Testing the Apex theory - Resonance management

The final frame: bit.ly/3j0gcRW
Table of contents:
00:00 - Intro
00:24 - Basic explanation of vibrations
02:00 - Apex frame theory
03:03 - The frame past design changes
05:00 - The test frame designs
07:57 - Results of the first pair of frames
09:57 - Alternate frame design to test
10:36 - Results of alternate frame design
11:52 - What I did with this information
12:28 - What I learned?
14:07 - Final thoughts
14:54 - Example of how unpredictable vibrations can be
Chris Rosser: • ImpulseRC Apex Vibrati...
This is someone you want to subscribe to
Stay tuned to www.FPVCycle.com. We've been working extremely hard on a bunch of new things.
▼Join the FB group for more discussion: FPVCycle - Kabab FPV
/ 379155946182689
▼ Tips?
Patreon: bit.ly/2oGLP9b
$1/mo from 1/4 of my subscribers will make me almost quit my job. Thank you
-PayPal: www.paypal.me/kababfpv
-Bitcoin: 1E4XZXoD4rS6MYWVWuenY7Kw2M1YgyNpoQ
-Nano: xrb_1188ek5bd7tb9kw67asnp9o65byuoaxbtpxuznx1ribf75x78awywmakj55f
-ETH: 0xD067F93811f2eC31CB1928901002dfdf9A492EF5
-LTC: La2KWYAjp4VT2Lq2M76pbjzXWvRyjpfCtG

[@Kabab]

Table of contents: 00:00​ - Intro 00:24​ - Basic explanation of vibrations 02:00​ - Apex frame theory 03:03​ - The frame past design changes 05:00​ - The test frame designs 07:57​ - Results of the first pair of frames 09:57​ - Alternate frame design to test 10:36​ - Results of alternate frame design 11:52​ - What I did with this information 12:28​ - What I learned? 14:07​ - Final thoughts 14:54​ - Example of how unpredictable vibrations can be Chris Rosser: www.youtube.com/watch?v=u5bTa...​ This is someone you want to subscribe to Stay tuned to www.FPVCycle.com. We've been working extremely hard on a bunch of new things. ▼Join the FB group for more discussion: FPVCycle - Kabab FPV facebook.com/groups/37915...​ ▼ Tips? Patreon: bit.ly/2oGLP9b​ $1/mo from 1/4 of my subscribers will make me almost quit my job. Thank you -PayPal: www.paypal.me/kababfpv​ -Bitcoin: 1E4XZXoD4rS6MYWVWuenY7Kw2M1YgyNpoQ -Nano: xrb_1188ek5bd7tb9kw67asnp9o65byuoaxbtpxuznx1ribf75x78awywmakj55f -ETH: 0xD067F93811f2eC31CB1928901002dfdf9A492EF5 -LTC: La2KWYAjp4VT2Lq2M76pbjzXWvRyjpfCtG

[@zenith9671]

This hobby is turning all of us into scientists

[@Kabab]

😂

[@WKfpv]

That's what most hobbies do 😉

[@redbeardthepilot3168]

I believe you meant to say “Aerospace Engineers”, and yes.

[@minkorrh]

You wish. Even with a background in electronics and a better than average grasp of physics it's doing nothing but lightening my wallet.

[@erikbrown7927]

@@minkorrh most people don’t have a background in physics and an understanding of electronics. I do purely because of this hobby. I wouldn’t underestimate the ability fpv has in turning average people who like drones into people who now have the two things you mentioned

[@ChrisRosser]

Awesome video! It's wonderful to see theory and practice align. I'm feeling like this is just the beginning 😁

[@aakashjana6225]

And I am feeling like I don't deserve to build quads anymore coz I am not even a quarter educated. Just a computer engineering student.

[@charlesmarseille123]

Well, your damping theory has a good start ;) Though, the comparison between tf3 and tf4 lacks the mention of the different base plate distance from center to corner holes. This must change the fundamental modes of the arms, thus shifting frequency.

[@444haluk]

I love the fact hobbist are thinking about hypothesis, testing results and sharing the lessons. This is top-notch. You can even write a paper.

[@Director-M]

You're one of the most valuable members of our community. Thank you for being not only motivated, but also *selfless*! Huge inspiration..

[@Asu01]

