timers? Set one up in output compare mode to generate the OneShot125 signal, and update it as throttle values change. Takes no MCU time when it's running on its own.

Yeah I suppose I could just figure it out experimentally and see where BLHeli_32 starts cutting out. I'm not sure why I was so fixated on finding a timing spec considering I'll probably only ever use one ESC in this thrust stand. I was also thinking it would be difficult to inject the 30uS pulse to trigger the telemetry dump without possibly disrupting an in-progress throttle pulse but I can just count the overflow interrupts and force a 30uS pulse instead of a throttle pulse for one cycle out of every 200 or so assuming the ESC doesn't need to be refreshed every loop period.

Thanks for putting my feet back on the ground, I was clearly having a bad programmer brain day yesterday.

# ¿ Jan 9, 2019 21:50

CapnBry: Jul 15, 2002; I got this goin'; Grimey Drawer

It was pretty easy to set up a OneShot125 signal from code at 2kHz (or 1953.125Hz) and have it request telemetry with a 30uS pulse every 256 cycles (telemetry update at 7.8125Hz). My Code Here (gist)

It didn't work though. I could calibrate the ESC and control the motor no problem but no data on the telemetry line. That's when I saw in the BLHeli_32 manual it says "telemetry can only be requested via Dshot or Proshot" (sad trombone). Since even the slowest DSHOT150 is too fast for the atmega328 (I can only get up to 62.5KHz using hardware timers) I resolved to get up this morning to learn how to program STM32 DMA so I could write a DSHOT implementation. Before I got started I thought I'd make sure I had the latest BLHeli_32 firmware on the ESC. When I upgraded it, holy smokes they added Auto Telemetry in Rev32.6. Turned that on and I don't have to DSHOT at all because it just continuously sends telemetry every 32ms. Project done!

Aikon AK32 SEFM 35Amp 2-6S ESC Speed Control W/ BLHeli_32 $12
5kg load cell and HX711 ADC $4
Arduino Nano $4
3D printed stand $1
Button / potentiometer / wire / screws $2
Total cost: $22 (although I had everything except the ESC so the total cost of the project for me was $12)

I get the full deal for that too: Force, ESC Voltage / Current, RPM, and Temperature. I will say I need to design a better stand. It is terrifying to have a small piece of plastic clamped to your desk with a spinning blade of death pushing out 1000g of wind force.

# ¿ Jan 19, 2019 19:25

CapnBry: Jul 15, 2002; I got this goin'; Grimey Drawer

Yeah exactly, you definitely would have it off on a multirotor because there's no way to tell who sent the telemetry data, and you'd also need to sync the autotelemetry in some way to prevent two motors from trying to talk at the same time. TTL serial media access layer needed!

I'm not sure what the use case is for autotelemetry but it is perfect for this application, no analog measurements required and no extra tach sensor. I could see it being a work-around for wanting to get telemetry but having trouble with the length of the run for DSHOT. The protocol is only 115.2kbaud so there should be fewer transmission errors and also you can probably drop half the packets or more and still have a pretty complete picture of what's going on at the ESC. You still wouldn't enable it on more than one motor though unless you got UARTs for miles.

EDIT: And this is what the semi-finished thrust stand looks like. The potentiometer is manual mode and if you hold the button for 2 seconds it starts an automated sweep test 20% to 100% throttle in 10% increments. You have to keep the button held through the test (which takes about 8 seconds) as sort of a deadman switch in case things start getting out of hand. The CSV is just dumped out the USB serial port and I can copy paste the data into a spreadsheet.

CapnBry fucked around with this message at 00:11 on Jan 21, 2019

# ¿ Jan 20, 2019 15:30

CapnBry: Jul 15, 2002; I got this goin'; Grimey Drawer

Ambihelical Hexnut posted:

I don't know why the gps receiver needs to receive a signal back from the FC but I've always run them with both connected.

