Resonance analysis of the Armattan Marmotte, Shaken not stirred!

Hi everyone,
Lots of people asked me to do a video looking into an Armattan frame. So here it is! Hopefully this video will explain what's up with the Marmotte and what lessons we can learn from the vibration problems many people report with this frame.
Thanks,
Chris
Comments and feedback always welcome!
Song: Markvard - One with the ocean (Vlog No Copyright Music)
Music provided by Vlog No Copyright Music.
Video Link: • Markvard - One with th...

[@christianmmoser]

thank you very much for this interesting analysis. i propose that you bring it further. what really would be very interesting for all the marmotte owners is how to set up betaflight to make the best out of this quad.

[@naomarik]

What you're doing is extremely great. I hope you do this to all popular frames. Would love to see what the Apex is doing.

[@ChrisRosser]

I hear the apex has good vibration performance. I would be happy to model it if someone can provide a gyro-scaled blackbox log and CAD of the frame. Alternatively if you are interested you could sponsor me an Apex to measure up, model and test fly and I'd make a video detailing the results of the analysis.

[@ChrisRosser]

Apex video is up kzbin.info/www/bejne/q2bFhZRuhd5sgas

[@flaparoundfpv8632]

Remember when Armattan tried to sell us space carbon, but everyone told them that aerospace pre-preg isn't meant to be durable, and that the resin system would be too hard to withstand a crash? And then remember when Chris doubled down on the marketing claims, but everyone kept breaking their frames and Chris lied about how many were breaking and went back and deleted everyone's comments on FB and KZbin from months before, saying that the frames would break too easy? and then remember when Armattan had to admit the new carbon broke too easy and then they stopped selling it? Pepperidge Farm remembers.

[@Kabab]

This is an intesting way to simulate things. If I can find you on FB, I have a design concept that may be intesting for you to think about. Not any specific design, just a way to make things that may be a lot worse or a lot better depending on the design. Interested to hear your take on it.

[@ChrisRosser]

Can't believe I missed this comment. I've sent you an email to connect with you about frame design. Looking forward to speaking with you soon 😁

[@maceliusfpv370]

Did the collective thoughts of a few hundred FPV geeks just manifest the ultimate collab? Perhaps we could learn something by evaluating the Glide.. but I know you're focused on the prototype 5.

[@RCRitterFPV]

I was just doing a search to see if he tested the prototype5 ... then I find this thing of beauty. can't wait to see what becomes of this.

[@mouseFPV]

@@RCRitterFPV really want to see a pt5 frame have this done. my logs look really good sans some weird noise at 80hz but I would love an in depth

[@slowentropy4531]

Years of flying and intuition combined with a highly skilled and motivated engineer, this will be very interesting to follow. Best of luck and enjoy the journey.

[@MPQuads237]

Nice, this confirms my experiences. The issue is even worse on analog Marmotte where there are no middle standoffs.

[@doctorjmobilefilming8194]

Nice video! I am experiencing exactly the same with my Marmotte, and same black box data. Keep making more of this! Blunty & jb sent me here too 😁

[@CapnBry]

Thanks for making these and showing how they correlate to the blackbox logs. I have an 3" Armattan Tadpole with a crazy-high off-the-charts noise at 340Hz on the roll axis only and now I'm understanding the spectrum plot that shows how the vibration stays at that same frequency at different throttles and I am guessing it is exactly this. Now I can stop spending hours debugging it!

[@ChrisRosser]

Glad that I was able to help you out 👍

[@maggoffm7586]

Nice analysis 👍. Also the one with the apex frame. I would love to see a noise analysis of a badger frame and a comparison of marmotte,badger and apex. Great work, dude 👏

[@davidlondon3111]

Thanks Chris! New subscriber. I learned something

[@ChrisRosser]

Thanks for your support!

[@travis6768]

Fascinating...thanks for doing ths!

[@yawnastyfpv6942]

Keeping the focus on Armattan frames, how would you say the Gecko 3 stacks up? Would you say the Vibration would be amplified given the unibody frames design and short proximity to the source?

[@mikebergman1817]

Yes! Found you through Jb and ItsBlunty, and I'm so happy you're doing this!! Any way you would consider doing the 399 super G??

