450 Quadcopter Frame From PolyFlex Pressure Pipe

[kpixels]

I've been flying mostly Edge Racing Light 2205 2480kv on my 5" freestyle quads and one of my baby tricopters. They are Smooth and light. Starting to play around more with T-Motors lately too. I really don't like the uber notchy over magneted motors that a lot of racers prefer. The Steel motors are supposed to be smooth, but I'm not a big fan of his. Recently went with some Tmotor instead on a Strix Screech build. Wanted low 1600kv 6S motors for this build. Playing with 6s not for uber high wattage and power, but for all around efficiency. High Voltage at low current for same wattage = less thermal loss. Also should equate to less sag. Run mostly 30A ESCs on my 5" and up quads.

Zonked out looking at motors last night. Will probably go with Tmotor, I appreciate this info very much...

 

[PsyBorg]

Can already see the changes. Nice. Going by this flight and the first one with the pid changes you made I would say you could go up another 10 on roll P and 15 or 20 on pitch P. For I gains I set mine a .4 or 40 depending what FC for pitch and roll and .5 or 50 for yaw and they are all very stable for the conditions I fly in no matter what I fly.

 

[kpixels]

Can already see the changes. Nice. Going by this flight and the first one with the pid changes you made I would say you could go up another 10 on roll P and 15 or 20 on pitch P. For I gains I set mine a .4 or 40 depending what FC for pitch and roll and .5 or 50 for yaw and they are all very stable for the conditions I fly in no matter what I fly.

Hi PsyBorg, Thanks for sharing this info. Luckily I saw this before I went to the field and had these numbers with me. Starting with version2 for P I bumped it up 10 on the roll and 15 on the pitch like you said. My roll and pitch for I were both already 5 higher than yours so I left them first time. It would probably be easier to give the different settings and how it acted each time.
last video;
P-r=50, p=68, y=70 _ I-r=45, p=54, y=45 _ D-r=30, p=35

was bouncy descending
P-r=60, p=83, y=70 _ I-r=45, p=54, y=45 _ D-r=30, p=35

Less bouncy but spins out bad at end of loops, flips and rolls (I was curious about D)
P-r=60, p=83, y=70 _ I-r=45, p=54, y=45 _ D-r=35, p=40

Could not control direction (seemed to get worse during flight) had to crash land
P-r=60, p=83, y=70 _ I-r=40, p=50, y=45 _ D-r=30, p=35

It's long story but I couldn't make changes to the pids in the field. It was still soggy so I left the laptop in the car. I should be able to change the pids by the OSD menu but I couldn't recall the controller positions to trigger the menu. Can't even find it online, I found it by accident before. For version2 I'm waiting for an video cable for the VTX.

Good news is I hadn't changed any pids for version1 before, so I started off with your numbers (+ defaults) and that thing felt pretty good compared to version2 using same numbers. I should mention that I bumped TPA to .35 on both.
To be continued...

Note: numbers were wrong first time

 

[kpixels]

Police Want My Quadcopter!


Seriously, officer liked it and said he was thinking about getting one. He was sitting in the lot and I asked him if he would mind if I fly there to test some settings and he said sure, no problem.

He said he thought it would be a matter of time that their patrol cars would have UAVs. I was in complete agreement that it was inevitable told him I thought they would be a great asset. He was a cool officer!

 

[PsyBorg]

Having the numbers is good so we can see where you are. If possible can you distinguish which is quad 1 and which is quad 2 please.

What I have learned is beta flight has a huge range of tuneable area between 20 and 80 seems to be the happy spot for the larger quads. What I think you are seeing is when you crossed that 80 mark you went from slow oscillations. to a tighter tune and now have crossed into the place where P is too high and starting to cause fast oscillations. D does in fact help with that. Stingerswarm actually tunes like that taking P up to where it starts them faster oscillations then cuts them back with D. This is fine just be aware of motor temps doing this.

We are into territory I am not 100% on for tuning an AP ship so I am going by my own general knowledge for pid tuning where I KNOW I have good control in most cases where I would be flying like with higher average wind conditions.