I'm not sure about the other specifics of that flight controller, but at least in iNAV, the GPS needs both TX and RX because when the GPS starts up in 9600 baud 1hz mode, iNAV uses the UART TX to send a bunch of initialization commands to change that (as well as some other settings I can't remember). Some GPS have onboard flash to store the settings so it only needs to be done once and saved so it might be possible to use the TX for something else as long as the baud, parity, start, and stop bits are the same, but not smartaudio which is 4800 baud. I think SmartAudio can use SoftSerial though, right?

# ¿ Feb 24, 2019 21:37

CapnBry: Jul 15, 2002; I got this goin'; Grimey Drawer

evil_bunnY posted:

Modern inav with the new servo config is actually the bomb, TYVM

Speaking of, I have a switch I want to make two stage flaperons through iNAV. Their FLAPERON mode is a fixed amount so I can't use that. I tried to mix another channel and add it to the ailerons in the mixer, but that makes them go up as well as down, instead of just the two two-stage down I want. Is there any way to add an offset in the new iNAV mixer or do I need to set up the transmitter to scale and offset the channel?

# ¿ Mar 7, 2019 14:33

CapnBry: Jul 15, 2002; I got this goin'; Grimey Drawer

evil_bunnY posted:

Let me gently caress with it tonight and see. If you�ve got an up-down config working you might want to just configure the up as no flaps and be done with it tho

Yah it was pretty easy with just setting the switch to enable Mode: FLAPERON and even have them extend slowly. I obviously don't need two stage flaps for my DTFB planes but I dunno, what else is there to do with my free time? Learn to be a better pilot? Pffff. Who has time for that when I can program my ESC, transmitter, inputs, mixes, special functions, lua scripts, inav mixer, inav PIDs, OSD, goggles, and blackbox tools exports to do marginally useful things which I'll never be able to remember the switch combinations to activate them :-D

I'm happy to hear I didn't just miss a big glaring obvious inav channel offset feature. It still isn't too bad to set the switch's channel to basically be positive only in the transmitter.

# ¿ Mar 7, 2019 22:30

CapnBry: Jul 15, 2002; I got this goin'; Grimey Drawer

Golluk posted:

Got myself one of those ImmersionRC power meters. I have a number of various antenna and Vtx, so it will be interesting to measure and compare them. And to find the best channels to use.

I have one of those too that I got for testing 2.4GHz transmitter output and it is only really useful for testing if you hook it directly to the transmitter with a cable and not trying to sniff it with the included little antenna. Getting the power meter's antenna just perfectly aligned with a transmitter antenna is a great way to compare numbers because just 5mm in any direction and your reading could be 20-30% different.

It would be nice though if the USB charging port also dumped the data and accepted commands to switch bands and modes so it could do sweeps and stuff like you want but it is still a pretty neat device for troubleshooting.

# ¿ Mar 15, 2019 20:59

CapnBry: Jul 15, 2002; I got this goin'; Grimey Drawer

My Rhythmic Crotch posted:

iNav is also very wonky about servo trim, last I checked anyway. It does that weird averaging thing where it tries to guess what your servo trims are set to. It's just dumb.

I always thought iNAV wanted your transmitter trims to be zeroed since "trim" received in any stabilized flight mode will be interpreted as a rotational rate. Two ticks of down trim on the elevator may make you fly straight in manual mode but in horizon mode it is "continuously pitch up at 3 degrees per second". I get around this by flying in MANUAL (aka PASSTHROUGH) and getting all my trims set, iterating if they are so far off that I need to physically adjust the surfaces. Once I have that all worked out, I just enter the center points into the configurator and zero the transmitter. I know they have a SERVO AUTOTRIM mode but that requires a dedicated switch since you flip it on and hold the stick to keep you flying straight and level for at least 2 seconds and it averages the stick position to set the new center points. I say dedicated switch because you need to keep the switch on until you land and disarm and use stick commands to save.

