Building my own Quadcopter

[FDS]

Capacitors only smooth out voltage spikes. I run them just on the main battery input, wherever that is. You don’t need 4 on the ESC’s as well.
I would look at getting a cheap carbon frame, the ZMR250 clones are under $15 and are going to be much more rigid than anything printed. That would eliminate vibration from the equation, then you can look at other potential issues.

 

[PsyBorg]

At this point mate you need to make a decision to either continue this quest with the gear you have on the 3d print frame or to step into the realm of traditional quad building. Either way I will try and get you sorted out albeit at the expense that comes with experimentation.

If the white smoke left the motor that means to me it was over heating from being rev'd up and down too much by the FC and was over worked. More then likely when that motor shorted you can say good by to the esc as well. Its almost a guarantee the phase that fried in the motor now has a direct short thru the esc.

We will never know if the recent changes are the cause or this is results from previous heat issues and it finally gave up the ghost.

What ever your choice its probably best to film each flight so we have a better idea what direction to steer you in to achieve your goal. I have a feeling the idle noise you describe is due to yaw being so off track from anything I have ever flown. An example is after I spent weeks tuning my build on 4s the graduation to 5s added a grinding type noise at idle. that was due to too high yaw p and d settings most of that went away when I figured out where the issue was and cut yaw gains nearly in half.

Anyways dont fret we have all been there. I recently smoked too high end motors AND a 70 dollar 4 in one esc. Set me back about 6 months by the time I can buy the replacement to get back in the air. That was caused by my own skipping my routine of preflight and having screws work loose. No ones fault but my own.

 

[Tammeo]

That's a very bad news,
I was able to update from an F3 to an F4 Flight Controller (as suggested by others) and I replaced the motors with a cheap "series" of 2207 (instead of the previous 2205RS) but I did not replaced the ESC...

I tested the circuits after soldering the new components and the check was ok (no beep of short circuit from multimeter)

the plan was to test tonight the Flight COntroller plugging it to the PC to calibrate the parameters... how can I check if the ESCs are ok or not?

 

[PsyBorg]

Continuity check between the motor phase outputs back to the positive and negative inputs should revel the shorted phase on the esc. after the burnt up motor is unsoldered.

 

[Tammeo]

it's been a while since my last update,

basically Psy was right and one of the ESC was fucked up by the FC. so I decided to stick to my "3d printed parts" plan and I've bought a lot of replacements. in fact they were so many I had spares enough to build a second 3d printed quadcopter (same design).

#1: Main Quadcopter

1. FC: matek F405-CTR
2. ESC: Wraith 32bit
3. Motors: 2207 2600kv Samguk
4. Props: 5040
5. TVX: rush tank Race edition
6. Camera: Foxeer Predator 4

#2: Trainer Quadcopter

1. PDB: matek XT60 (probably fake one)
2. FC: Gootoqomo betaflight Omnibus F4 (obviously a fake one)
3. ESC: Racestar RS30a V2
4. Motors: 2207 2750kv iflight
5. Props: 5040
6. TVX: NONE
7. Camera: NONE

I've set up the FC parameters following your suggestion and yesterday I did my maiden flight for both of them...
the Trainer ( #2) had the same behaviour of the previous ones: esc overheat, motor overheat, I had to stop to avoid motor failure.
the Main (#1) was unbelievably smooth, no twitch, no issue at all, motors and esc were extremely cold, I depleted two "3cells 2200mah" batteries with no crash, no overheat and lot of fun. (at least from me)

so, this experienceI proved to me that:

1. my 3d printed design was solid enough for my skill level (maybe when I will be able to do aerobatics I will have to switch to a carbon chassis)
2. the "twitch" that screwed my first quad is somehow related to the FC.

Now I am installing a Matek F405-STD FC onto the # Trainer Quad to prove my theory.
if it will be confirmed then I swear I will not use non "genuine" parts anymore on any of my future quads.

 

[Tammeo]

Here is my 1st flight in acro Mode... Please Be So Kind to watch it and Point out Everything that seems Weird to you... i need every feedback to be able to improve my Skills ad the design of my quadcopter.

 

[PsyBorg]

Congratz on the build and maiden success.

Right out of the gate I see two MAJOR issues. First dont be flying anywhere near or over those roads like you did in the first part of that video. Things like that are why we are getting hammered with rules and laws and all the governmental mess. Dont be "that guy" contributing to the issues and giving the anti drone community more ammunition. You got that whole huge field to be learning in with an unproven craft.