I've been into contact with Chris as well (I was the first to reach him when he made the Reddit post because we share the same interest and I want to give some suggestions on his frame), and we've discussed quite a few things about frame design and resonance. He even made an analysis on my frame! (though the blackbox data is flawed because the bolts weren't tighten) I'm glad that you acknowledge his theory in this video. About his theory on Apex's mounting approach and how it affects frame resonance, I had some doubts regarding its effectiveness and your results further confirm that. In my opinion, the Apex can be so good is because it "cheats" its own way to achieve rigidity. Apex has squished-X geometry, that means they can lower motor to motor distance which results in higher resonant frequency (at cost of flight characteristic). Apex also has very beefy arms (5.5x12mm at narrowest point), which suppress vibrations effectively (at the cost of weight). All of these are pretty much the exact opposite approach of your Glide frame (true-X and lightweight). Both of these frames have different objectives and so it's not exactly comparable, but it's a good example of what you need to sacrifice in order to gain something. What I've learned from my experience is that to achieve good resonance characteristic, it comes down to 2 main things. A good frame should have adequate arm coupling. Frames are manufactured with tolerance, and even the slightest of wiggle will result in undesired vibration (we're talking hundred of Hertz here). This can be achieved by increasing contact area, or using some sort of interlock mechanism where all of the four arms rest on each other. The usage of lock nut instead of press nut helps to improve arm coupling, because you can tighten lock nut much tighter than you can tighten hex screws with press nuts (I've been avoiding press nuts with my recent frames). A good frame also should have suitable cross-sectional arm area for the application, with higher ratio being better at suppressing roll and pitch vibration and lower ration being better at suppressing yaw and torsional vibration. It can be tricky to strike a good balance between these four vibration characters, but you can cheat it by having arms that are wider and thicker, or by adding braces and such. These cheats come at the cost of added weight and complexity, so it really should be the last resort in your frame design. His video on my frame: kzbin.info/www/bejne/p2PXZ2SXiM-dbK8

[@Kabab]

Yeah I never made the glide with the intention of dampening. It's all durability, weight, value and workability. The prototype 5 has some friction built into it and it does do better but only about 10% better. In this case I present in the video, I feel like it's about 10% better with the apex mount as well. So maybe it's just that, a 10% improvement......BUT, 10% of an enormous spike is gonna be super valuable....

[@tylerelliottFPV]

I am sooo happy to see this video. I haven’t even watched it yet, but have been convinced as of lately I need an Apex frame and seeing Mr. Kabab mention what’s in my head is perfect.

[@friendlynomad9840]

Awesome video! I'm happy to hear you're working with Chris. You guys are entertaining. The passion is obvious. :D

[@BrandonBeans]

I'm loving watching this whole dynamic play out. SCIENCE

[@PropagndaX]

“It’s in a tree now” Story of my life 😂😂🤘🏽

[@wyvern4588]

I know! If we could discover how to make quads not get stuck in trees, we will be truly at the pinnacle of human evolution. Right now, when I get stuck in a tree I revert to cave man and throw sticks and shoes at it- unga bunga.

[@PropagndaX]

@@wyvern4588 UGH UGH 😂

[@sleepingbearffg5008]

Lol love your sense of humor.. God bless

[@MSR.]

Love the testing and theory discussions, thanks 🙏 Bob 👍🤘👊🙌

[@strawhatsam]

Really smart idea of using the two screws to see if the theory makes a difference. Coloumb damping (damping by friction) seems like a very elegant way to reduce the amplitude of vibrations given the mechanical nature of these frames. My concern though is, just as you said, making the frame as rigid as possible so as to push those frequencies as high as possible so it's easier for the FC to filter it out. But if there are ways to introduce friction damping without reducing frame stiffness, then you can cut down on the tonality (spikey spikes) of the noise and spread it too. You're totally right about preferring low amplitude and wider spread of frequencies. I think this route is much easier to tune out and is healthier for the ESCs. Although it's the same amount of raw noise/energy coming from the motors, if you can spread that energy out like butter on toast then you got the awesome sauce!

[@MCsCreations]

Fantastic, Bob. I've never seen this level of testing in this industry... Top of notch! Stay safe there with your family! 🖖😊

[@WEEBER13]

LOVE this content! Can't wait to see you JB's channel

[@MrShagiFpv]

That " Vibration analysis" was so funny. Looked like the drone was swimming.

[@lawrencemartin8205]

Awesome video. Thanks. Im the type of buyer that just wants to buy a good frame etc.. but im also wanting to try to learn along the way also and learn something new understand why something is the way it is. So thank you to both you and chris.

[@Dangeresque14]

I love that you're willing to test this stuff regardless of the results. I could never get an original 65mm toothpick to fly smooth. It was vibration central. But, I have a 6s power pick 5inch that flies phenomenally. I like that you always work to improve! This was super interesting and as a prior owner of the original, I can't wait for the new "tooth fairy".

[@donphobos]

CAE for the win, nice to see that you get help with these issues.

[@onemanmob6756]

I am excited to see what comes out from this co-operation - i.e. your future designs. Also - flying exclusively micros, I'm curious about the resonance performance of the BT and TP3 frames

[@DutchRC]

years from now archioligists will tell people these were musical tuning forks ;) Interesting stuff though and for a Second there I thought you were going to advocate pre-build quads ;) that would be a nice opportunity to really drive this home

[@davehrcfpv]

Intriguing. Keep up the great work! FEA on a quad - it works for race cars and bridges!

[@JaraFPV]

0:19 That!! Is how you actually test for what you can call the "BEST" frame! It's not just designing .. It's also how it will perform in crashes

[@mikebergman1817]

Bring on the Toothfairy Apocolypse Edition. Half of us would probably fly to cali, just to climb that tree and buy the one thats currently still stuck.. This video was super interesting, and I had a feeling when I started watching Chris's channel a few weeks ago, that you would be the one to bring actual testing and further development, based on the analysis. I'm really excited to see how this goes! I'm curious if this can be as easily applied to smaller platforms, like the TP3 and Babytooth..

