I'm an Aerospace engineer and I just discovered your channel as I was doing research to pick up my first hobby-grade multi-rotor. You've definitely got some good stuff and I think you are spot on with this video. Multi-rotors rely on their FCs to stay in the air and the better the data going into them, the better the FCs can manage the overall performance of the vehicle. There is a downside to soft-mounting, though. The soft-mount trick works because it allows the motors and the props to "wiggle around" on their mounting pad, so the mounting material actually absorbs/damps out a lot of the high frequency vibrations originating on the rotor side of the interface, such as vibrations from an iffy bearing. The problem is that this also allows the rotor head to slightly change angles based on the vibrations being generated by the system and the larger torques being created by the motion of the aircraft. These motions should be pretty small, but even a few fractions of a degree could be enough to cause the control logic to have to take action to actively try to suppress this motion in the control loop. Will you feel any of this action? Probably not. However, it could result in a slight "softening" of some of the control inputs by introducing a slight "lag" in the control channel. Imaging the nice rigid rotor arms are flapping around slightly under load, so it takes a fraction of a second for the system to settle down and return to a steady state when it is upset or you introduce a new control input. Again, this will probably be way below the threshold of what most folks would notice, especially for a vehicle in this class, but it is a real byproduct of having anything "soft" in a system that expects everything to be rigid. Again, great video! As a noob to this corner of RC, I have learned a lot from you so far.
I enjoy listening to you Joshua because you keep talking without stutter,coughing,excuses of any kind and distracting from the subject.You have a knack of keeping to the point and my attention.Good job sir.As far as the vibrations are concerned I’d say insulating the arm to the frame ,if possible would be less tedious and also the motors can still keep their ridged mounting without the wobbling.
Thanks much again for your (very valuable) time Joshua. Just at the first motor here. The tape I had sitting around here in OZ is Scotch Outdoor. Cut mine at about 60mm before folding to create a couple of small outer corner 'bumpers'.
"It Done Got Better".LOL Great analysis as ALWAYS!!!! Thank you So Much for taking the time to actually show us some RAW Data!!!:) You are Sharp as a TAC My Friend!!! BTW. Thanks *Johnny FPV*, and *Willard FPV* ALSO!!!
Perfect timing on this vid :) was going to soft mount tomorrow in together with an motor swap. with you confirming the possible benefits I got some new thoughts on how I'm going to do it. I run an tricopter with bracket mounts and the fc is fixed to all. RxExplorer mini. so you gave me thoughts of the end mounts of my boms :) Keep up the good work and help us thats new to this :)
I spent a great deal of time on this issue in respect of tail rotor gyros for model helicopters back in the 1990s and I'm convinced that isolating the motors from the frame is a better solution than trying to rely on soft mounting of the FC on its own. My reason for saying this has several factors. Firstly, the range of frequencies of vibration in the airframe with the motors hard mounted is wide - the beat frequencies can obviously be anything from zero upwards. When it comes to soft mounting the FC you have two competing requirements - you want the FC's attitude to accurately track the attitude of the airframe that, in rapid roll, pitch and yaw accelerations (OK, we old heli guys will tell you quads don't really do decent yaw acceleration but I digress 8-)=) requires a low compliance to the mount while the need to isolate low frequency vibrations requires a high compliance to reduce the resonant frequency of the mount. By isolating the motors from the frame you are only needing to isolate the relatively high frequencies of each individual motor and not the much lower beat frequencies that exist in the frame with hard mounted motors. I think this easily explains why, despite the motor being quite firmly mounted, you are getting a significant reduction in the low-frequency components in the gyro traces. There is another effect which I think explains why you are seeing the relatively slow meanderings of the raw gyro traces (especially the yaw axis). This arises from the fact that, although the sensors are 3-axis, each element of the sensor is actually only a single axis device - sensitive only to rotation about the axis it is currently pointing. The effect of this is that simultaneous vibrations in pitch, roll and yaw can cause a single axis sensor to have a non-zero mean output when there is no net rotation about its nominal sensitive axis. To see how this comes about consider the following movement of the quad and consider what the yaw gyro actually sees. Tip quad 90 degrees nose down Roll quad 90 degrees right Tip the quad 90 degrees nose up (back to level) Yaw the quad 90 degrees left The quad is pointing in its original direction but the yaw axis gyro only saw the last of these motions and its mean output through this sequence is thus displaced to the left! The important thing is to realise this happens to a degree even when the angles of motion in the three axes are very small. So when you have the sensor experiencing simultaneous vibrations in all three axes its output will generally be displaced (the degree and direction depending on the amplitude and relative phases of the pitch, roll, and yaw vibrations and this accounts for the stochastic nature of the disturbances). The result is to get small amounts of drift in the yaw axis (reduced if there is a magnetometer to give another heading reference to add to the Kalman 'mix'). Twitches in pitch and roll are also to be expected as it takes time for the accelerometers to take gyro errors out via the Kalman filter. The final important point is that this vibrationally induced drift effect occurs even if the individual gyro sensors are “perfect” i.e even if they don't have any g-sensitivity or out-of-axis rotation sensitivities.