[@gerthyfpv7577]

Thanks for doing these tests. Love seeing what you come up with but I am curious how large of a sample you have for this frame as well as what your build and PID tune is along with the betaflight version. I can only imagine that a good tune on a bad frame will show less noise than a bad tune on a good frame. I have multiple builds on the marmotte, 6s with Xing 2208 1800kv BF4.2 with bi-dshot and rpm filtering(my youtube pic), and I have a lot of black box recordings where I am not seeing the same thing you are. My logs are actually extremely clean. Would you be interested in checking out the flight logs of my Marmotte to compare to yours? Let me know and I can put them on google drive.

[@ChrisRosser]

I collected a few logs from the community to check my results were representative. I haven't found the tune to affect the gyro_scaled plot too much provided the gains are kept relatively modest. If you have debug mode = gyro_scaled logs that look very clean then I'm happy to take a look. You can email me them at the address for the channel 😁

[@WKfpv]

Loving this series, how about the split top plate designs like the lumenier qav-s? And I would love to see some analysis of the top of the line frames like the impulserc apex

[@ChrisRosser]

A split top plate design would be very interesting. I don't own a QAV-S but if you would be interested to sponsor a video I can put together a test quad, fly it and publish the analysis!

[@fractalengineering]

Mate may I ask you a couple things; 1-In your analysis did you actually input the proper material properties (I mean the full set of orthotropic specifications) or is it just some anisotropic simplified model? I highly doubt that you have access to the space CF data; but you certainly know that any such analysis, whether it be static or dynamic is pretty much moot without the model following the real world as close as possible, right? 2-You mention the lipo leveraging the top plate into such vibration. But did you consider that the real world lipo attachment may have a certain degree of dampening due to the lipo strap flexibility? Or did you just clamp a brick on the top plate making it behave as one solid part?

[@ChrisRosser]

Hi Titansteam, thanks for your comment! I use an anisotropic approximation of the carbon fibre. It is not a perfect replication of real carbon fibre, but the difference in resonant frequencies between the simulation and the black box log is only around 10%. The goal of the model isn't necessarily to perfectly replicate reality but to approximate it closely enough to allow the mode shapes to be visualised. The lipo attachment is modelled by a battery pad with the same softness as a typical gel battery pad such as Ummagrip.

[@Perforator2000]

Great work. I'd be really interested in seeing some larger frames for six or seven inch props. Maybe the 6" Glide or the 6" Hyperlow RS+.

[@ChrisRosser]

I would love to help out. I have a 7in GepRC Mark 4 HD7 would you be interested in that? Alternatively if you would be interested in sponsoring a video I'll simulate any frame you nominate!

[@TERBANGFPV]

Hi Chris.. before using blackbox to find frame resonance, i always checking the resonance by tapping the frame with my finger nail. When it vibes (like having echos) i can tell that the frame is having resonance issue. And this afternoon i picked up my marmotte frame from my drawer 🤭 and adding electrical tape at every spots where the carbon frame meet the aluminum/titanium part. And after that i tapping the frame with my finger nail and the vibes are way shorter (looks cleaner from resonance).. this makes me want to re build it again and see the noise profile with bb log. Maybe ill do this in next few days and hopefully it doesnt waste my time rebuilding it hahaha. Maybe the electical tape can help a bit just like prototype 5 frame from kabab. 😊 Cheers

[@ChrisRosser]

Yes electrical tape can definitely help in some situations, but there have also been cases where it makes things worse. Mechanical damping grease works better and more consistently than electrical tape and lots of people are using it to great effect on their builds. I have a video on it if you are interested.

[@TERBANGFPV]

@@ChrisRosser oh it can make it worse too? Yep i watched the video.. but i cant get the grease here in my local

[@MUNKYFPV]

Really interesting thanks 👍😎 I notice in your resonance analysis you are using the HD version of the frame which has a longer body and top plate but has an extra stand off compared to the standard frame. I actually had the standard Marmotte and noticed a big spike around 250hz in blackbox and after breaking the bottom plate converted it to a Badger under warantee. The seperate arms seemed to reduce the noise in blackbox, not sure if they are thicker and narrower than the single piece bottom plate but maybe the carbon is also cut out of the sheet in an orientation to make them more rigid? It’s still a noisy frame but the filtering in betaflight does a really good job and it’s actually a lovely quad to fly. My biggest gripe about it is the arms break too easily compared to my more traditional frames.