[@Crftbt]

KababFPV, what a legend!

[@hateeternalmaver]

Sounds like it's recorded in a construction site that's inside of a military base that is doing an exercise *while* crossing the ocean on a huge battleship... =D You fought through it so hard and you make it barely even noticable. Loved the 'lower amplitude over broader spectrum rather than higher spikes on narrower bandwidth' conclusion. Thanks man.

[@Kabab]

🙈

[@MPQuads237]

Good talk again. Very informational. Thnanks.

[@hojiqabait]

Main vibration caused by joints i.e. fasteners. Major problems are mostly on hardware precision and standards. Thicker bottom and top plate with countersunk always help for tight fit. Resonances are not always a bad thing. Dampeners can reduce vibrations easily, but give less response to gyro. I go for 195mm unibody X. 😉

[@SneakyB]

Interesting. I might send some of my designs for analysis :)

[@southwalesdrones281]

Seriously cool nerdie stuff! Love it

[@PropagndaX]

If you put the frame close enough to your ear, you can hear its resonant frequency

[@user-hg9ew4in8j]

Quite useful video.. great

[@uavtech]

Exciting to see folks raising the bar and doing more and more engineering analysis on FPV products!! 11:30 - I would look at a waterfall graphs of the noise vs. throttle too. Also run through PID Toolbox as the spectrograph can be deceiving at times since cumulative when comparing logs (flight length, etc...) Of course, variables in BF need to be the same; which I assume are since you are just moving equipment. 12:30 - WHAT WE LEARNED? Proper Engineering is not working by feel or the seat of your pants! Anyone not saying "AMEN" to that statement, yeah, I'm talking to YOU.

[@Kabab]

Proper engineering in my mind is working with both the real world and the simulation world. You can't have one without the other and both are equally important. I choose not to show the waterfall graphs here because they're harder to read for most because it's not as obvious if you're not used to reading them.

[@uavtech]

@@Kabab 100% agree. Simulation just helps us not do so many empirical tests. The fact is empirical testing is so costly and time consuming we short-cut and don't get to the best solution. Simulation is so easy and fast to tweak parameter, you can quickly come to the best solution and then empirically test it. It is awesome you, in partnering with Chris, are bring finite element analysis to FPV frame design; especially when trying to make the lightest frames. This will save you so many headaches.

[@ChrisRosser]

@@uavtech I would add that simulation can sometimes show things that are very difficult to measure in real life. I would think it would be very challenging to visualise these mode shapes of vibration through any real world test because the displacements are submillimeter, and there isa strong possibility that adding sensors would change the response due to the extra mass on these small systems.

[@uavtech]

@@ChrisRosser great point!! Hopefully you guys can weave this into to convo with JB today. There is a tendency sometimes in FPV for folks to attack tools (like Blackbox) because they don't like what it says or they don't understand it. It is a tool (like a ruler or laser level), that shows you things clearly that are very hard to discern in just flying alone; again - like a laser level it is very accurate at measuring real-world to help show you what is actually happening (as you know).

[@ag.page.]

Hey Bob. Just to know- when you said you transferred everything from one frame to another did you transfer the exact same props too? Your prototypes you gave me vibrate more than other props I fly, they fly so nice tho. So much control for acrobatics and fast too. 👍🏻👍🏻👍🏻. Cant wait to buy the final product

[@spencerthorp]

You are so cool! Keep it up!

[@gmivisualsjason3729]

Excellent

[@DriftaholiC]

We need custom autoclaved carbon fiber. Tubular arms etc. Carbon is cool because it resists stretching and compressing very well, the thick plates don't do as much for the rigidity as something hollow would do.

[@PortalFPV]

I have been really enjoying designing my own frames for the past year. The way you design is so far beyond what I ever even think about o.O

[@JLfpv]

I just saw Chris's channel this am. Awesome stuff. Hope to see the quietest frames we've seen yet coming out. My rooster is NOISY. but I still love it. Been my basher for yrs now.

[@JazzXP]

"Strapping motors to 4 chopsticks" I kinda wanna see that! Great analysis though, definite food for thought.

[@marc_frank]

basically tubular los frames

[@mikev2116]

If you had vibration dampening gaskets or tape you could maybe isolate those "bad vibes" even more favorably.

[@Kabab]

I'm thinking of something like that but the issue with any rubbery material is that it induces bounce. We don't want any bounce, we need friction. So I'm thinking of a sort of shim...

[@Plur307]

I was commenting over 2 years ago for you to look into "ekoa" composites. They are like carbon fiber but made of flax fiber. They have amazing vibration dampening properties. I have built tubular frames using ekoa from rockwest composites and i think they fly amazing. Only the outer few layers are ekoa, it is still mostly carbon fiber.

