First Quad DIY build considering T4 quadcopter

[voidstar]

I've heard of the "melting" print problems -- yikes, even just keeping them covered (out of direct sunlight thru a window) might not be enough? I try to keep them covered by a towel or box at least, but might also need to keep a window cracked. Or you know how Insight's have those solar roofs, to keep an internal fan blowing while away from the car -- those need to become standard

I did notice the bundle of bullet connectors truly felt "heavier" just holding them. But yes, "swap-ability" is why I kept the black plastic signal connectors for the ESC's also -- if I was going to make them swappable, I need the 5 motor+power cables on bullet connectors, and the 3 "signal cables" (well, lower power and signal) also on their own connectors. 30A was probably overkill (Brendan used 20A on his original). I'm not sure how the Phantom or Mavic get such long battery life -- like over 25 minutes? I'd assume mostly very efficient motors?


As for the landing gear -- if I keep them "straight down", yes, they tend to catch. So I found that pivoting them like 45degrees before a launch, they work a lot better. If they were slightly larger, they might also work as "prop protectors" (for walls/structures, not so much for trees). And I do think ABS may be the way to go for the legs/landing gear, and PLA for the arms. For the legs, I used a glue gun to better secure the ends.

Also, hard to see in the pictures, but I added a sort of "integrated roll bar" into the top body piece. About 10mm thick rib around the center of the body. Then a similar "anti-crush bar" up front behind the camera, another 7mm thick bar. These keep that top a bit more rigid if it ever does get stepped or pressed on (in one of the crash-flips, I had the camera get sheared off since the top body came lose) -- without the reinforcement, that would tend to pop-out the screws securing the top (which I am also looking for slightly longer screws for that, they're M3 I think). Not that we're going for BattleBots, but this is kind of a tank

 

[wafflejock]

Still have to battle gravity and the concrete if you're flying over that

Regarding plastics, PETG is a nice alternative as well, sits somewhere between PLA and ABS in terms of rigidity and strength. PETG is like coke bottle material (PET typically I think) so it has more give to it than PLA but doesn't warp while printing nearly as much as ABS (only down side I've experienced with PETG is some more stringing, or dripping, than usual during what should be filament free moves).

Regarding DJI they claim 'up to 23 minutes' which means ideal conditions hovering not actually flying since any amount of moving forward is energy you aren't spending staying in a hover and I assume this is something like 20ft off the gnd hovering not climb to 400ft then hover against the wind. They also say they have a 4S 4480mAh or so battery which is 17Wh total energy or so, can compare with your battery Wh to get some idea of total energy capacity in the battery. I think all BLDC motors are something like 80-90% efficient so I don't think they are getting much more out of that.

 

[voidstar]

Testing with a 4S lipo has worked out great!

Without the (heavy) landing gear, the 3S still works -- but perhaps barely (or maybe its the props?). Anyhow, the 4S has been night and day difference -- HOLD and LOITER modes are working great (no more weird vertical oscillation).

First FPV field test! I accidentally ordered a 6000mah 4S instead of 4500mah, so it was too tall to fit into the battery compartment. But to just test out the electrical aspect, we strapped on the 6000mah. The EagleTree Current Sensor is rated up to 6S; it seemed to be regulating the 12V and 5V paths just fine, and giving that little extra "ooomphf" to the motors to stay steady.

The camera is bouncy. Since everything else is checking out well, that's the part I'll address next: dampening for the camera.

 

[voidstar]

Wanted to report an update...

Both my T4's are still flying, but have had some surgery. Against very heavy winds today, both these T4's lasted 11 minutes on 4S 4500mah.

PART 1: REPAIRS

Ladybug T4 (red/black) crashed, but only broke a couple of the landing gears and a propeller. With the landing gear, a couple of the Y-shaped clips also snapped. Overall not a huge deal, had backups already printed, and was back up in the air quickly.

For Original T4 (clear/black), I replaced all the arms with the original T4 arm design. I had tried an alternative arm that provided more cooling to the motors, by being sliced diagonally along the end-cups for holding the motors. But these got weaker after some earlier crashes, to where they couldn't even hold the props at idle without cracking. Arm replacements are time consuming procedure (removing the whole motor and re-doing the ESC wiring), but they are all done now.


