Amazing work! Tying the waterfall plots to the FEA simulation is a real science upgrade for the hobby! Love it!

@ChrisRosser

3 жыл бұрын

Thank you! Hopefully it's going to bring about improved frame designs in the future.

Next level info! Nice to see someone showing diagrams of concepts that I’ve wondered about! Keep up the good work! Subbed

The fact that I know nothing about physics and I still understood everything just shows how perfect this video is ! This channel will quickly become very popular. I am currently tuning a 7 inch, and you just put maths on my observations. Thanks a lot !

@ChrisRosser

3 жыл бұрын

I'm glad you enjoyed the video! Good luck tuning your 7in. I've posted another video where I add some simple carbon parts to the frame to see how they can improve certain vibration issues which might be of interest to you. Are there other topics that you would like me to do a video about?

@LenordeFPV

3 жыл бұрын

@@ChrisRosser I had just finished designing and ordering thicker arms when I saw that last video. I guess it will be the next step! I would love to see how other factors play a role in frame resonance: quadcopter configuration (true-x, deadcat...), arms fixation on the frame, center of gravity... And does the arm shape/profile also make a difference with overall equal cross-section?

@ChrisRosser

3 жыл бұрын

@@LenordeFPV Alright! I think you'll really love my next video. I'm going to be looking at damping within frames and how certain design decisions can reduce frame vibrations. 😁

@LenordeFPV

3 жыл бұрын

@@ChrisRosser Nice, I can't wait to see it 😀

Amazing analysis 🧐 thanks

@ChrisRosser

3 жыл бұрын

My pleasure! Glad you enjoyed it.

Hi Chris, a hypothesis that could explain the pitch axis response on the 160 Hz, is when the stack is bouncing up and down, the capacitor (that is ahead or behind the stack) causes a lever on the ESC and it causes the entire stack to swing on the pitch axis. It would be very interesting if you could model the stack with its supports and the masses that are attached to it, since the modal response of the stack could be different from the way the drone vibrates. Sorry for my english, but it's no my native language 😊.

Bedankt

Great video, thank you! I happen to use both frames... Would you share your tunes for these? BTW my 5" has 2204s and 7" 2508s/1200kv. I want to use Gemfan 7042 bi-blades.. Or should I start from scratch with the help of your videos?

this is just pure gold! may you elaborate on what someone can do to get rid of jello? or may you look into what jello on a basic level really is? I guess you would need to get the framerate of the camera and shutter speed into consideration. I fear that the only thing you could really do to reduce jello is to fix the hardware issues, or do you think you can "tune jello" out? thanks in advance!

Oustanding work! I have a question though: do you think it makes a difference whether I mount the arms inboard (above the bottom plate, as the frame is designed to be), or below the bottom plate? Because on my build I'm having issues with space, and mounting the arms below the plate would free up a lot of room inside the quad, but I don't know if it's a good idea. The little carbon reinforcing plates would fit, I tried, so I could use them, just upside down and on the bottom. Thanks

Great video! Having the 7 inch frame got some questions here 1. What about flying footage? Do you get jello (without ND) ? 2. Do the 5,6,7 inch arms fit on the same frame?

very useful video .... A question, I have a 7 Polloco frame that I modified by designing new more resistant arms. In flight the vehicle has a low frequency vibration that I can not eliminate with the pid. He can't sit still. Is it possible to do something or can I throw away the frame?

Amazing stuff! No one makes similar content on the KZread AFAIK. Interesting to know your opinion whether X shape arms connecting in the middle are better vibration wise then a dead cat config.

@ChrisRosser

3 жыл бұрын

The length of the arms is a big factor. I would say true X often has longer arms than dead cat which can be a concern. Unsure as to whether the interlocking arms will increase stiffness for the very small movements we are concerned with. My gut feel would say dead cat might be better all else being equal.

How do you calculate the Young's Modulus of your carbon? What are you using to measure the stiffness? Or, are you looking up the value from a data sheet? I ask because I have a lot of different frames and the arm stiffness varies quite a bit even when the cross-sectional dimensions are very similar. It seems to be affected by both the direction of the carbon weave, and the type of carbon fiber (e.g. T300 vs T700), but I don't know how much those things would affect your calculations.

@ChrisRosser

3 жыл бұрын

I'm just using a typical value provided here: www.performance-composites.com/carbonfibre/mechanicalproperties_2.asp. As a result the simulation frequencies never perfectly match the blackbox but the mode shapes are still representative and occur at frequencies close enough to reality that matching them up is super easy.

both have low freq spike at around 40-60hz.. and this usually can be removed by removing all gyro lpf and rising dterm lpf.. is this caused by those lpf (too. much latency?)

You could try taping the motor wires down on the arms over some moon gel, that stuff is excellent for vibration damping. It looks like they made the arms a bit skinny for the 7" though tbh. Excellent video 😁 by the way.