[@Kabab]

I just looked it up. It's damn expensive and only seems to come in tubes. Using tubes for arms on a production frame is a bit difficult. Hard to stop them from twisting and if you drill holes through them they tend to want to fracture

[@Plur307]

@@Kabab I agree using tubes is very hard for a production frame. I just wish the material could start being used in multirotors as it seems like a perfect application.

[@TheEviling]

If you're feeling like trying something odd, you could try make pyramid-shaped bars on the frame. Basically moving the frame away from being a flat thing to a more three dimensional thing. The best way to remove flex is usually to add support in the direction of the flex.

[@Kabab]

The 3D frame design idea is definitely good. The issue is that it's very very hard to beat flat carbon in terms of durability, function, workability, ease of use, cost and overall value. You might get marginally better flight performance but you'll still have all the vibrations to deal with and you'll have spent heaps of money and effort getting it out the door. It's definitely something I hope to see more of in the future but to me it doesn't make sense right now.

[@verdi6092]

@@Kabab just a thought, how about sandwhiching a hard rubber in between the carbon arms or the frame? Like a tennis racket material

[@Kabab]

@@verdi6092 the issue with anything rubber or rubber like is that it's springy so it introduces bounce which can make things a lot worse.

[@verdi6092]

@@Kabab then maybe u could use a titanium i beam like for the arm

[@Kabab]

@@verdi6092 anything like that just needs to be tested. Maybe it's a better way to go but it's super hard to simulate

[@gavinvales8928]

this becomes especially important with 7 inch builds

[@oneistar6661]

Push, push and it will crack. Eagerly awaiting the times when the resonance problem will be solved and we're all spend more time flying than adjusting and filtering. Thank you.

[@Kabab]

I've got a rather unique concept I'm testing in a week or so when the samples arrive. This may actually resolve these issues pretty much completely....when you see it, you think, 'oh yeah no duh that will work'. So I'm very hopeful

[@David_Nielsen]

When you doubled up the arms to 8mm thickness, do you mean one arm cut from 8mm carbon or two 4mm arms pressed together? I think based on the theory the latter would introduce a lot more surface area for friction to occur when the frame flexes, and possibly add to the damping? Or maybe the normal force out in the middle of two pressed arms just isn't enough to create significant friction forces because there's no nearby bolts, so all you would get is significantly weaker arms that resonate at a lower frequency? If you didn't already, it might be interesting to test (maybe even 3, 4, 5 + separate arms pressed together)?

[@Kabab]

It was 2 sets of arms double stacked. In this case there was some dampening but hard to say if it was from friction. You need something thay is rigid and has movement in the right places. I don't know if a double stacked arm would have any movement in the direction we need

[@David_Nielsen]

@@Kabab hmm, I think you're right. I was also forgetting that friction only depends on normal force and friction coefficient, and it doesn't care about surface area.

[@michaelvonfeldt9629]

Yeah!

[@0ADVISOR0]

Why does the flex analysis go through the battery? Shouldn't be the battery considered as a separate item?

[@Kabab]

It's strapped to the quad so it's considered part of the structure. He puts appropriate weight motors and battery on the frame.

[@Quick-Flash]

You ever thought of using rubber where you mount the motor arms? So that it rubs against some sort rubber instead of carbon?

[@Kabab]

Yeah I've tried tpu and Corey/mewo has tried various materials too. It tends to increase amplitudes. You specifically need friction, not bounce.

[@mouseFPV]

Glad to see you looking at this. Better filters on the software side, and less need for them on the hardware side, is where we have room to improve flight performance in a measureable way. Would love to see how motors affect the frame performanc. Dig up those xing Cyber motors and do a 5s throttle sweep on a Prototype. Then pop your 5s motors on a Prototype frame and compare the bb logs or pop them into pid toolbox. Would love to see the real frequency analysis.

[@happyvalley808]

@kababfpv just for grind you should soft mount the motors to see where that takes us? The black box will tell all

[@Kabab]

Yeah I'd rather skip that. Even if it helps a lot I wouldn't do it. Such a pain to manage

[@chrismillersrcfamily]

Just a thought I’m sure you thought about it already. Cross bars on front and back arms? Not side to side. Like the CMW shocker? So would that change the vibration on pitch? Sorry for all the questions. I was listening just in case there was a quiz at the end. 😉

[@Kabab]

Good questions, hard to answer. The solution in my eyes is finding more ways to dampen rather than other ways to stiffen. We can already make stiff and tune the frame for a setup. Damping however lowers the amplitude of all

[@valden]

Hey Kabab, I have some airblade 1905 2400kv and some lumenier chief 2004 2600kv. Think either of these would work ok for a light analog build of the Toothfairy 5 inch?