Second get rid of that horizon osd stuff on the screen. Its unsafe as you are busy paying more attention to that then watching where you are actually going and potentially going to hit. There are enough physical references in view to tell you what attitude you are currently in. Start thinking that you need to be looking further out in front of you to plan ahead for the flight path or for seeing upcoming obstacles or potential hazards sooner.

Other then that you did well and are on the right track to become a more skilled pilot over time. The quad could use a little more tuning as its standard betaflight stock sloppy / lazy flight but that will also come with time as you learn more and understand more about how it all works and what to change to make it fly a little tighter and safer having more precise control.

Oh yeah.. and you are one of those guys I hate / envy being able to stand up and fly and not falling over. I have to sit most times when I fly or I start doing the Stevie Wonder bob n weave and fall over.

 

[cranialrectosis]

Start thinking that you need to be looking further out in front of you to plan ahead for the flight path or for seeing upcoming obstacles or potential hazards sooner.

This^^^^

I find that if/when I am thinking ahead and seeing ahead I fly better, faster and safer. You want to anticipate how your actions will impact others and your flight. Don't look at the copter. Look at where the copter is going to go and how that positions you for the next move.

Flying is like chess or driving a semi. If you are reacting you are losing.

That is beautiful work btw @Tammeo. You should be proud of your accomplishments. If this were easy it would be common and not nearly as cool.

 

[Tammeo]

pilot question:
I cannot control the drift on U shape turn.. I understand that I should compensate the “forward” momentum which continue to push me away from pivot point (after the 180deg yaw) but I cannot understand how to do it effectively... so every Uturn I find myself with right stick full forward waiting for thrust to compensate inertia if I started the turn with too much speed.


setup questionS:

1. Is it normal to see “low voltage” warning during high speed maneuvers even if I have a capacitor installed on battery wires to compensate voltage drops?
2. is it normal to have 5 to 7 mins of flight time with my setup using a 4S lipo 1300mAh C100?
3. is it normal to see 70A of max current on the OSD telemetry report at the end of the video?

training question:
Do you think flying the same drone in third person view will be useful for my “pilot skill” or do I have to fly in FPV mode only?

@PsyBorg
U r right. I do not want to be “that kind o guy” it will not happen again
When u speak about sloppy controls are you speaking of PID tuning.

@cranialrectosis
This project was one of my hardest one: it started a year ago with me building and improving a cheap 3D printer to be able to design and print my own quad. No experience at all with RC quadcopter, FC, betaflight, l’iPod, etc... therefore I am super excited of my achievements

 

[PsyBorg]

Quads unlike fixed wing Require coordination on all 4 channels when not flying in assisted (auto level) modes. To confuse things even more for your brain the faster you go the more tilt happens where at 45 degrees pitch, roll and yaw axis are equal to make a turn. THEN as you go even faster and the pitch axis exceeds 45 degrees roll and yaw axis ratios start to reverse to where yaw becomes roll and roll becomes yaw.

Its best to sort that out in your brain by holding the quad in your hand and physically changing the angles to get the visual proof your brain will need to understand better what happens there.

Set up answers..

1. Yes that is very common with lower C rated batteries and is called sag. You should not however be seeing that a lot on a 100c battery. If you do there are one of two issues happening. main one probably being the current sensor in your particular setup is not properly calibrated. That or you are over propped. I will go more with the sensor issue as you have not stated super hot motors as an issue. It could go back to tuning as well to a degree. More on that later.

2. Four to seven minutes on a 1300mah pack can be done regularly with cruise type flying and good motor / prop combos. The more aggressive you fly the more current you pull so that equals shorter air time. I average 6 to 10 minutes and have gotten close to 12 minutes actual air time on my 2inch Gremlin doing docile flights before.

3. 70 amp max. Yes it is very easy to see numbers like that depending on the flight and how aggressive you fly. I have seen numbers approaching 150 amps on normal flights where I am pushing hard. Another thing that can spike max current at the flight log page is air mode. If you bump something in flight or land with it on you can see strange max currents that will drive you bonkers.. I have seen spikes over 180 before and it racked my brain until I realized that it was air mode spiking the currents on hard landings when I tried to land with super high camera tilts. Now I just lay the quad sideways so I can see the ground and disarm and let the quad flop on the ground so I dont fry something doing that.

Training

Nothing is better for skills then actual stick time. Where most people go wrong early is constantly changing things. Anything from new builds to different props to tuning changes. Flying in the way you intend to fly is best. Keeping everything as consistent as possible builds muscle memory and a "feel" for the aircraft. You will never get that if you are constantly changing things. Its best if you have a quad that flys ok and is not over heating to fly that for a while before trying any other changes.