For cheap flight controllers / stabilization I use the Omnibus F4 Pro V2, which has gyro, baro, current sensing, OSD, microsd, 4 servos, 2 motors, and 3 UARTs and I've gotten them from time to time for $15-$18 on Aliexpress or eBay. Add a BN-180 GPS (no compass) for $12 more and you've got more fixed wing drone technology than many of the world's air forces. Plus it gives you fun blackbox logs to look at and see how many deceleration Gs your crap landings inflict on your model!

Unrelated, I 3D printed some landing gear for my DTFB Mustang, since the scoop on the bottom makes a fantastic arresting hook if you don't have landing gear. Educational experience: don't mount the landing gear directly on the CG mark.
https://www.youtube.com/watch?v=CxXZoKFdON0

# ¿ Mar 15, 2019 21:22

CapnBry: Jul 15, 2002; I got this goin'; Grimey Drawer

Golluk posted:

The fact that prop didn't shatter the first time it hit cement, tells me that plane needs a bigger engine :P I do like the rubber band shocks though.

Haha my prop budget would be sky high if i had motors powerful enough to shatter props on a ground strike. The rubber band shocks are neat but they sort of operate in reverse of how they should. It takes the most force for them to move from their fully extended position and it takes less and less force the more they absorb despite the rubber band getting tighter. When in full shock-absorbed mode, 100% of the force is trying to pull the wheel forward and 0% of it is pushing back against the ground.

quote:

It does actually have some data you can get over USB with telnet, but it was pretty limited from what I recall. Mostly read, no write. Being able to switch frequency over usb would be very handy.

Oh snap see here I didn't even think that they would have a USB interface for it but the ImmersionRC people clearly are pros. If you connect to it over serial/USB, you can set the frequency with 'f#' where # starts at 0=32MHz and goes up to 15=6000MHz. 'd' and 'e' read back the dbm, 'd' being average and 'e' being peak. There's also 'p' which reads back like 1923.11234 (not frequency specific but goes down the higher the receive power) and 's#' which sets the span (0=10ms, 4=160ms). So it does look like it would be possible to make an app to build a crude spectrum analyzer from it after all!

# ¿ Mar 16, 2019 15:19

CapnBry: Jul 15, 2002; I got this goin'; Grimey Drawer

i own every Bionicle posted:

Using a Taranis, is it possible to control a servo output from a telemetry input? If so, is a lua script the way to go? I have an airplane with a gyro that has pretty simple gain control (one channel controls the gain, it does not have full PID control) and I�d like to reduce gain as the airspeed increases.

You can do this with a mixer LUA script. Completely untested, even for syntax but here's the gist.

code:

local function run(speed)
  -- return between -1024 and +1024
  -- this example returns 100% (1024) at speed=0
  -- and 25% (-512) at speed=150
  return 1024 - ((1024+512) * (speed / 150.0))
end

return { 
  input = { {"your_speed_telemetry_name", SOURCE} },
  output = { "what_to_call_your_output_except_limited_to_4_characters" },
  run = run
}

Then set the MIXER source for whatever channel you want to output it on to what_to_call_your_output_except_limited_to_4_characters. You can also add inputs of type VALUE to allow you to set the max speed and reduction right on the radio (theoretically, I've never tried this) without editing the LUA.

CapnBry fucked around with this message at 20:31 on Mar 25, 2019

# ¿ Mar 25, 2019 20:29

CapnBry: Jul 15, 2002; I got this goin'; Grimey Drawer

i own every Bionicle posted:

Thank you! I did some searching and couldn�t find anything like this even though I knew it should be possible.

I actually saw some changes coming in the opentx source code to allow the use of telemetry items as mixer sources so this might be all built into the radio UI by next release. I could also be misinterpreting the code changes though.