[@Kabab]

Yeah if you want something 250g, those motors would work. Use a 500-700mah 4s to keep it under or around 250g. You could run up to a 1500mah 4s however

[@alfskaar489]

This is currently my favourite "The Phat Ultralight FPV Racing 5" Frame" if you take a look at the how the arms are keyed together and the stiffness of the frame you will understand why. :-)

[@Kabab]

The arm mounts on that look similar to my Floss 2 way back when. The frame setup is also quite a bit different than what I've proposed here. What I've got is a top mount battery with a vista and is designed to run 2204 and 23XX size motors. Very different animals. That racing frame looks good however

[@yusofhelmie5058]

Hi, would it make a difference if the fc screw goes all the way to the top plate as well

[@Kabab]

It wouldn't make a difference but it would make routing the battery strap a lot more difficult

[@wyvern4588]

Interesting. I designed a few frames- stretched plus frames. First design was a simple 2 piece construction, main plate shared the fore and aft motors, and the bolt on "arms" on ether end were one piece, bolted to the upper third of the main frame. Immediately I knew I had a noisy quad- sure enough roll axis was crazy noisy, no doubt there were harmonics going on. I went back to the drawing board and now I have two (rather 3), the first is a multi-piece, 4 short arms on a wide plate- I have yet to test, but the second version of the mono frame I have is abnormally quiet, to the point where I thought the black box log was broken- nope, swapped out all different motors, props esc and FC, same story, almost no noise whatsoever. Only problem is I valued strength and resonance over sheer weight- at 4mm thick, it's about 90 grams with hardware- not a tank, but not light- so next design is going to be thinner with additional arm reinforcement plates to beef them up- as loosing an arm means the whole frame is junk. I still have to try the multi-peice frame, but I'm not sure it will be good, it's ugly, heavy at 105 grams and incorporates sandwiching the arms between the plates that may have an inherent resonance flaw- unless the geometry of the arms play a huge roll in this frame, I expect there to be much more noise. This certainly gives me some Ideas to consider when I make a final revision to the multi-peice frame, or, If I decide to go with an X configuration quad (I sure do love my + frames though!)

[@Kabab]

that's one of the issues with an analysis like this. You want other features too and we have all these filter tools so maybe it's not such a bad thing to favor other features and deal with a little noise.

[@wyvern4588]

@@Kabab True, It takes quite a bit of noise for the quad to start to show signs something is wrong, the popular MPU6000 gyro is great at throwing out bad data, but the ultra-fast refreshing ICM20608 will show signs earlier. After all, many people fly with bent bells and damaged props- they aren't going to notice a difference in frames!

[@princesfpvdrones]

How fascinating!

[@gievepix12]

Hello KababFPV i would like to ask you about your graphs. All graphs that you provided into the video are having x axis limited from 0-1000hz. Does that mean that higher frequencies are not that important for drone flight? Because i'm imagining this system to fly on frequencies much higher than 1000hz. #question

[@Kabab]

0-100hz is typical control frequency area. I'd you have no frequency below 200hz, you're pretty much fully safe. All the frequencies above 200 are much easier to filter out. Anything above 400hz is a total non-issue for us unless the amplitude is super high.

[@MrGone2k]

On the topic of resonance, this gun video overlaps the topic with an irregular barrel design - kzbin.info/www/bejne/Z3OXpZShoJqjg5Y maybe some irregular features in the arms would deaden those troublesome frequencies.

[@paulbredt5607]

It is interesting you are doing all this analysis. Seems like you should have 3 screws on each motor since 3 points define a plane. But overall, I would suggest making a few duplicate quads and see if the results are repeatable.

[@Kabab]

The motor doesn't twist on the arm. The entire arm twists a bit under the motor weight

[@csg2000]

So I know the industry standard is flat sheet components that key together, but have you ever thought about testing a 3d unibody design? The layup would be a lot more work than machining flat stock, but it could be super interesting. I started designing a unibody carbon whoop as a fun composites experiment.

[@Kabab]

There's been a bunch of 3D unibody/molded carbon designs. They're several times more expensive and offer no tangible benefits unfortunately. That's why we don't keep seeing them.

[@David_Nielsen]

Would a kevlar or kevlar-carbon hybrid drone frame possibly have less vibration because of its higher material damping factor? Its not as stiff as carbon so it might just lower the frequency of vibrations and be worse, but it might be interesting to explore? I'm kind of surprised I don't see much kevlar used in this hobby, as it might also provide better impact resistance for those that don't care as much about vibrations.

[@Kabab]

I've tried using kevlar layers in carbon. It doesn't help at all. Makes everything worse but not much worse.

[@hetlarfje]

They Shoud build a apex frame with holes in the front bottom plate and the arms to reach stack screws and stack mounting like stingy v2 frame in rear bottom plate

[@hetlarfje]

Be shure the stack Bolts are sunken deep enough. Otherwise the arms make contact with the stack screws only and all arm vibrations will go to the stack

[@olafschermann1592]

Aren‘t single, higher spikes better than lower amplitude at a wider spectrum because you can use notch filters against that single spike frequency?

[@Kabab]

Overall, if you have a single moderate spike, it is better but with so many different components out there, it's impossible to tell if that spike will remain from build to build. It's not unlikely for that peak to be sky high and no filter will really be able to deal with it. In the case of the apex arm mount, you get overall more noise but less amplitude. Maybe only 10% but 10% off a huge spike is still a huge number. Still a tossup, I'm testing a couple other small features as well to figure out exactly what the mount mechanism is doing...