@ChrisRosser

3 жыл бұрын

Yes, I think you are right. Its not really a design optimized for 7inch at all. But I can learn more from noisy frames than quiet ones sometimes! I will definitely get hold of some moon gel. Thanks for the suggestion. 😁

@jonc67uk

3 жыл бұрын

@@ChrisRosser There was a huge thread on rcgroups back in the day on arm materials & vibration damping for the long range photo squad. You may be pleasantly surprised if you stick some balsa plates vertically under the arms to stiffen them & improve airflow over the arms from the props in terms of turbulence. It turned out that wood was the best arm material for vibration damping, spruce for preference. Aluminium tube arms were the worst, with carbon tube arms transmitting HF vibration quite badly just for irony with the current miniquad frame material of choice lol.

I have the 5" (original) frame and 6" (clone) frame. The former doing super on stock tune and it has survived for more than two years! The ladder is a vibration hell and I have been tuning and tweaking and never really gotten it better than barely flyable. I think the clone frame worse tolerances worsen further the vibrations. In the end I gave up bought another 6" frame.

I will experiment with adhesive sound dampening mat strips under the arms.

So, the other part that isn't intuitive is the frequency at which 7" props themselves resonate. In most cases, it is between 127Hz and 148Hz that the props will resonate at, and that can also drive a fair bit of the observed noise

@ChrisRosser

3 жыл бұрын

Great observation. On my 7inch the props vibrate at 130Hz and on the 5inch it's 270Hz. Have you noticed prop vibrations in your blackbox?

Awesome I love your frame analysis. I’m going to give black box logging a try. You have helped me understand a little more how to use it. I was wondering did you put a tune on the GepRC Mark 4 HD ? Or do you have any suggestions for that frame ? I have the vista nebula pro 20x20 Tmotor F7 FC ESC 55A Ultra esc. I converted an analog Mark 4 with crossfire and DJI V2 goggles. I just need some help on a Bettaflight 4.2 or 4.2.8. Tune.

@ChrisRosser

3 жыл бұрын

You can try my standard filter recommendation. Its on the community page. 👍

@fpvraver

3 жыл бұрын

@@ChrisRosser community page ? Is it Bettaflight 4.2 ?

@ChrisRosser

3 жыл бұрын

@@fpvraver kzread.infocommunity Yes its 4.2

@fpvraver

3 жыл бұрын

@@ChrisRosser nice thank you 🙏🏻 your the best !!!!!! Supper fast 💨!!!! I clicked on the link you sent me it takes me to your community link but I can’tfind the tune for the GepRC Mark4? I’m not that good with all this social media. Sorry lol

@ChrisRosser

3 жыл бұрын

@@fpvraver I just have a standard tune for clean builds. kzread.info/dron/45_6KVAAZG_iOgzksGswMw.htmlcommunity?lb=UgxIdA2nSEQHTP-041R4AaABCQ These settings should be a good starting point for most builds. If you have noise issues try lowering D term low pass but do not go lower than 70Hz. If that still isn't working you may have mechanical issues. No gyro low pass filters. No static notch filters. No yaw low pass. No D notches. Dynamic notch filter ON 100-400Hz Q=250 width percent 0 RPM filters ON (min 100Hz) D term: 1 Biquad low pass at 100Hz

Hi. Great video. You should test a Rebel Ruxus. It has 10mm thick arms in their 5", 6" and 7" frames. My Ruxus 7" flies great with almost no tuning and I believe it is because of the thicker arms and compact design.

@AlvaSpeed

3 жыл бұрын

I believe the bigger the quad the thicker the arms and quad in general should be.

@ChrisRosser

3 жыл бұрын

Wow, that frame almost looks like it has vertical arms they are so thick! How much battery can you carry with your 7 inch?

@AlvaSpeed

3 жыл бұрын

@@ChrisRosser I fly for fun and performance. Not interested in LR at the moment. I usually fly with 6S 1300mah and have used 2 of those batts in parallel (2600mah). It can carry bigger batteries if needed. Right now on Xing 2806.5 1800kv and this week will change to BH 2810 1500kv (more power...I hope) Rebel makes some motor guards that cover the whole tip of the arm, so your motor mounts on a thin layer of 3D printed material that somehow may help to reduce vibrations. But overall I believe thicker arms are needed on anything above 5". And bigger motors than what people are now using for correct prop handling.

@ChrisRosser

3 жыл бұрын

@@AlvaSpeed Hi Alva, We are definitely on the same wavelength. 2806.5 is definitely the minimum for 7 inch and 2810 sounds really nice! A thin layer of TPU could help with vibrations by increasing the damping. If you could produce a gyro_scaled blackbox before and after that would be amazing! I can then look into modelling it and seeing what we can find out. What I'm seeing definitely suggests even thicker arms for 7inch. It should scale linearly with length I think so 5mm for 5inch and 7mm for 7inch at a minimum.

@AlvaSpeed

3 жыл бұрын

@@ChrisRosser I have never used black box...hahaha. I will try to learn....