I noticed you have said you had issues with bouncy descents. That is more because you are dropping down thru a huge mass of disturbed air called prop wash. You can tune out a fair portion of it but there is no way you will ever be able to come straight down in a more then painstakingly slow manor without getting that on any tune. This is where most pilots that are new or rely on auto level modes get into trouble. The quad starts to get wobbly and they panic and drop even faster to get down to the ground. That can escalate those wobbles beyond what the FC is capable to process and that's when things go south and total control is lost.

Now that you are in the air in a reasonably controlled manor you need to train your thinking not only to know where you are going but where you have been as well. Descend super slowly when coming down in your own prop wash and if things get wobble hold in a hover for a second then resume OR move slightly in any direction to take yourself out of the bad air. For example. When the freestyle pilots do a power loop you notice on the back side where they are powering back up to catch themselves the good pilots always are moving forward ever so slightly. This is to stay in clean air as they apply more power to complete the move. Turning is the same, you are always moving just slightly to stay out of the prop wash for a smooth and steady turn.

Lets see if we cant smooth you out a bit with rate changes. In the screen shot to follow I have all my quads (except the Gremlin. Lower degree/second due to short wheel base). With the curve that is built in no expo is needed yet I am still able to make any quad I fly move a mm or two at a time if I want yet still be able to have a nice strong flip and roll rate. (mind you for the ap ships you are flying this is not recommended to be flipping and flopping with expensive camera gear). This should give you extremely fine movement control yet still give enough to do sharper turns when needed to avoid things or for certain film effects.

Lastly is TPA. That should not really come in to play on an ap ship as what that does is lessen pids the more throttle you use in a fast change like punch outs or hard cornering. If you get to where there is need for TPA changes you are already in a bad situation and they wont matter on an AP ship.

Soo.. when next you fly grab us some video and label each ship and the pids so we know exactly which is which so not to confuse the changes. THEN what I want you to do is set up on profile one the current best pid configuration for each quad and leave that alone. Use profiles 2 and three to make changes one at a time. Use one for a higher parameter change and the other for a lower parameter change. ONLY change one thing at a time to be able to compare them. Then you can basically swap profiles and fly for a bit, swap again fly different parameter, then fly again and fly current best to see what is the best of the three for what you want to do. That number gets used in profile one and you move to the next change in your tuning.

Here are common stick commands for making changes when not armed. I believe most OSD's are accessed by yaw center left for 5 seconds as well.



And here is a screenie for the rates to try.. if you use these be sure to change them in all three profiles or you will be in for a huge surprise going randomly from super smooth rates to less forgiving rates. Do not bother even looking at the pids as this is from tuning sessions long past and will more then likely set you back. You can how ever try the settings for horizon mode as that is set up for very smooth transitions in and out of forward flight with out the having to fight the FC to get moving and the sharp snap back when returning to center stick that angle mode has.

All I do is change RC rate between 1.9 and 2.05 for faster or slower degree per second and keep the Super rate to hold that soft curve and feel for center stick ranges.



Finally... Police.. 99% of them are cool. IF when they approach you and you dont tense up like you know you are doing something wrong, seem defensive, are mouthy and not respectful. If you greet them with a polite hello and how are you today sir then most are put at ease you are not a jerk and will act accordingly.


Sorry for the book post its a lot to process but hope it helps.

Changed to say keep Super rates the same and change RC rate.. sorry used KISS labels for those values.

 

[kpixels]

Having the numbers is good so we can see where you are. If possible can you distinguish which is quad 1 and which is quad 2 please.

WOW PsyBorg, Thanks! I'm reading your reply now just wanted to say off the bat, all those sets of numbers were for version2.

This would have been the numbers for version1
P-r=60, p=83, y=70 _ I-r=40, p=50, y=45 _ D-r=30, p=35
from default;
P-r=40, p=58, y=70 _ I-r=40, p=50, y=45 _ D-r=30, p=35
But it flew good. I'm working on a video compilation with settings in captions.

 

[kpixels]

Lastly is TPA. That should not really come in to play on an ap ship as what that does is lessen pids the more throttle you use in a fast change like punch outs or hard cornering. If you get to where there is need for TPA changes you are already in a bad situation and they wont matter on an AP ship.

I misunderstood what TPA does, I was going for softer. I put back to default of .10.