[@testaccount4191]

really stupid question but, can you not just put rubber on all of the contact points to help with the dampening including soft mounting the motors?

[@Kabab]

Soft mounting motors is a PITA. The screws don't stay put. Otherwise, yeah you could and it has the same effect as we've recently discovered with electrical tape on top of the arms.

[@kydawg1684]

FEA is so sweet.

[@shotbyarian]

have you tried chris's AOS 5.5 and what do you think about his frame designs?

[@Kabab]

I think his frame designs are nice however their inefficient from a manufacturing and weight standpoint if looking at the bigger picture and all the areas and functions of a frame. They do one thing well and sort of mostly give up on the rest. Nonetheless they will provide a consistent build because the resonance profile is under control. That does not mean you won't have any problems however. You're still putting four spinny things on the frame with props and a heap of other stuff that can cause problems...

[@shotbyarian]

@@Kabab thank you for your comment! will have to watch more videos and then decide .. your frame also looks nice

[@Imakilln]

i still dont understand why we don't just softmount the motors... Years ago I compared some abused but soft mounted motors to some new identical motors - soft mounting clearly won. Never tried it again since but damn that quad flew nice before i totalled it!

[@Kabab]

Soft mounting the motors tends to not do much anymore with the recent filters. Your motor screws end up coming loose and your motors fall off sometimes in flight too. It's also a pain in the ass to install. Sometimes the soft mounting mechanism would make it worse too because what we need is specifically friction, not bounce.

[@jackfpv8904]

@@Kabab I'm using 3d printed thin pad under my motors all the time. I have never check vibration from FC, but quad feels better.

[@Imakilln]

@@Kabab Yeah my experiment was back in the days before betaflight added the filters tab 😅. That said, the less noise the gyro sees the less filtering you need in the first place. To solve the in flight motor ejection issue I simply used some high strength loctite, never had a single bolt come out after the first accident! If I was to try it again I think some custom bolts with a short thread & long shank combined with a larger longer version of our fc gummies would be a good experiment.

[@rawpicsfpv3516]

does this have the same amount of space then the fouride? Does this fitt the vista and 20x20?

[@Kabab]

Yes it has holes for 20x20 mounting and fits a vista in the back for sure. It's designed to be built with a vista. It doesn't really work the same without it.

[@rawpicsfpv3516]

@@Kabab nice, how long till available? And will this also be the go to frame for 350-400g quads or is it too fragile for that?

[@Kabab]

@@rawpicsfpv3516 it'll hopefully be here in just about 2wks. Around 450g is the max expected weight.

[@ehunter2010ed]

Sorry I must have missed this part but apex meaning mr steele apex correct?? Or is there another apex ur referring to??

[@lawrencemartin8205]

@Eric Dunlap yes correct. Impulse rc apex frame.

[@ChadKovac]

I wonder if you could get someone with a motion amplification camera to video one of these things?

[@Kabab]

Unfortunately I don't have a high-speed camera in the tens of thousands of FPS

[@ChadKovac]

@@Kabab of course not.. I'm thinking maybe you could find a fellow content producer who does?

[@raast9835]

@@ChadKovac collaboration with slow mo guys would be the perfect one!

[@raast9835]

Good reflexions....have you compared same setup with continuous single bottom plate against bolted arms? Obviously the target is be able to replace one singoe arm but from stiffness stanpoint for same mass distribition, having a cut on the structure makes more flexible

[@Kabab]

Unibody frames are not good for anything unless it's 3" or under. There is literally only one benefit to unibody frames and that's the time it takes to actually put together the frame....because you don't need to bolt the arms on. Otherwise, every single other metric is improved by going with separate arms and body. Weight, durability, vibration management, repairability, aerodynamics....it's been about six years and I have been unable to come up with a metric that a unibody is better at other than the time it takes to put together just the frame which is pretty insignificant imo.

[@raast9835]

@@Kabab thanks for your response. I apreciate! I respect your experience as is the best teacher.

[@raast9835]

@@Kabab Just one comment about fft plots comparison, the metric to check is the area under the equivalent psd curve equal to mean root mean square. That are the rms values to compare with. In this case indeed the magnitude of vertical peaks is a way more important compared to the spread across freqs. In this case you are introducing the same amount of energy(same engines,boases,setup) and what we see is that the way the junction is modify, it affects the way it is responds. No energy absortion relevant apart for the coulomb damping modification...so the energy is invested across more freqs with less global amplitud. The rms could be a better metric to compare quantitatively as you will see how much is reduced between designs. But fully agree anyway in your rational! Good job!

[@joshtabor3288]

Seems like you could probably confirm the vibration problems if you used a high speed camera. Also, if you have a certain frequency that you're getting vibrations at, you could probably use a band stop software filter to eliminate that frequency.

[@Kabab]

These vibrations would require a really expensive high speed camera that I don't have access to. We're talking about 150-300hz. You'd need at least 75,000fps to be even begin to be able to see these oscillations. The black box does a good job of recording them however so it's not as hard to deal with when you know what it is.