I have also been thinking about building an FPV quad. I mean, I've got a transmitter and some micro receivers, that's like 90% of what I need right? :v:

That Eachine Tyro79 looks cool but I think I'd want something that can carry a gopro too. I'm struggling to keep myself from dropping $100-200 to just cruise around and look at things and possibly spy on girls, when I think I really prefer fixed wing flying. I do like putting things together though...

# ¿ Apr 6, 2019 14:10

CapnBry: Jul 15, 2002; I got this goin'; Grimey Drawer

Corky Romanovsky posted:

Fixed-wing, huh. Probably have a mishmash of parts, too. How well do quadcopter flight controllers do with different motors, props, and moment arms?

Yeah I've got a couple of F4 flight controllers, some micro receivers, FPV goggles, but not enough same motors or ESCs to just hot glue two paint stirrers to fashion a frame and give it a try.

I have used iNav in little nano goblins that weigh 200g and bigger 1400mm wingspan models and there's not much configuration to change between planes. You just set cruise throttle (or don't) then set how fast the model can pitch, roll, and yaw, rough out some pid numbers using that and you're done. I usually take it up once on manual mode to measure the rates and center the servos, then enter the observed values and we're good to go. I think fixed wing are really easy to tune compared to multirotors so you just need to be in the right ballpark to get decent stabilization. I mostly use it for being able to go into loiter mode because I've stood in another red ant pile in the field or forgot I left the transmitter in low power mode.

# ¿ Apr 7, 2019 12:36

CapnBry: Jul 15, 2002; I got this goin'; Grimey Drawer

Slash posted:

https://itchythumbs.co.uk/product/tiny-wiener-tiny-wing

I dunno what you're talking about!

I got a pair of 1104 8700KV motors for like $6 and mated it with a 6A ESC and a 2.5" prop and made a brushless F949 last year, but it eventually became so bent and flimsy you could only turn left if you gave it full rudder and full up elevator. I took the guts out and 3D printed this and covered it with saran plastic cling wrap. It flew okish but the scotch tape elevon attachment was pretty sloppy. It has so much power on 1S you could probably get the 50g thing into low earth orbit.

BurntCornMuffin posted:

How much stock should I take in this Wing Cube Loading calculator for determining performance characteristics? It looks like both that tiny wing is (guesstimating, since they didn't give wing surface) about 15g/dm^2, which I've gleaned is hard to fly for those new to fixed wing, and I saw some other popular wings (like the RMRC Nano Goblin) that have even heavier cubic wing loads. Is there a scale adjustment, or is it more of a really subjective thing that I'm caring about too much?

The (theoretical) good thing about cubic wing load is that it is supposed to be scaled already, so a 6 WCL on a 400mm model flies like a 6 WCL on a 2500mm model. The "wing loading" (not cubic) is size-specific though, so you can't compare wing loadings very well across models of different sizes.

quote:

For the wing shape, I had read that delta wings are more of a supersonic thing, and that a better subsonic wing shape would be elliptical or at least a tapered rectangular shape, but that was in the context of full size aircraft. Does that also hold at this scale? Is there a performance benefit for deltas in flying wings, or do manufacturers just do it to look cool?

The reason the wings are swept back on flying wing models is just for stability. Mostly it is because the weight of the motor in the back is so heavy you'd have a hard time getting the CG right, so they sweep the wings back, which brings the center of lift back which means your CG can move back too. "Plank" style planes with a rectangular wing and a pusher motor like the Nano Goblin have a long nose so you can put the battery far enough forward to make this unnecessary. Having the tips further back also increases the length of lever arm of the elevons sort of like why a traditional plane's tail is so far back, which makes the tips of the elevons more effective than their root ends.

quote:

What's a good way to estimate what sort of cruising airspeed my motor/airframe combination will give me? Also, do super-high kv motors, like the ones I have lying around from my tiny whoops have an adverse effect on fixed wings?