Being aware you need the changes is good so while you are flying think about what is happening and get used to identifying things like propwash and when it happens or washing out in hard turns like you have already experienced. THEN after you have managed to adapt to those things and understand how to lessen those effects as much as possibly by adjusting how you fly you can start to make changes. For instance prop wash and shaking after turning or doing a trick. Training your mind to think things like not falling straight down after a flip or a loop and easing back into forward flight slowly and smoothly to stay out of it to begin with. Or in turns to steer a little tighter and smoothly apply throttle a bit more to turn the craft harder and constantly keeping out of its own turbulence.

tuning.

You are correct. The sloppiness and washing out in the turns is due to weak pids. They are low at default for a larger range of gear to be safely be put in the air the first time without frying esc's or cooking motors. This will also contribute to some slightly higher amp draws as the escs fluctuate more to make the motors control the craft with in a wider set of parameters. If you are not experiencing heat issues I refer back to my comment on consistency and suggest not messing with them for a little while thus giving yourself more time to get the basics down where its all second nature. Other wise you will end up chasing a tune as your skills progress and become one of the obsessive people (like my ocd self) that try and remove every little shake and bobble and not focus on the fun of the hobby which is why we start in the first place hehe.

Anyways I am impressed with what you have learned so far as well as your flying at this stage. Keep getting the stick time. keep learning and asking the right questions and you WILL be a better then average pilot in six months to a year if you focus and stay consistent. Dont forget to have fun in the quest for perfection.


EDIT: Also If you are maxing stick deflection in turns and not being able to get them as tight as you need you can adjust rates to help with that. Its all ratios and proportions. Higher RC rate is stronger turning, Super rate is softening the center of the stick like Expo on fixed wing. NEVER use expo on a quad. (I know many will argue this) but heres the thing. It will help soften center stick and make it more smooth and predictable. However on those times when you need to be at full deflection the changes are violent and not consistent. Use super rate as your "expo" to get the feel then adjust RC rate to strengthen or weaken turning ability to match the type of flying you wish to do.

 

[cranialrectosis]

TLDR: Go slow. Post pics as you go. You will crash and destroy your copter, many times. Be prepared. We love flying things and obsess over them. Come back whenever you need for help.

 

[outthayr]

Here is my 1st flight in acro Mode...

Great for a first flight. I would try removing the OSD lines / horizon. I found them helpful at first, but later they started distracting me from actually just flying - especially in acro.

 

[Tammeo]

Here I am Again to update you on my FPV experience.

since last update... I've crashed, a lot!! as u foreseen...
I've reprinted the frame, many times but I had to face vibation and temperature issue, I've smoked a couple of motors and esc too...
then I began to study both PID tuning and filter settings... I've bought a carbon frame just to see the difference compared to the one I've designed and 3D printed... and I've found out you were right: My design was not stiff enough to avoid resonance in the range 80 - 150 Hz.
at the end, in order to fly with my 1st design, I had to enable both notch filter on D and rpm filter to "dampen" vibration which (in my opinion) were caused by suboptimal geometrical tolerances (wrong planarity & parallelism of control surfaces, poor stiffness etc...)

so I've decided to re design the frame, to make it smaller in size (from 250 down to 220) and stiffer (increasing thickness and using screws as pushrod to compress the 3D printed parts) , I've opted for a F405 STD Matek FC and a 4in1 ESC BlHeli32 60A coupled with all the other stuff I already mentioned in the previous post.

Here the design, and the 3rd flight I've made with this new design.
Interesting thing is that I can run the "STD PID values" without modifying the filters. (which is a big achievement) and I can focus now on my pilot skills without the fear of smoking motors anymore.
Next Step is to analyze Blackbox log and understand how to improve PID Tuning to make it more responsive while I learn some trick

thank you again guys for all the hint and tips which guided me through one of the most difficult hobbies I've ever decided to do.
At the end I've got cheap frame with acceptable performances which can be 3D printed in numbers with 30€ of PETG wire. If anyone is interested in trying my design please contact me I'll be glad to provide 3D models and Gcodes.

 

[Tammeo]

@PsyBorg @Ihichi Bolls

Could you be so kind to tell me what I have to focus next?

I fly at least one battery a day and I train at simulator (velocidrone) on the same track I fly for real (I’ve re modeled it to gain confidence)
I feel I have to continue without touching the quad setting in order to get comfortable with my quad... but I would like you to inspect my settings and point out everything that does not sound right to you. I have no local community to ask for advice so here I am one more time asking for help

Here under my set up:


and a quick video of the route I train for real and with the sim (skip the first part coz I took off in Horizon mode instead of acro...)