PARTS 2: SERVOS!

I also got external servos working. The X4R-SB receiver is great, by supporting 3 PWM channels along with SBUS for everything else. This means I got a 2-axis servo/gimbal working, then another 3rd channel for a cargo drop servo. Fun stuff! Also works with the X8R.



PART 3: AUDIO!

And I got a microphone finally working. The Vector has an augmented audio that play tones during mode changes and other such things, I didn't see this on the MicroVector. But anyway, the audio is passed thru to the Video TX. It works well for hearing the motors, but it has video interference noise also (kind of a signal static) -- plus I need to add a wind muffler.

The wind protector I'm going to try are these: TecUnite 5 Pack Mini-size Lapel Lavalier Microphone Furry Windscreen Muff
https://www.amazon.com/gp/product/B0798NZ23F/ref=oh_aui_detailpage_o00_s00?ie=UTF8&psc=1
The microphones themselves are cheapy ones from eBay, like $2. But small and pre-wired. I'm not sure how to extract out the static chop that seems to come from the video signal.


PART 4: VIDEO NOTES

Regarding video.... I'm using an EACHINE EV800D currently to capture video, which I think is recording at 720x576 (while the display itself may be 800x480). The camera I'm using is 700TVL, so it's still not ideal recording quality. May look into a more HD-capable headset that also has DVR, but the EV800D is good enough for practicing (on sale under $100).

​

I've worked on dampening the camera motion. Tried two different approaches:

(1) Using a dampening plate designed by squidmarks : https://www.thingiverse.com/thing:852391
EDIT: See Things File, then CameraShelf.STL
This works by inserting vibration shock absorption rubbers, something like these: https://www.amazon.com/10-PACK-Vibration-Absorption-Dampening/dp/B01F7T41GC/ref=sr_1_1?ie=UTF8

(2) Keeping the camera on the plate as done originally, but just adding anti-vibration "gel" (such as Kyosho Zeal Gel).

#1 is definitely the better way to go, but makes transport even bulkier.


PART 5: FLIGHT NOTES

Last thing I noticed, Ladybug T4 is using the MicroVector, and I noticed it pivots "differently" than the original T4 (that is using the Vector). I suspect maybe the MicroVector is more sensitive to its placement relative to the center of gravity?

I say this because Ladybug moves like a semi-truck -- meaning, it feels like it has a huge mass behind it. So while moving forward, then pivoting, it has a slower or less precise reaction (that can result in over-pivoting more than you expected). However, recall Ladybug also went thru a fairly hard crash, wonder if one of the motor shafts got slightly bent? The MicroVector software may have stuff to help compensate -- but for now, just have to fly it differently (slow down, then pivot).

But for both of these, the altitude hold is working reasonably well - you can still rise and go lower, but at medium power it holds the set altitude reasonably, so you can concentrate on just orienting along that altitude.

 

[voidstar]

My daughter calls her red/black one Ladybug. The white one has now survived quite a few flights, so time to name it: I'll call my white/black one Magpie!

Updated photo after a few flights: reverted to original T4 arms, added anti-vibration pad to the front. Two axis servos, but only using one (tilt). Back arms are braced, which I think helps in harder landings. And you can see the wind muffler poking out port side of the upper front bulk head.

 

[voidstar]

The muffled microphone is certainly an improvement at reducing the wind noise in flight! It's not fantastic (there is still interference static noise) but at least presentable. We've had many flights now, here is a more recent example in the link below. Also can compare the image stability with the prior video (certainly not Phantom-quality, but improved).

This video was done late in the day right at sunset plus overcast, so getting dark fast -- it is two back-to-back flights, with some bloopers from prior testing at the very end.

First part is high altitude, second part is low altitude- exercising Altitude Hold during both. [ caution, be prepared to mute or turn down audio -- motor noise isn't pleasant to everyone ]




I need to find (or print) a smaller, single-servo bracket-- the servos themselves work fine, but I think I need only Tilt and not also Pan.