I watched this video the same day it was posted but never soft mounted my motors out of laziness... I soft mounted them today and just WOW!! It really maid a difference. I had some problems with vibrations and I had soft mounted my FC but it wasn't enough and soft mounting the motors really got the problem fixed.
This has transformed my quad video footage. Had some similar tape at home, used it on the wizard which had suffered vibrations after a crash and now the view through the runcam eagle is silky smooth. Thanks for the vid!
Mr. Bardwell, I've been experiencing crazy noise in my FPV goggles, especially when at high throttle punch outs. I've ripped my quad apart so many times in an attempt troubleshoot the issue but to no avail. I switched cameras rerouted wires and still nothing. I was beginning to believe that it was my PID settings. At the time I was using Cleanflight but recently switched to Betaflight. I finally got my quad to fly super buttery and locked in, but the damn FPV goggles still had a isht ton of noise. I finally soft mounted my motors and now my video feed is solid. Thank you. I also like to thank you for many of your other videos as well. So thanks again! I'm in Sherman Oaks and would love to rip with you one day. Best regards, -The First African American Quad Racer - QUADristotle
Excellent. Glad it made a difference for you. You can also add capacitors to your ESCs which will probably fix the noise, if you didn't do that already.
I do in fact have nasty vibrations in my Tarot 130. 2 motors came buttery smooth, 2 sound and feel like the bearings have rusted out! Will try this and report back. Thanks for taking the time to test this and report back to the community.
Hello Josh! Saw this right now here in 2018 and i do have some technical knowhow in filterdesign to share that might help understand why there are still noises in the Gyro Signal (also an idea on how to reduce vibrations even further). Filters in genereal: There are basically two kinds of filters, HIGH PASS Filters which cut out only low frequencies from a signal, and LOW PASS Filters which cut out only high frequencies. You can also combine those two Types (but that doesn't matter here). As you can imagine, what is used in the gyro-application is a LOW PASS filter, as we want to reduce high frequencie noises that mess around with our gyro-feed. The reason why you still have noise after the filters are applied: However, the term "CUT OUT a frequencie" doesn't really apply to what a filter is able to do. Filters, no matter if analogue or digital (what is used in every gyro application for quads) can only dampen a noise down to a level where it doesnt matter anymore, and they do it gradually from a fixed frequencie on. (See: upload.wikimedia.org/wikipedia/commons/thumb/6/60/Butterworth_response.svg/350px-Butterworth_response.svg.png After the fixed Filter Frequencie (cutoff frequencie in the graph), the damping is not straigt, it is gradually) The gradual decrease of the noise intensity above this fixed frequencie is the real reason you can still see noise after the filter applied, because it is simply too close or even before the fixed frequencie. Other Way to reduce vibrations (not tested, since i never flew anwhere else than in the sim lol): To do that, you could also try to softmount the arms that hold the motors. This approach is not possible with every frame, since you would need the arms to be mounted individually, but it should work on the wizard.
Thanks Joshua. I'll try this tonight. My quad flew away yesterday but lucky it went directly straight up to the sky since I was hovering. My setup is Emax Eco 2306 2400kv, Spedix GS45 4in1, Hyperlite F4 OSD v2 flight controller and 6" props. I could not believe what I saw and heard in the span of 2 seconds. I was hovering the quad and it was about 10ft away from me when suddenly, it flew to 150ft high and I was so friggin scared since I'm a newbie to this. I wasn't aware that flight controllers should be soft mounted. I was told yesterday in one of the Discord channels. I'll fix it later. I'll try soft mounting the motors too, Thank you Joshua!