[@ChrisRosser]

The individual arms do add thickness and therefore stiffness to the design. Armattan are making a tradeoff for durability with these lifetime warrantied frames. They are more noisy because the frames are designed to be more flexible/tougher in crashes. Apex video incoming!

[@ShebanFPV]

Wow! Very well shown and explained! Thank you. I have Marmotte and Badger, enjoying them both. Now, seeing all these resonance figures, could someone help to setup filters in order to mitigate those vibrations and oscillations please?

[@ChrisRosser]

Have a look at my filter tuning video. It might help.

[@Mekanikern]

Hi Chris! I must say I really appreciate your true scientific approach on these difficult topics. I have a question for you. Well, I actually have many questions but let's take one thing at the time. Looking at this typical Armattan approach - a wide relatively thin arm, if we cut a triangular shaped part of carbon sheet and glue it 90 degrees on top or bottom of the arms, would that not make a huge difference in arm/frame stiffness for very little weight? In my mind that would stiffen up the frame and raise the frame resonance to a much higher frequency in all axis?

[@ChrisRosser]

If you can get it really well bonded it should help a lot.

[@HotboiEngineering]

Hey Chris, love the content and as an engineering student its incredible to see the stuff I learn and do in class applied to a real world application. I have some Solidworks files for the Ummagawd Moongoat and FPV Cycle Glide, and have blackbox logs for both. Would you be interested in running a vibration analysis on either? Also, you mention the wide, flat arms of the Marmot being the main contributor to the resonate mode in the roll direction. The Rotor Riot CL1 VS frame has arms made of 6mm carbon, but they are rather skinny. Interested in your thoughts on how that affects roll and pitch resonance and what other resonate modes may result from that design.

[@ChrisRosser]

Glad you enjoyed the video! Yes I would love your SW files for the moongoat and glide with the BB files (gyro_scaled if possible!). Thick skinny arms are better than wide thin arms because the worst resonant modes usually involve the arms bending up and down and as you will know you get a cubic contribution to second moment of area with thickness in a vertical bending mode.

[@ChrisRosser]

You can find my email in the about section of my channel, looking forward to hearing from you.

[@ibnfpv]

General question about frame resonance , don’t we should look on the spectrum with the throttle curve and see if the noise is constant on the same frequency and not changing

[@ChrisRosser]

Yep that's how to spot a resonance. But you can see them in the frequency plot clearly as a spike also.

[@yonirodriguez2936]

Nice video! Thanks again! By the way, have you got any suggestion or information for a resonance analysis of the wizard x220 HV, thanks in advance!

[@ChrisRosser]

No plans to look at that frame, are you still flying it? I thought it went out of production years ago.

[@HormigaZ]

Love your work. Keep it up. Can you do the Drib Skyliner from Rotor riot?

[@ChrisRosser]

I would be happy to, but I would need a blackbox log and CAD for the frame or a frame to measure up and test fly. Would you be interested in sponsoring the video?

[@MrDirrrty]

I wonder if the Badger with removable arms has the same resonance issue?

[@KwaddawK]

Im interested in this too having broke my marmotte and swapped it out for the badger as I bet a lot of other people did as well.

[@ChrisRosser]

Geometrically they are very similar so should have similar modes of vibration but removable arms introduce significant damping at the joints which can reduce the amplitude of the resonance in the frame.

[@beefycol]

Great work. Turn the music down please. So if its got 250hz resonance what should the betaflight filters be on?

[@ChrisRosser]

No problem, I'll turn it down 6dB on future videos. If you have a 250Hz resonance you want to make sure that you have the dynamic notch filter on and set to a range where it can find it and attenuate it. You could also consider a static notch at that frequency for the gyro or Dterm

[@mbug7973]

I really like this video and I have some ideas for improvement: It´s a pity that altough you analyzed the resonance problems in a very precise way you did not show how to get rid of all the gyro noise by setting the appropriate filters. I really like your new series about filtering, PID tuning, etc, but especially for the filtering I would have loved to see some examples of successfull filtering approaches. Like "this is our noise problem, it is caused by this and that reason, I would set the filters here and here, this is the profile before and this is the profile after, magic!". Please keep up your awesome work!