I think you're overthinking things now, but all fixed wings I've flown (400mm to 1600mm) stall between 10-20mph, with most being around 14-17mph. You can really get into the weeds with trying to figure out what motor size, kv, battery cell count, and prop size are appropriate. That link might not work because that's a paid site, but you can find motors they do have and get an approximate idea of what size you're looking for. You just want thrust to weight of 1.0-1.5 I'd say. I've flown as low as 0.75 thrust to weight but it feels pretty underpowered and can be hard to control for a new flyer since having power can get you out of almost any bad situation.

High KV motors going balls out I can't imagine there would be a problem as long as you're getting the proper amount of thrust. A guideline I've seen for power is about 75-100W/lb of all-up-weight, so a ~225g/8oz plane you'd want 37.5-50W, which is 10-13A on 1S, 5-6.75A on 2S, or 3.3-4.5A on 3S. On the model I posted a picture of, I get about 3A draw on 1S so I can estimate that the max weight of the airframe it is fun to carry is about 67.5g.

CapnBry fucked around with this message at 19:27 on Apr 11, 2019

# ¿ Apr 11, 2019 19:23

CapnBry: Jul 15, 2002; I got this goin'; Grimey Drawer

Warbird posted:

loving moran tries to ruin things for everyone.

"'... this incident shows why the federal government must mandate a remote identification system for airborne drones as soon as possible' said Adam Lisberg, the corporate communication director at DJI Technology."

You're so right. That government mandate that excludes drones from being able to fly over or near sporting events clearly works so well that this can be solved with more legislation. He also said that this pilot had subverted the geofencing restrictions in the DJI firmware, so even if the drone would have had a government-mandated drone remote identification system it likely would have also been defeated. What we need are enforcement drones and awesome aerial footage of high speed drone chases and destructive takedowns by law enforcement.

# ¿ Apr 12, 2019 14:17

CapnBry: Jul 15, 2002; I got this goin'; Grimey Drawer

ImplicitAssembler posted:

I'd forgotten about that extra set of low hanging wires
https://www.youtube.com/watch?v=y798FTce1E0

Where's the end of the video! Did it all work out or was your model sliced neatly in half down the middle? poo poo really comes out of nowhere on the FPV camera. My kryptonite is tree branches that go from not being seen to "this is so big you'll never get around" it in a fraction of a second. You had no chance with that wire.

Does anyone here have a Section 107 exemption for FPV flying? I fly around the small park in my neighborhood and was hoping to venture out a little further to explore out of LOS. I'm wondering if it is something legally doable without costing thousands of dollars in licensing and registration fees to the FAA and FCC. Are all the long-range equipment you can buy from any online retailer purchased by felons? I already fly my park sometimes without a separate observer so I guess that's me too, but I prefer to follow the laws.

# ¿ Apr 29, 2019 17:14

CapnBry: Jul 15, 2002; I got this goin'; Grimey Drawer

BurntCornMuffin posted:

Additionally, getting something that uses FrSky or Spektrum transmitters/receivers, which are both longstanding communications protocols for rc aerial hardware in general, reduces the cost of future dabbling since they're so widespread in the hobby scene already.

For budget transmitters, FlySky FS-i6 for $40 (no receiver) can't be beat. The mixer is limited to two mixes, but you won't run into that on a drone since the TX only needs to send straight signals and let the flight controller do the mixing. Micro receivers are like $12 too. 80% of my models still have FlySky protocol receivers in them and I've gotten range with tons of trees in the way up to 1km which is plenty for a small drone. I know some people have gotten 20km in a clear LOS too. It can also have its software updated to do 10 channels if 6 isn't enough.

# ¿ May 27, 2019 17:50

CapnBry: Jul 15, 2002; I got this goin'; Grimey Drawer

moron izzard posted:

wanted to see how small of an ardupilot capable wing I could build. omnibus f4 nano, 7a blheli32 ESC (turnigy, which im pretty sure are rebadging of the racerstar tattoo series), that motor, among other things

I wholeheartedly support this kind of fun. It also comes with 4x LED modules and I'm not being sarcastic when I say that's pretty sweet.