Great idea Joshua, people in general DO forget the physics behind many things in the RC world. Great advice, even if it's old news. Thank ya sir and...See you in the air. Wayne
Don't let theory ruin a good experiment! AWESOME! I get excited when my hypothesis is wrong! Most people try to force their hypothesis into the result. Not good science folks. Your killing it as always Brother! Keep it up.
I learned this years ago when I soft mounted my wife's thunder stick 2000 to the headboard! I have slept like a baby since lol good stuff man thanks for sharing this information with us 👍🏽👍🏽
Thanks for taking the time to do this experiment and post your results. I think I may give the soft mount a try with the same tape and see if I can tell a difference. I'm running the TBS Powercube on my builds and for the most part they fly smooth, but curious to see if this makes things even better.
Very nice. Thanks for speaking on this. I soft-mounted mine and thought I could notice a difference, but also wasn't sure if it was just in my head. haha ... thanks again!
Your videos are amazing! Thank you for being so scientific in your processes. Too many multirotor channels lean hard on their opinions vs objective results. Subscribed!
Finally a video with some objective data and proofs. Thanks Joshua. Instead of electrical tape I'm planning in using silicone backing sheets, 1mm thickness. Softer, cheap, doesen't leave a sticky mess and is high-temperature resistent. Also, I think is better using only 2 screws for motor (A' la FGA or MrSteele), it will give more wiggle room. Edit: [On a side note: If you find the perfect silicone backing sheets already in your kitchen, better ignore them and order some new one from Amazon. Your significant other may disapprove.]
Thanks for this vid. I like the fact that you take the time to show us these theories and whether they work or don't work, rather than random pilot on the internet saying "it works" or "it's b.s." Without any real evidence.
Finally! Somebody did a proper video on this! Gonna have to give it a try on my quad as well now. Might not see any change right now, but as my motors wear out and get noisy, this should help keep it flying the same.
LoL at the beginning. I have similar shakes at around mid throttle, but the gyro traces are really clean. Haven't been able to tune it out and have been at it for months. It was buttery smooth with my earlier setup 20A little bees and 2204 Cobras. After switching to Xrotor 20A micros and Emax RS 2205 the shakes started.
Willard, Johnny and Kevin all had vids about the same time on this - good flight videos after doing it but no empirical data. Again, thanks for performing the test and publishing Joshua, this is exactly why vlogging _ISN"T_ killing the hobby but instead educating and making everyone a better pilot.
Vlogging with information is fine. Just because a guy is on camera talking doesn't make it vlogging. I am personally bored by vlogging about vapid stuff, but to each their own.
I asked you the other day about soft mounting motors . . . I got the rubber one's off amazon. . . . Currently not using them because I'm one of the people who aren't really having that problem. . . And they are a HUGE pain to get on correctly . . Then you run the risk of motors flopping off . . . And were talking about super glue . .. my quad flies fine . . Thanks for the advice !
Thanks Joshua! I have a friend who soft mounted his motors. His hd footage before showed blurry leaves on trees but afterward the leaves were crisp. Thanks for the black box backing to support his experience!
That's a benefit I don't see mentioned much in the discussions of soft mounting: less jello in HD camera footage. Thanks, Joshua, this is very compelling data. You are the Galileo to all the Aristotelian pundits out there.
3m 4010 clear mounting tape works great. When a similar discussion went on in the rx helicopter world it was settled(mostly) that this is the best tape to mount gyros and flybarless units. I've been using it to soft mount my motors and it works well for me. Like the grey stuff it's a bit thicker than 4 layers of electrical tape but it compresses without loosing its ability to dampen vibrations. I am curious to see how the 3D printed mounts work since they would be much cleaner and easier.
Ok now that I have watched the whole thing, that's really telling. I suspected that we would see results like this after thinking it through a bit. It's the same reason electricians put rubber grommets on big transformers when they hang them. It dampens the vibrations and keeps the "hum" down.
Thanks Joshua. I was anxiously waiting for your test, but was thinking "Joshua wouldn't touch this topic with a ten foot pole, would he?". Anyway, great results! A couple of us did post similar controlled tests on Facebook, which confirm your observations. Most who payed any attention to it were convinced instantly, but the vast majority were laughing too hard to notice.