Soo.. when next you fly grab us some video and label each ship and the pids so we know exactly which is which so not to confuse the changes. THEN what I want you to do is set up on profile one the current best pid configuration for each quad and leave that alone. Use profiles 2 and three to make changes one at a time. Use one for a higher parameter change and the other for a lower parameter change. ONLY change one thing at a time to be able to compare them. Then you can basically swap profiles and fly for a bit, swap again fly different parameter, then fly again and fly current best to see what is the best of the three for what you want to do. That number gets used in profile one and you move to the next change in your tuning.

This video includes all the changes made to version 2 so far;

I'm uploading the version1 flight footage now which starts out with the same settings as version2 last clip...

Finally... Police.. 99% of them are cool. IF when they approach you and you dont tense up like you know you are doing something wrong, seem defensive, are mouthy and not respectful. If you greet them with a polite hello and how are you today sir then most are put at ease you are not a jerk and will act accordingly.

10-4!

 

[kpixels]

Here are common stick commands for making changes when not armed. I believe most OSD's are accessed by yaw center left for 5 seconds as well.

I just remembered that the instructions for the OSD menu I stumbled across before shows up (very briefly) on the FPV screen when the FC boots up. This is different than stick commands because you can access all of the settings that you can in CleanFlight via OSD.
Thrust=Middle, Yaw=Right, Pitch=Full (for Illuminati32)

And here is a screenie for the rates to try.. if you use these be sure to change them in all three profiles or you will be in for a huge surprise going randomly from super smooth rates to less forgiving rates. Do not bother even looking at the pids as this is from tuning sessions long past and will more then likely set you back. You can how ever try the settings for horizon mode as that is set up for very smooth transitions in and out of forward flight with out the having to fight the FC to get moving and the sharp snap back when returning to center stick that angle mode has.

All I do is change RC rate between 1.9 and 2.05 for faster or slower degree per second and keep the Super rate to hold that soft curve and feel for center stick ranges.

We have crystal blue skies right now so I'll back off the changes to version2 (not taking laptop) to something in the middle, and I'll be able to tinker with version1 settings in the field with OSD.

Here is the flight of version1 after first changes;

Changed to;
P-r=60, p=83, y=70 _ I-r=40, p=50, y=45 _ D-r=30, p=35
from default;
P-r=40, p=58, y=70 _ I-r=40, p=50, y=45 _ D-r=30, p=35

Thanks,
Ken

 

[PsyBorg]

Ok. It may be time for version 3. Roll is looking pretty unstable still and possibly worse from the original. There may be some twisting action going on with the frame I think. I also think I am still not on the same page as you with this build. Is this meant to be an ap ship to be flown like a phantom for video or a freestyle ship as you keep doing flips and hard maneuvers?

In any case it may be time to get a wider center section and maybe think about one if not two more center tubes. It is obvious this has plenty of power for the weight its carrying so I don't see that being a factor just yet. Pulling the camera back more towards center will help on Pitch tuning and you can raise it up a tad more to keep the props out of view if you have to but getting it more stable is the first goal.

 

[kpixels]

Ok. It may be time for version 3. Roll is looking pretty unstable still and possibly worse from the original. There may be some twisting action going on with the frame I think. I also think I am still not on the same page as you with this build.

I'm down for version3! I got this distracting idea last fall while standing in Lowes rounding up stuff to remodel my bathroom. I had three boxes of motors sitting on the shelf that now fly (I'd say) and building these got me thru the crappy winter. For me, the main point is for exploring possible camera platforms.

Is this meant to be an ap ship to be flown like a phantom for video or a freestyle ship as you keep doing flips and hard maneuvers?

No, I have a Phantom4 and 2 Xiaomi Mi rigs but they are good for cinematic look. I also like having the capability of a more organic look, like from a bird with tilting horizon and all, but more of a cruising or soaring (exciting) feeling that I'm ultimately after. That's what got me into tricopters and this actually started with a tricopter build.

I would like to offer a range of footage capabilities including from a proper freestyle 5 inch. I realize that I will never get that 5" like freestyle footage out of these boats. You sounded like you wanted to push them and try and break them. I was just looking at it like if it will do flips, rolls and loops then why not.

In any case it may be time to get a wider center section and maybe think about one if not two more center tubes. It is obvious this has plenty of power for the weight its carrying so I don't see that being a factor just yet. Pulling the camera back more towards center will help on Pitch tuning and you can raise it up a tad more to keep the props out of view if you have to but getting it more stable is the first goal.