After seeing it come through the FPV reviewer sites, I kinda wanted that new ZOHD Talon GT "Rebel", that boxy looking gently caress. It's sort of just a giant crate you can load up with gear and send into the air. What I don't have is enough influence to get one for free, and geez $200? Last time I had 200 spare dollars was before I started this hobby. Instead I made my own out of parts I had on hand, a sheet of dollar tree foam, and a rigid insulation project panel. Total new parts cost, under $10.

# ¿ Jun 1, 2019 01:54

CapnBry: Jul 15, 2002; I got this goin'; Grimey Drawer

Judge Schnoopy posted:

But the real question was "will my Rx on ibus work with my fc on sbus"

Something to note, SBUS is inverted and IBUS isn't, so don't connect it to an SBUS input put that has a hardware inverter on it or it won't work. When connecting things over IBUS, I usually combine the IBUS and telemetry into a single pin and connect it to the TX pin of the UART I am going to use. The TX pin associated with SBUS usually does not have the inverter so it solves both problems. Downside: you need a diode and resistor to combine the wires using this pseudo-schematic. I also have to change things like 5 times between "Ports" and "Protocol" in the iNAV's configurator before I can get it to stick properly for some reason.

If you only want to connect IBUS in, then you can't use that trick though and just need a non-inverted RX UART on the FC.

# ¿ Jun 18, 2019 11:45

CapnBry: Jul 15, 2002; I got this goin'; Grimey Drawer

Judge Schnoopy posted:

Jeez, I'm glad I gave up and settled with ppm.
Maybe someday I'll upgrade to a big boy frsky x7, use real sbus, and move on from crazy hacks.

Haha oh every system has its own hacks for some reason. Like, why is SBUS inverted? Who makes an inverted serial protocol when we have standard polarity TTL? Oh that's because SBUS was a hack at the time, and now it is just most things account for it despite it being dumb. FrSky also has the newer F.PORT protocol, which is non-inverted (hooray!) but bi-directional (what?) since telemetry is carried on the same wire. It runs on the S.PORT wire, which is also inverted, so you need to either have a double inverter that works in both directions, or solder onto the FrSky receiver a new wire that accesses uninverted S.PORT and run that to your uninverted TX (not RX) on the flight controller. So, juuuuuust a bit less crazy hacky than the way IBUS needs to be set up to work, but still well situated the wtf hax pavilion. FrSky also just completely redid their air protocol, replacing ACCST with ACCESS which only works on certain receivers, is a one way "upgrade", and I think that only does SBUS so we're back to full hacks again.

When I need a micro receiver, I go with a Flit10+ which does FlySky IBUS and includes the pieces needed to run telemetry and the receiver on the same wire and is about $12. FrSky has the nice advantage that any data can be sent back and forth from opentx / Taranis, which allows you to change flight controller settings from your transmitter, but you pay twice as much for the luxury, the protocol has a ton of data drops, and the RXSR I have is EMF noisy as hell compared to all my different FlySky receivers.

# ¿ Jun 19, 2019 13:27

CapnBry: Jul 15, 2002; I got this goin'; Grimey Drawer

I'm pretty impressed too, after watching that Rotor Riot video. I mean the cost adds up pretty fast, but if you compare it to just a set of HDOs, a Runcam Split, R-XSR, and a decent VTX the price isn't out of line. The resolution is incredible by comparison though. It would have been awesome to include at least a composite video input to the goggles so you could still fly all your analog equipment with an external diversity receiver or something, as having $500 goggles that only work with the $170 (?) air unit is a bit steep for people like me with a few different FPV setups.

I just did some digging on if the latency has improved in OpenHD this year and looks like it is still about 125ms "glass-to-glass", with 110ms being a best case scenario. Getting better video quality and doing it with such low latency and integrating control and telemetry as well is really amazing and abruptly stops the digital HD for FPV is impossible discussion completely, or at least until we find out all the limitations or problems with the new DJI system.