Use a slightly longer screw and "soft mount" the screw side as well then you escape the contact of 1 axis of the screw. Bore out the holes in the frame and line it with mounting tape or plastidip, then it is fully isolated. Has anyone seen a spring setup? Having thoughts of the old car cd changers that still skipped. All this makes a difference but how much? Most of us are park flyers not aerial photography pros, so is it worth it for the hobbyist? FPV racer? Would be interested in seeing fpv with the data of each in a more controlled setting ie indoors with set parameters like a warehouse loop. Thank you for making great content, super informative. You Rock!
Quality!! 😁 Very much enjoyed watching. And a I am not only talking about the purely scientific part of "measuring". Received it as more entertaining than potentially initially intended by you. Very good!
I beat the hell out of my quad, fly PC props and don't change them often, and I don't like swapping out motors once per month. I had great results from the electrical tape trick. However, after exposing my quad to moisture and mud and abuse, the tape started to peel and trap dirt underneath. Last night I applied 5 coats of liquid electrical tape to my mounts. I'm pretty happy with its adherence to the carbon fiber, and it even coated both sides of the screw holes. I'll be giving it a test fly today. Thanks for your comprehensive workup, Joshua!
+1 on the soft mounting motors. Had a frame with yaw twitches I couldn't resolve. Soft mounted the motors with electrical tape after watching Johnny's video, and everything is right with the world again. Definite +1 on the need for loctite however. Only 3 packs through the quad and everything was coming apart.
Have a go at putting silicone washers under aluminium 'cone' washers on the bolt side of the arm to supplement what you've already done. I've done so in the past while playing around with reducing vibration for an APM board (with mixed success). It would be very interesting to see what data comes out a black box log with modern hardware.
@Joshua Bardwell: I´m working as a engineer in the semiconductor manufacturing industrie and know "a bit" of PID, gyros, dampening, dynamics of vibrations etc. Because to produce wafers you realy need a good understanding how to keep your PID controllers dealing with the slightest vibrations. What do you think why it´s possible to do lithography in some nanometers resolution to produce computer chips? :-) I´m realy glad you did that video. After all the haters and people knowhing nothing about this just laughing and shouting "hypetrain" you did a great job showing the results. You´re right with most of your assumptions and did a realy great video. There are some other reasons why the connection of the screwheads are not a so big deal, not only the surface area, but your conclusion is damn right. Of course soft mounting is reducing the vibrations that your PID controler has to deal with. And of course, some people will feel a bigger effect that others, this is not only because of the motors, but also about the frame design. There are even enough quads with so less vibrations, that the gyro filters will output a clean enough signal. The most important thing beside getting rid of the vibrations is, that you indeed (in some cases) can tune your PID´s much better, resulting in higher P and lower D and a better tuned quad. I´m personaly sick and tired of discussing with haters who do not even listen and trying to understand, just trolling around and making jokes of others. Just look at all the jokes and memes in the Facebook groups. :-( Nice Job Joshua Bardwell, you have one subscriber more. Just one last hint: If you look at the spectrum of the vibrations without soft mounting, you may notice that the biggest amplitude varies in frequency between differend quads. If you would like to push the soft mounting further you could do that with selecting different hard soft mounting materials, because they will dampening better on specific frequencies. This can be achived with polyurethane elastomers in different shore hardness. There is no vodoo, this is just known engineering in some industries. :-) Keep on with your great videos!
I love this, watched it too many times haha.. The next big thing for 2019 I hope you guys cover and stating soon, is frame resonance and vibration and how it impacts your flight experience. I recently tried to build a 6s using an iFlight battleaxe frame. I recon if you want a prime example of a bad frame that will burn your stuff on day one, its this one. I can send you blackbox data on this as well.
Great information. Time to cast some polybushes and drill out the mounting holes so even the screws don't touch! PU is available in a range of hardness grades. I build my micro quads from acetal because that's what I have available and I can cut it on my laser. Many people have been sceptical of acetal because it is not as stiff as carbon. However for a micro I think it is stiff enough. The fact that it is more flexible than carbon might mean less vibration is transmitted to the FC.