The camera booms and mounts are my own design I've been working on. They're intended to eliminate the props and prop shadows from the footage (and jello and vibes). I was thinking, since I would never fly without the camera, I should have placed the FC at the center of mass with camera mounted. I would probably change the FC placement if I built another.

I got much better flights today. It was cold but a beautiful sky. I somehow pulled some rolls while diving! Haven't looked at footage yet though. Ken

 

[kpixels]

Ok. It may be time for version 3. Roll is looking pretty unstable still and possibly worse from the original. There may be some twisting action going on with the frame I think. I also think I am still not on the same page as you with this build. Is this meant to be an ap ship to be flown like a phantom for video or a freestyle ship as you keep doing flips and hard maneuvers?

In any case it may be time to get a wider center section and maybe think about one if not two more center tubes. It is obvious this has plenty of power for the weight its carrying so I don't see that being a factor just yet. Pulling the camera back more towards center will help on Pitch tuning and you can raise it up a tad more to keep the props out of view if you have to but getting it more stable is the first goal.

Actually, I should probably wrap up this project. It is what it is.
I've gotten more than I expected out of this experience but don't have much time for it.
I really appreciate your support, and I'm glad to be a part of this community,
Ken

 

[PsyBorg]

Bummer. This has a good potential for making cheaper AP ships as well as maybe scaling down to trainer level freestyle quads with a little more effort. Well you can always come back to it over time. Specially after learning how to set them up without me pestering you.

 

[LitterBug]

Ken,
We need to meet up sometime and compare notes. I'd really like to see those puppies in person. I'll let you try out some of my builds so you can get an idea of where you want to go with your freestyle build. Weather was great this weekend, but got 0 packs in outside. Got some indoor time in Friday night tuning the 6" Mini Tri and some 2" builds. Wanted to get out Saturday, but spent a good part of the day trying to figure out what was wrong with the FPV on my AWK210 RunCam Split build. Turns out, it was the RunCam Split.... Really need to get out tonight before that next line of cold and rain hits late tonight.

Cheers!
LitterBug

 

[kpixels]

Bummer. This has a good potential for making cheaper AP ships as well as maybe scaling down to trainer level freestyle quads with a little more effort. Well you can always come back to it over time. Specially after learning how to set them up without me pestering you.

Thanks PsyBorg, it's cool.
Anyone can pick this up if they want and I'd be glad to help if someone tried and got stuck. Another reason for doing this besides the camera platform was to familiarize with various flight controllers so I needed more knowledge about pids anyways, which will be ongoing. To me it's motivating but if you call it pestering, it was a good kind that I'll benefit from and can appreciate.
Ken

 

[kpixels]

Ken,
We need to meet up sometime and compare notes. I'd really like to see those puppies in person. I'll let you try out some of my builds so you can get an idea of where you want to go with your freestyle build.

That would be AWESOME!

 

[kpixels]

Poly Quadcopter PIDs And Camera Boom/Mount

Polyflex quadcopter version2 after some tricky PID tuning (still tuning) and modified camera mount. The camera boom is now removable and most of the jello and vibes have been eliminated. Flights of this intensity last more than ten minutes with a 2800 mah (224g) battery. Had two flights in this tight area with no close calls. Still learning to fly in advanced modes...

 

[PsyBorg]

PROGRESS!!!

You are starting to get the hang of tuning now. Seems you are figuring out what it takes to tame one of these beasties. Flying skills are progressing as well. A lot smoother in the coordinated turns. the more you fly in acro the faster you will progress and the smoother you will get. Keep it up mate.

 

[LitterBug]

Glad you made it outside. It was sunny when I got home, so put a pack on to do some 6S testing on my F550. As soon as the charger beeped that it was ready, it was snowing hard... :-(

Gotta love OH this time of year. LOL!

Cheers!
LB

 

[PsyBorg]

Shame your weather is not like ours.. what they tell you here is.. "If you dont like the weather wait five minutes it WILL change." I have seen all four seasons happen in the course of less then two hours. Its the aliens at area 51 doing their government funded weather experiments I tell you!

 

[LitterBug]

Well, actually.... Ken is west of me, and I firmly believe his flight was to seed the clouds and make it snow on me!

Cheers!
LitterBug