 

[Pascal Linder]

Hi Voidstar,

I started also to have a closer look at the T4 quadcopter from Brendan22 since I got a 3D printer and wanted to use it for something "meaningful"
In October last year I started to print the copter's body and arms and started creating an amazon wishlist for the electronic parts.

As you mentioned in your post (btw, thanks very much for sharing your journey! I think I will read it through in more detail during a quite day) the components now 4+ years later are quite hard to find. Since I didn't want to start from scratch I decided to choose the Pixhawk as flight controller (don't ask my why I choose this as a starting point, I had never had a copter before this), then was looking at what ESC I could choose since the QBrain was no more available.
I was lucky enough to find a similar 4in1 ESC, the Skywalker Quattro from Hobbywing on eBay. I just ordered 2
Looking at motors I was a bit overwhelmed by the huge choice available, so I finally focused first on the TX/RX part. Choose the combination FrSky X8R / Taranis Q X7.
Finally I ended up in ordering 4 T-Motor F40 Pro II 2600kV and a set of 10 x 4.7 SlowFly propellers.

Testing the RX/TX using a little servo gave me already some good hope that I may be able to make it. I was already so happy to control a servo-motor using this big remote

Then started the even more interesting of trying to make this fitting together. The flight controller/Compass/GPs were not a big deal. The motors gave me more headache since I didn't gave too much attention when I ordered them. They aren't high enough, propellers couldn't even be mounted. I was almost about to start cutting the arms!
Finally it came to me that I could mount the arm top-down, which seems not to disturb a lot, just the aesthetic which is not very fantastic!

I did the first flight (well not a big flight) on January 4th after 9pm, all dark outside, but I could not wait for the next day after all these days/nights of DYI!
I had a lot of issues to configure the Pixhawk. I think it has some flaws so ordered a second. I had to deactivate a few security checks to finally let me arm the system and get the 4 motors spin for the first time!

During this first short flight, it did not take me more than 2 minutes to brake one of the first legs, i was a bit nervous and unsure on the control.
Back at work the next days there wasn't very much time for my quadcopter project, but this weekend I went again to Thingiverse and that's when I found the link to your page and all the extensions you made. Well done! Your copters look really great, I like the colore scheme you used, especially the red/black.

While writing these lines I am printing the second leg. They are really big compared to the original ones, I hope they will allow me to do smoother landings.

Here a picture of my build. It is not 100% finished, still some cable management to be done and the case needs to be fixed with more than only 2 screws. I did my first first with a big piece of tape around the whole body



In the next couple of weeks I will try to focus on the FPV/camera part. Hope I can build something similar to yours.

Kind regards from France!
Pascal

 

[voidstar]

Pascal: That's awesome! Indeed, the parts list has "evolved" over the years. I recall also spending quite some time deciding on a size suitable motor.

Once you get more confident with the kit, it gets fun to start making "experimental" arms (like truss-style arms that are more open -- so you can put the ESCs inline with the arms; but interesting you found a workable 4-in-1 ESC controller, just careful on the heat if it gets hot enough to melt/warp the print). I just use TinkerCad to modify the stock designs.

IDK about those low-profile motors, but nice idea on inverting the arms! As long as you've got the KV/torque to lift the weight -- I still recommend the 4500mah 4S. With the stock body, it's a tight fit for the 4S. I don't use the plastic battery clip anymore (it never printed perfectly for me anyway), instead I use a Velcro strap thru the slot.

And yea, I gave up on those short legs -- they're gonna break and are a pain to replace (at least for me). I still like the reinforced version I came up with.


In hindsight -- after borrowing a Mavic Pro for awhile -- I now really appreciate folding arms and folding props. I can't really fly at home, so I have to transport everything to more suitable areas. I do like the size of this class of drone -- stable in the air and easy to visually keep an eye on it. But this T4 design can be printed without supports, which is very convenient.

 

[Jackson]

The camera is bouncy. Since everything else is checking out well, that's the part I'll address next: dampening for the camera.

IDK about just dampening, seems like more to it than that.