Mass vs. Vibrations; if the mass of the entire frame is isolated from the motors but hard mounted to the FC we have the mass working for us. Alternatively, when the entire frame of the quad is vibrating along with 99% of the mass and then we expect the mass of the soft mounted FC ( a few grams) to stabilize on a few o-rings we get very poor results. This makes me think of back in the day when David Windestal started making his Tricopter frames, the camera plate was suspended from the flight frame and included the mass of the battery. Now off to print some flexible motor mounts! Thanks for bringing this to light Joshua, not all of us have time to lurk in the forums or the tolerance to wade though people's BS! :-)
fpvYou interesting use of Corrosion-X. Do you need to re-apply bearing oil afterwards? (Like how a more refined oil will act as a solvent on a less refined oil -- I don't know how Corrosion-X compares to bearing oil like Scorpion.)
My quad flys so much better since i softmounted! really happy with the difference! You are a master of tuning josh! I would kill to get a glimpse of your final PID's for the wizard Mine came with 2 really bad motors as well :(
Awesome video Josh. I like your comedic additions, namely your opening and homeopathic comments! Not to mention your clear explanation of why soft mounting isn't BS. Sadly I think most of the soft mounting haters on the rotor riot Facebook group are not the kind of people that enjoy your nuanced and scientific approach to these kinds of questions. Much easier to just make a meme than think critically
To prevent the heads of the screws touching the arm, you can apply a spring between the head and the arm. By choosing the strength of the spring, you can also define the force of pressing the motor against the foam. Maybe a "Tellerfeder" (sorry, I don't know the English word) can also do a good job.
I read somewhere before, yes 2 years ago, I think 3 years ago either about soft mounting these motors. This was when we still fly these 450mm hex, octo, all those bunch of big multirotor.
I've been doing this for a long time, even before it was talked about like now, the benefits became clear very quickly. Now I can even buy pads specifically for this on eBay. Place them under the motors and with the motor protector under the arm the screws even tighten onto that, not the carbon. You can heard the motors sounding cleaner in the air, to me this is a must because of the benefits I achieve.
Now the verdict has been finalized. I guess people will use torque hex key on specification material(e.g. silicon, tape, washer, etc) to fine tune the softmount. Sooner or later, there will be specify products for the softmount
Nice vid thanks josh. PS There's a lot of noise across the whole spek analyser- especially in areas low frequency which betaflight can't filter. I had a similar problem and it ended up being ground loop by connecting both vtx grounds might be worth double checking eaachine didn't balls it up?
Excellent presentation. From the first time I saw the soft mount videos, it made sense. Will I do it? Probably not. I'm not that good a pilot and I don't race, yet. All my birds are low budget quads. So I don't really think any amount of insulation from vibration will make much difference. But for those that have spent tons of money on their gear, this should probably be something for them to try. There will always be naysayers. That's the way the internet rolls. :)
awesome. I was too lazy to do this myself. I just tested a 1.5mm TPU pad on my motors but I need to get longer screws to feel comfortable. So ideally, the motor screw holes should be outside the base of the motor because then you can use a lock nut and then you don't need to cinch them down so hard.
Yo great video bro 👍 I've just changed the bearing out on my x220 wizard as yeah they Was rattley lol! now there nice and tight and run nice and smooth. I could feel the vibration from the motors by just spinning the motors by hand and definitely could hear it when flying 😆 And I've also tpu soft mounted them too and wow I can feel the difference in them just by hand through the frame with spinning them 👍 great job I'll have to fly it tomorrow as I have a esc to replace as that crapped out on me just over night 😆 so I'll give you a update on my findings 👍
Love your videos Joshua. I wonder what the results would look like if you had done this test with electrical tape. I'm assuming the 3M tape has a better effect due to it's thickness. I want a cleaner look, so I'm planning on using the rubber o-rings that I use for the FC, good idea you think?
Thanks for sharing. Very interesting. I started FPV back in 2012 when things like the QAV400 were the go to FPV quads, which back then was actually described as a small aerobatic frame, ha. Most of the popular frames around then had soft mounting built into the design. At the time I just figured they stopped including it in designs because software and filtering had improved and 250 miniquads didn't need it because of the smaller scale.....obviously not entirely true. I think maybe the Blackout 250 miniquad (now old too) moved the trend away from soft mounting and vibration dampening. When the Blackout 250 came out people were mostly using smaller 1800 sized motors on it. I guess now that motors have gradually moved back to larger sizes with bigger magnets and more torque; vibration issues are coming to attention again. Wonder if frames will start to move back to including vibration dampening into the design.
Yes yes yes! I backed out the screws on my motors and bye bye vibration. Ran the same motors on a different style frame, no vibration. Frame design must just resonate at a certain frequency.