[@joshtabor3288]

@@Kabab I think you could actually start seeing it around 600fps - 1000fps (you need to sample a minimum of 2x the frequency you want to catch, and more is better) Even that's probably still hard to get ahold of though. I'm not sure if the vibration would be high enough amplitude to be noticeable even if you could get the fps. I'm also curious about the bandpass filtering. Obviously, you can't get rid of the resonance with software filtering, but you could probably remove the effect it has on your PIDs, which would help.

[@marc_frank]

have you tried mounting the fc to the topplate?

[@Kabab]

Overall the whole frame is vibrating. You do get different vibrations from the top plate but they're still there. Also, depending on the top plate design, it could be worse. This arm mount mechanism is specifically damping so it going to help no matter where the FC is mounted.

[@marc_frank]

@@Kabab if we could only get more powerful processors to do a FFT filter basically converting gyro data to the frequency domain, chopping off the unwanted frequencies, and then converting the frequency data of the selected portion back to raw movement with an inverse FFT all the non movement data would be just gone the issue is not enough resolution in the product the FFT's currently used to position the dynamic filters don't use all the data

[@timl4081]

Looks like my builds with 5-10 different screw types, brands etc.

[@natural_nc7230]

How did Charpu ever get his quad off the ground? Just kidding I'm just getting back in to this hobby after a year. If you recommend something I'm probably going to buy it. I really enjoy your passion for FPV. Thanks

[@thomasfischer1580]

9:45 lol go climb up get it😂 i just recently climbed a 50ish ft tree because my 5in was stuck at the very top.

[@user-lu2cy9xv2z]

It seems to me that improvements with different types of mounting to the bottom plate can't get higher 1% because all the troubles comes from audible fork which all the arms with flat form appear. you can check all the resonance frequencies with just drum tuner app for Android, there is no need to fly

[@Kabab]

yes you can actually use many phone apps to pick up the resonance. The simulation just tells you all the resonant frequencies. The bottom plate doesn't cause vibrations but it can dampen them. That's what I'm testing here.

[@user-lu2cy9xv2z]

@@Kabab I do believe that the only thing which will decrease the amplitude in blackbox is softmounting motors first of all as the reason of the vibrations and than softmounting fc. The second way to make arms as hard as you can or add some "thickness" in the design. If you look into bbl of any cinewhoop with iflight bumblebee like construction, you'll see that all the vibrations are above 250-300hz. The most interesting way to kill resonance is avoiding it by ESC... For example you have resonance on 100Hz, this frequency appear on 10000rpm, so the flight controller understand it and trying to avoid this rpms and for example 2 or 3 even harmonics. And so 2500, 5000, 10000, 20000 and 40000 rpms will be avoided

[@rcvg69420]

Maybe we take a page from the dark ages... Some electrical tape between the arms and center plates?

[@Kabab]

that could possibly work but I think it might induce more bounce than friction. Bounce will amplify things

[@Bruno-cb5gk]

Hmm, I wonder what would happen if you made "I beam" arms, as in use two plates per arm and offset them with a vertical plate. In theory that should increase stiffness quite a bit, though maybe as you seem to be alluding in this video we should focus less on stiffness and more on damping.

[@Kabab]

Yeah stiff is needed but it seems some potential friction needs to be left in the system as well.

[@samuelyoung2671]

Imagine that would be pretty bulky tho, as well as add some drag. idk?

[@Bruno-cb5gk]

@@samuelyoung2671 Yeah, it'd probably work best with a Remix style frame since that already has a much lower bottom plate. I think the drag penalty wouldn't be too significant, especially if you won't be racing. Might try to run some simultions of my own.

[@samuelyoung2671]

@@Bruno-cb5gk Would be pretty cool to break the flat carbon plate meta.

[@samuelyoung2671]

That being said considering how well quads fly now, Im skeptical the return on investment for this sort of innovation is there. unfortunate really

[@samtny1]

I thought you couldn't really compare overall amplitude of one chart to another, due to the fact the Y axis / scale is not a constant but instead more of a "fitting" algorithm based on the flight you are looking at?

[@Kabab]

They're set the same and the flights were very similar so they should be very comparable. I have the waterfall charts but chose not to show them here because they're more difficult to read. Happy to send them to you if you like. They paint the same picture.

[@scottmilano2940]

Do you have, or can you point me to a TF2 assembly video? Edit: A google image search helped a lot, I think I have it.

[@alxgag3]

Someone should make an online calculator where you could upload a stl of a frame and get the simulation results back. I see Chris uses Ansys, but I believe you could do this in Fusion 360 as well (sadly not in "personal version"), prob some open source tools out there also

[@Kabab]

That would be quite convenient

[@brezovprut4431]

What do you think about 3D printed frames?

[@Kabab]

They're alright in small sizes. Harder to do a larger print that works. Even in parts. The issue you'll ultimately run into is thay in order to acheive the same overall strength as carbon fiber, you'll end up using more weight in printed materials than carbon and you'll have an aerodynamic disaster. Not like flat carbon is so aerodynamic but it's at least not giant.