# ¿ Aug 1, 2019 15:33

CapnBry: Jul 15, 2002; I got this goin'; Grimey Drawer

CapnBry posted:

It would have been awesome to include at least a composite video input to the goggles so you could still fly all your analog equipment with an external diversity receiver or something

Oh snap, turns out the goggles do have an analog a/v input, at least according to Bardwell. That's nice that you can still fly cheap analog stuff with just an external analog receiver. That makes the goggles a lot more useful especially one they make some better face pads for them.

# ¿ Aug 2, 2019 22:02

CapnBry: Jul 15, 2002; I got this goin'; Grimey Drawer

My Rhythmic Crotch posted:

Picasim is free and works with Taranis

Wow that is pretty complete for even commercial software, and amazing for free. Why didn't I find this when I was looking for simulators?

I use Real Flight 8 (RF8) Horizon Hobby Edition (I think they're all HH edition now). It works with any HID joystick too. I use a receiver and a Teensy microcontroller programmed to be a joystick to make a wireless link and just read the iBUS packets and convert them to joysticks. I thought at the time that I needed to have the latency to make it more real, but flying airplanes I don't think the added 30ms or so of latency makes any difference. My only complaints about it are that the wind resets to 0 if you change scenes or restart the application and that the UI is sort of half finished. I do like that you can edit the built in planes and change motors or throws or move the CG to see how it affects flight and you can also build your own planes. Its VR support is ok but pretty crap considering there's UI panels that always seem to be in my way, the performance is terrible for the small scenes it has to render, and planes become an unidentifiable speck pretty quickly.

The other downside is that it is hella pricy at $100 compared to most of my homemade planes, which are under $10 in materials.

# ¿ Aug 3, 2019 18:27

CapnBry: Jul 15, 2002; I got this goin'; Grimey Drawer

sigseven posted:

A little fear of a battery fire is not a bad thing. That style of charger is pretty typical for 1S, but it's best to not leave batteries charging unsupervised regardless.

I was wondering the other day when I realized I have three 1S batteries that have sat fully charged for like 3 months, are 1S batteries just so cheap that the everyone just charges them up to full all the time instead of doing storage charge? I hadn't been using storage charge since I fly almost every day, but my (3S) 1400mAh batteries are more like 1000mAh now after maybe 100 cycles and never settling below 3.6V so I'm starting to wonder if I should be charging to storage charge rather than topping them off at the end of every day. I have no way of doing storage charge on 1S though, so I just hooked a power resistor to them to run them down a bit.

# ¿ Aug 9, 2019 19:56

CapnBry: Jul 15, 2002; I got this goin'; Grimey Drawer

I've got a VTX that fell into the swamp for a minute or two and now it powers up and works in pit mode but when it fires up to 200mW, the range is like 5ft. I checked the antenna connector and it shows as a dead short across the center output pin and GND, even with no antenna. I've gone over the area with a giant magnifier and don't see any connections bridged or corroded and even checked under the can. I know the easy answer is to just replace it, but i can do SMD replacement so I don't want to drop another $25 if I can get away with swapping a 3 cent part or something. Has anyone ever seen any info on VTX repair?

# ¿ Aug 20, 2019 14:56

CapnBry: Jul 15, 2002; I got this goin'; Grimey Drawer

bring back old gbs posted:

if you can't bring it back to life with some dollar store rubbing alcohol and a thorough scrubbing with a cheapo toothbrush it's probably just toast.

Thanks for the ideas. I did the ISO+toothbrush route and reflowed the antenna connection points, then went through and checked all the diodes and caps for a short and didn't find anything. I'm hesitant to put it in the reflow oven due to the plastic connector which will melt way before any of the ground traces even get close. I think it is toast, although I'm not sure what's wrong with it. Stupid high density high power components! Appreciate the suggestions though, good idea to check the caps and stuff.