[@ChrisRosser]

I have a video on tuning your filters. Check it out!

[@mbug7973]

@@ChrisRosser hey Chris, thank you for your answer! I know this video and I love it. And it still lacks somewhat of the feature I'd love to see. For example: your goal is to teach us how filters work, how to set them, how to analyze the causes of Gyro Noise etc. And this is what you have done wonderfully in your latest series and also in your frame resonance models. What I would love to see in addition is something like practical examples. Think about giving an exercise to your students for practicing purposes. Show us a Gyro Noise profile of a quad with specific problems in the form of gyro_scaled black box log. Describe how you would analyze the root cause of the problem, how you would set which filter, and then show how the noise profile looks after successful filtering. So that we can see something like a before / after effect. I am very sure you can do this, you are a great teacher! :-). Knowing all the theory is great, but the practical examples with successful solutions are often the things which give you the big aha-effect. I mean the ultimate goal is that we as your students become able to analyze and solve our quad problems so that we can make every crappy quad fly well. Maybe this helps you a bit.

[@ChrisRosser]

@@mbug7973 Thanks I'll try and work an examples video into my schedule.

[@mbug7973]

@@ChrisRosser this would be amazing!

[@tehllama42]

I'd be curious to see how this changes after adjusting some of the material properties of the carbon - I've found these frames to fly OK (still in need of a lot of filtering) when new, but as soon as delamination starts happening as a process, they go right off a cliff in terms of flyability.

[@rongrandt7960]

How do I get a hold of you. I did some testing on the marmotte and have some interesting info.

[@ChrisRosser]

You can find my email in the about section of my channel. 😁

[@BelovedFPV]

TBS Source One V3 Please :)

[@stevo43068]

No suggestions to try and improve kwads using this frame?

[@mattraymond2935]

He suggested struts

[@dod.fpv_]

I somehow missed this video entirely :D i designed a 6mm unibody bottomplate for the marmotte. Any chance of you testing it? would be interesting to see if it made a differance. I saw an increase in flight performance at least and its much better in crashes now than it was before

[@SKoToBoT]

How would you help filter these problems

[@neuriomat]

I Have the Armattan Badger 6", nearly the same Frame and have exact the same Noise Spike. especially with biprops the Spike is even much bigger. I would really like to know why biprops produce significant more resonance. What exactly is the "Frame Design" you need? Black Box Log is clear, but Frame Design not. Greetings

[@mfrfpv9079]

Where to get the braces that go between the arms?

[@RCRitterFPV]

Sorry just typing while watching while kids yelling... soo.... apologies, I'll do this ramble often.. But I'm a Patreon now so :) can we somehow apply these known resonance as a filter with the onboard gyros / FC... also does, how the force is applied make much difference? slower smoother application of power by esc tuning change how these manifest? smooth motors, notchy motors, aggressive timing... or motor RPMs and resonant frequencies. we know X frame resonates at Y Hz. reduce motor power on Z Rpm's..... or something...

[@ChrisRosser]

Absolutely! With the new dynamic notches in BF 4.3 we should be able to set the right number of notches based on the number of vibrational modes present in the frame and the filter code should find and crush the resonances. ESC tuning can definitely help as can PID tuning but to improve vibration we have to detune PIDs and ESCs so its not ideal for flight performance where we would like the gains to be as high as possible.

[@mattraymond2935]

Aramttan should make 6mm narrow 4 in 1 esc arms like the cl1 vannystyle

[@ChrisRosser]

That would certainly help. They should also try to squeeze another standoff into the design so the top plate has 8 points of contact rather than 4 or 6

[@Vesty20]

Do the badger for a comparison!

[@robbinvanhoek121]

Interesting stuff! For my marmotte the 250Hz resonance is completely absent in the blackbox.. i only see a significant resonance at 335Hz on both roll and pitch. If anything then on roll i get a second (must less pronounced) resonance at 206Hz. How would you explain this?

[@ChrisRosser]

Do you have space carbon by any chance?

[@robbinvanhoek121]

@@ChrisRosser yes indeed!