[@brezovprut4431]

@@Kabab Alright, so for larger frames even with variable ribs and infills we aren't able to achieve carbonfiber-like strength?

[@alaingosselin8764]

Very c👀l thanks for sharing 😃✌️

[@jeroenjegerings]

Is the 7" frame now being revised because of the apex theory?

[@Kabab]

No we actually have it. I just haven't had a chance to do QC on the batch and list it on the site

[@jeroenjegerings]

@@Kabab Good to know! Instant buy for me, together with the 7" motors!

[@churwck]

When Can I buy it?

[@Kabab]

Hopefully just a couple weeks

[@m34nb34n]

😂😂😂🌟🛩🛩💯noone else giggled like a fool cause a loud Plane was Interrupting a guy outside talking about Drones? Aaaahahhahaah ❤❤🤘🤘💪💪

[@maceliusfpv370]

Looks up from building 5" Fouride LR after making some repairs to his Glide Hammer: Finally! Can We please have thicker 5" arms for it now please :D

[@bob.w3984]

Try 6 blade props and see if that has a relationship to the vibration..

[@Kabab]

It does. Should be less vibrations because of the lower overall rpm

[@maryjane136]

Interesting. But I think we do not have this discussion if the MEMS Sensor IC is located on a daughter PCB like developers test new components. Are there FC's on the market with external SPI? If not, this would open a door to integrate other MEMS components like from Analog devices for example as well. And I would be able to patch BF to use my choice of motion sensor. Having a daughter PCB, mounting ESC and FC could be done traditional without dampers. And for the "gyro board" we can think about to install it in a new way, i.e. like a studio microphone. Instead of AOI I think a better system would be a module based one. Not sure about the last sentence, probably you guys know some cons and disagree.

[@Kabab]

There are actually many boards that have a separate gyro board. The FC sometimes comes with different gyros to plug in too. I'm not sure why this fell out of style

[@rawpicsfpv3516]

Any chance to get this in 4"? Maybe for 2004 4s 859mah...

[@Kabab]

maybe 4" but as with all 4" quads, even my own, they're hard to justify when 5" is just a little bigger and does everything better. It will however come in 3" because that is smaller.

[@craigcunningham8058]

Hope you give more attention to smaller frames capable of carrying Vista HD Nebula Pro. A lot of drone pilots like myself do not have great wide open areas to fly so 5" quads become to limited. give us a 3-4 incher, a trimmed down Fouride, for personal flying experience. Loose all GoPro crap. We have HD !! I enjoy flying not looking at movies after the event. Make some quads optimized around the Vista/Nebula Pro. Most hobbyists will not be flying big Go-Pro boxes around in the coming future. Would also be cool to see noise traces you show in video with a control solution such as the 4mm stiff triangulated single piece bottom plate of the Babyhawk 2 HD. Just to let us see you are not polishing a turd.

[@Kabab]

there's a 3" version of this coming too

[@aakashjana6225]

This frame looks similar to ummagawd's 2fiddy and he had frame resonance problems.

[@Kabab]

Anything skinny and light with big motors will have resonance issues. 😅

[@MoppelMat]

The side bars did not work! Reading this while dremeling and cutting my side bars for my project 399 super g+… 🤦‍♂️

[@Kabab]

Doesn't work on every frame.

[@MoppelMat]

@@Kabab i finished it and it worked great! 🎉

[@doncunningham3124]

I bought one of those Prometheus titanium frames. My common sense must have been taking a nature walk at the time. I can't believe I wasted $170. on a wind chime. It's too light to be a paperweight. Damn!

[@Kabab]

😐 it would make a nice trophy or something

[@leflamantrosefpv1725]

unibody?? (i never broke one and for the analisis is better no??)

[@Kabab]

The only thing better about a unibody is thay it's a little easier to build the first time because you don't need to put it together. Otherwise, there's nothing better about unibody at all.

[@leflamantrosefpv1725]

​@@Kabab ok that was a suggestion, you are doing a great job, thanks for your videos! :)

[@smartfpv3992]

Ok fine, but I still have to turn on my filters. Doesn't matter if it's more or less vibrations, as long as there are vibrations I have to turn on the filters. So I completly agree with you and your theory that shifting the frequencys higher is more useful. Or did you change your mind now?

[@Kabab]

No this is the right way to go except that you need to introduce friction dampening and increase frequency at the same time. That's where it gets difficult

[@smartfpv3992]

@@Kabab yeah especially when you don't have access to this kind of simulation. How can you know if you are shifting the frequencys up in your design?

[@justhackitfpv]

I lost it because I'm an idiot 🤣 I say This to myself when I crash

[@ariafpv]

Yesss, finally I hope more people will try and prove hypotheses by actual "science" instead of just talking about feel

[@donnygfpv7548]

I now wonder if the TP3 arm cross brace is better or worse, I haven’t personally been able to test this since I don’t have a cross brace or black box on my current build