EDIT: And just for fun I fired it up on the bench and it seems to work perfectly again, or at least on 200mW I can go all the way outside the house and down the driveway with good signal. It still shows a short across the antenna but maybe the off state for the PA is conducting? Either way, woohoo! I also have noticed that whoever assembled the thing put a thermal pad on the RF chip (check) and a second thermal pad except rotated 90 degrees from where the output amp is.

So one to the upper left of the crosshairs and one on the far left corner where the board is the coolest instead of over where the crosshairs are. I'm guessing that contributed to why the output of this thing was so awful. 25mW=20mW, 500mW=220mW, 800mW=250mW. Guess now I have to get some silicone thermal pad to replace this because I've gotten some bench dirt on the existing squares.

CapnBry fucked around with this message at 19:31 on Aug 20, 2019

# ¿ Aug 20, 2019 19:06

CapnBry: Jul 15, 2002; I got this goin'; Grimey Drawer

I have been flying fixed wing for a while now and had discounted quadcopters as something I'd think was boring. I mean what do you do, just float around and look at stuff? But two weeks ago I built my first quad out of parts I had on hand and ~$90 worth of Banggood specials.

I had RealFlight 8, which I feel has a very good physics model, but their quads are all your granddad's quads-- mostly big giant bulky behemoths. It was good to learn the basics of how poo poo worked, but pretty boring to fly around their sparse environments without purpose. I bought Liftoff and it's fun on a bun. I pretty much learned everything I know about FPV flying from doing "race" mode every night for an hour. It took me a few days before I could complete a lap, partially because I was playing with about 250ms of latency, but mostly because it is really hard to shake the fixed wing muscle memory. "Just point your nose where you want to be" does not work at all on a quad unless you're going at full tilt. I'm objectively 30,000x better at quad flying now and I'm flying more with just the hands doing what they're supposed to instead of having the brain say "ok now lower the throttle so you go down more, and don't yaw just roll slightly... there! now level out". I fly using my Taranis QX7 to a spare receiver that I've connected over USB to be a joystick.

Flying a quad FPV is far more exciting than fixed wing. You have such control over where you're going and can fly sideways as fast as you can any other direction. You can fling yourself and float through gaps. You can cruise super slow to check out where you're going to go. It's just nuts how much fun it is. I do agree that Liftoff feels about right, at least in the limited capacity I can fly. I did just throw my quad into the ground when attempting my first real life dive and the ground came at me much faster than I expected.

Sorry I got so off topic, but (in my order of preference)
Liftoff - Beautiful, fun, some arcadey aspects, decent selection of 5"ish modern quads and online multiplayer
Velocidrone - Lots of user tracks, good simulation, weekly challenge races, okish graphics
Drone Racing League - Good simulation of their heavy race quads
RealFlight 8 - Good simulation, but boring and expensive and lacks modern equipment (but you can make your own if you want to learn their editor).
Does PicaSim even have quads? I didn't see them in there

EDIT: I am considering buying an HDMI to Composite video adapter so I can play Liftoff in my goggles over an FPV link, but Liftoff has ~100ms latency from my controller to seeing the stick move on my 60Hz display, so I think I've got enough latency without adding another wireless link.

CapnBry fucked around with this message at 14:13 on Aug 28, 2019

# ¿ Aug 28, 2019 14:10

Adbot: ADBOT LOVES YOU

# ¿ Apr 25, 2024 21:56

CapnBry: Jul 15, 2002; I got this goin'; Grimey Drawer

Hey, when am I going to stop leaning and tilting my body for no reason when flying FPV? I feel ridiculous when I take off the goggles and I'm facing a completely different direction than when I started the pack.

Although clearly I need to turn my body a little more sometimes

# ¿ Sep 5, 2019 22:29

The Something Awful Forums > Discussion > Hobbies, Crafts, & Houses > RC aerial vehicles: because we don't have flying brooms

«‹›3 »