[@ChrisRosser]

@@robbinvanhoek121 So that explains it. The extra stiffness of the space carbon pushes the resonance up from 250Hz to 335Hz. This suggests that the space carbon is ~34% stiffer than the regular carbon!

[@robbinvanhoek121]

@@ChrisRosser thats pretty cool! is that anything i should be concerned about? do i need a notch there?

[@ChrisRosser]

@@robbinvanhoek121 The dynamic notch filter in BF should clean it up so I wouldn't worry about adding another notch.

[@videotube320]

Is this the reason my motors are hot and I struggle to find a setting to keep them cool?

[@ChrisRosser]

Hard to know. But if you have lots of noise in the BB log its possible.

[@JaredMcConnellFerzen]

This is why I am replacing my chameleon-ti with the Apex, I could not get the noise out. It would resonate so badly at 80% throttle that my FPV video would have noise introduced, and I was loosing screws and my lens on my FPV camera became so damaged from vibration I had to replace it.

[@ChrisRosser]

Yes, lots of people have the same experience I'm afraid! That's why I designed the AOS 5

[@seewell8168]

The science behind why I’ve been so dissatisfied with each Armattan frame I’ve tried 😁.

[@ChrisRosser]

I've reached out to Armattan to offer to help them with their next design... perhaps the next frame will surprise you?

[@zachcarrizales5038]

Doesn't the anisotropic behavior effect this type of simulation?

[@ChrisRosser]

The material is anisotropic as modelled and it only needs to be an approximation to visualise the mode shape at a similar frequency to what's seen in the blackbox log

[@ariafpv]

@@ChrisRosser What exact material do you use? Does it matter a lot if I choose e.g. ABS or steel to simulate it?

[@ChrisRosser]

@@ariafpv You can find some material parameters online for carbon fibre. That's what I've used. It doesn't matter too much whether your parameters are absolutely spot on as I am more interested in the mode shapes rather than the exact frequency. I get the frequency from the BB logs

[@JohnWhite84]

Great now I gotta sell all my Marmottes

[@IconicProps]

I have this frame with premium parts, F7 stack, imperial motors, stubby antenna, dji polar vista, that I have set up to what seems perfectly acceptable, and it vibrates like a loose hubcap. I have no clue what to do. Has anyone else had a terrible amount of vibration in this frame? It flies and sounds like its got a broken prop. Brand new pbj's

[@ChrisRosser]

Best option: Transfer the build to a different frame. The marmotte has issues with resonance performance. If I'm allowed to recommend my frame the AOS 5.5 that is the highest performing 5.5" frame I know of regarding resonance. Quick option: Try BF4.3 and hope the SDFT can do a better job of controlling that resonance.

[@mfrfpv9079]

The testing is flawed this frame is designed for descreate esc it is part of the structure and makes a difference would make a good revisit if it shows in the data.

[@Aridrona]

Dont forget your frame needs to look cool

[@ChrisRosser]

#1 priority I assure you 😋

[@Krunked]

nice man -- ive been saying this about armattan frames forever now.. theyre terrrrrrible. itsblunty sent me here

[@ChrisRosser]

I'm glad you enjoyed the video, it means a lot coming from a tuning guru like yourself! I hope that a deep dive into why the problem exists helps Armattan make better frames in the future. Itsblunty and I have been talking a bit about this, please share with others you think would be interested.

[@weberentertainment2962]

@@ChrisRosser chris from armattan will not listen to critisism, and i quote "what does a guy with only under 200 subs on YT know about anything" XD i will never go back to armattan!

[@ChrisRosser]

@@weberentertainment2962 Did Chris see my video?

[@weberentertainment2962]

@@ChrisRosser honestly I don't think he did, but the video was shared on the armattan Facebook groupe, the comment section on that thread was shut off.. He didn't want to hear about it, dismissed it with we know more than he does, but no real answers to the questions of people just trying to debate it🤷‍♂️

[@ChrisRosser]

@@weberentertainment2962 Thanks for the update. It's a shame that they don't seem interested in engaging with the community about the problem and doing an updated version with improvements. It wouldn't be difficult to fix these issues :(

[@AleilerFPV]

интересный обзор. а что ты думаешь про вот эту раму? kzbin.info/www/bejne/ppyYn2yBm8qXg68