Man....I can already see how deep the rabbit hole goes. It seems to me the main stack screws and the material they are made of also have an impact on the overall noise getting to the flight controller...? Is it possible you would get different results with different materials (ie. titanium, aluminum, steel)? Obviously we already tackle vibrations with the gummies on the stack screws, but would it be possible to "float" the stack on a separate platform with dampening underneath it? The individual stack platform could possibly be made from the dampening material? I have a Mark 4 frame myself! lol

@ChrisRosser

3 жыл бұрын

Vibration isolation of the FC is an interesting question. Mechanical damping and isolation adds delay to the PID loop just like software filtering so damping/isolation is not a perfect solution. Ideally you want a reasonably hard mounting so you get the best possible measurement of the movement of the quad. But you obviously need to reduce the vibrations to avoid oscillation and hot motors.

@angularfpv6572

3 жыл бұрын

@@ChrisRosser makes sense. Thanks! Thanks for indulging my "thinking out loud"ness lol

Hmmm, I haven't tried yet, but is it possible to see resonance with a thermal camera? Or does the vibration not cause enough heat to detect?

@ChrisRosser

Жыл бұрын

I doubt there would be much vibrational heating unfortunately. The amount of energy is very small.

Hello! I just bought geprc mark 4 5. As you was showing 5 and 7 in beginning of video it looked like 7 inch version is just 5 but with longer arms, is it true? can i fit 7 arms on 5 body? thanks

@ChrisRosser

2 жыл бұрын

You can but I wouldn't recommend it! The arms are too flexible.

My 5" is the Geprc mk4. I finished my Geprc mk4 7" today but before I did anything, I used Loctite Hysol to glue an extra 3.5mm (aprox) to the underside of the 7" arms. The arms are about 8.5mm because 5mm didn't seem to make sense for 7". I haven't tried tuning it yet but damn am I impressed at how it flies on stock Betaflight. I had a theory. The carbon I glued to the arms is a 4x4 twill weave which is stiffer than 2x2. Also the Hysol is not as hard as the epoxy the carbon is impregnated with, so possibly the Hysol in the lamination would help to dampen some of the vibrations and of course the thicker arms should help keep them from turning to rubber from resonance. Making them thicker and stiffer might raise the frequency from lower to higher. What do you think?

@ChrisRosser

2 жыл бұрын

That's a really good idea! making the arms thicker will make a huge improvement. You could even try some damping grease to further reduce vibrations.

@CarbonGlassMan

2 жыл бұрын

@@ChrisRosser I noticed today that the body is the weak point. I can twist the main body pretty easily now. I'm going to do something to stiffen it up too.

It would be interesting to see how soft mounting the motors affects these outcomes.

@ChrisRosser

3 жыл бұрын

Yes quite a few people have enquired about that. Do you soft mount your motors? If so what effect did it have?

@Perforator2000

3 жыл бұрын

@@ChrisRosser I have soft mounted the motors on a couple of different six inch quads when they were built, but I haven't run tests to see the effect. I would just need to capture black box logs with and without soft mounting, I suppose. The theory seems sound that it reduces vibrations reaching the gyro and is essentially latency-free filtering. I know Joshua Bardwell did some tests a few years ago and showed that it made a difference.

@ChrisRosser

3 жыл бұрын

@@Perforator2000 If you have the time and can collect gyro_scaled logs with and without I'll do some simulations and make a video looking into it in detail. Does that work for you? If so you can reach me on email kzread.infoabout or Discord Far_Humor_420#4205

@Perforator2000

3 жыл бұрын

@@ChrisRosser I can possibly do that. I need to order a few parts before I'll be able to do it, but I'll be in touch. I'll be using a 6" Hyperlow RS+ frame with 6mm arms.

@ChrisRosser

3 жыл бұрын

@@Perforator2000 I look forward to hearing from you if and when you reach out :D

What are your filter recommendations for Mark4 HD5 and Mark4 HD7?

@ChrisRosser

3 жыл бұрын

I would start with my standard recommendation kzread.infocommunity. If its not enough filtering you'll need to lower the cut-off frequencies gradually until it is OK.

@BuzzEye_

3 жыл бұрын

@@ChrisRosser thank you, I’ll try

Your videos are awesome regarding the information. However the background music gets annoying after a short while. You propably felt as if your video was lacking without the music. It would not have been! Please keep doing what you are doing.

@ChrisRosser

3 жыл бұрын

Thanks J. I've turned the music down a bit on the more recent videos. If it's still annoying I can try some videos without!

Hey Chris. My name is Zach. Would you be interested in setting up a video call regarding a lot of these concepts, and especially about some of the mechanics around yaw. I will join your patreon and go from there.

@ChrisRosser

3 жыл бұрын

Hi Zach, you're spot on! I'm happy to set up a video call with you and would really appreciate you checking out my Patreon. Please get in touch on email you can find it here kzread.infoabout or on Discord Far_Humor_420#4205