Very nice and technical video showing some difference in soft mounting. I do agree with the mounting tape and I would even try small o-rings because it is thicker and would dampen the vibration more. My question is how tight was the first try and how tight was the tightened test? Is there an objective guage on how to tighten the screws and have a difference and not have your motor fall off? Do we tighten all the way and back off 1/16? Your videos are always very informative. Thanks
Awesome idea! Have you thought about adding a bit of padding to the screw head side when mounting? It may create even more vibration isolation as it reduces the problem you brought up in the beginning (how the quad still has contact to the frame), as well as possibly allowing you to put more torq on the screws for added safety. Maybe it will make the ride that much more "juicy"? Haha
Shout out to all my haterz
Oh come on, you don't have any haters
I'm not one, I agreed with you when I first saw your video.
Hahahaha! Feels good, doesn't it?
now you should soft mount your glasses, so juice doesnt splash out when you put them on the table.. hence #evenmorejuicy
Johnny FPV my hats off to you sir
As an engineer myself, I applaud you for answering the mail with some science. I am upping my Patreon contribution.
I'm an Aerospace engineer and I just discovered your channel as I was doing research to pick up my first hobby-grade multi-rotor. You've definitely got some good stuff and I think you are spot on with this video. Multi-rotors rely on their FCs to stay in the air and the better the data going into them, the better the FCs can manage the overall performance of the vehicle. There is a downside to soft-mounting, though.
The soft-mount trick works because it allows the motors and the props to "wiggle around" on their mounting pad, so the mounting material actually absorbs/damps out a lot of the high frequency vibrations originating on the rotor side of the interface, such as vibrations from an iffy bearing. The problem is that this also allows the rotor head to slightly change angles based on the vibrations being generated by the system and the larger torques being created by the motion of the aircraft.
These motions should be pretty small, but even a few fractions of a degree could be enough to cause the control logic to have to take action to actively try to suppress this motion in the control loop. Will you feel any of this action? Probably not. However, it could result in a slight "softening" of some of the control inputs by introducing a slight "lag" in the control channel. Imaging the nice rigid rotor arms are flapping around slightly under load, so it takes a fraction of a second for the system to settle down and return to a steady state when it is upset or you introduce a new control input.
Again, this will probably be way below the threshold of what most folks would notice, especially for a vehicle in this class, but it is a real byproduct of having anything "soft" in a system that expects everything to be rigid.
Again, great video! As a noob to this corner of RC, I have learned a lot from you so far.
"Take it from a guy who's looked at alot of black boxes" could be the best out of context line ever.
Glad I'm not the only one who laughed at that.
"Johnny says it makes his quad more JUICY..." Best quote of 2017. 😂
I enjoy listening to you Joshua because you keep talking without stutter,coughing,excuses of any kind and distracting from the subject.You have a knack of keeping to the point and my attention.Good job sir.As far as the vibrations are concerned I’d say insulating the arm to the frame ,if possible would be less tedious and also the motors can still keep their ridged mounting without the wobbling.
Thanks much again for your (very valuable) time Joshua.
Just at the first motor here. The tape I had sitting around here in OZ is Scotch Outdoor.
Cut mine at about 60mm before folding to create a couple of small outer corner 'bumpers'.
Charpu tape just went up to $30 a roll now.
The comedic value in the first part of this video is beyond words....
"It Done Got Better".LOL
Great analysis as ALWAYS!!!!
Thank you So Much for taking the time to actually show us some RAW Data!!!:)
You are Sharp as a TAC My Friend!!!
BTW. Thanks *Johnny FPV*, and *Willard FPV* ALSO!!!
Perfect timing on this vid :) was going to soft mount tomorrow in together with an motor swap. with you confirming the possible benefits I got some new thoughts on how I'm going to do it. I run an tricopter with bracket mounts and the fc is fixed to all. RxExplorer mini. so you gave me thoughts of the end mounts of my boms :) Keep up the good work and help us thats new to this :)
I spent a great deal of time on this issue in respect of tail rotor gyros for model helicopters back in the 1990s and I'm convinced that isolating the motors from the frame is a better solution than trying to rely on soft mounting of the FC on its own. My reason for saying this has several factors. Firstly, the range of frequencies of vibration in the airframe with the motors hard mounted is wide - the beat frequencies can obviously be anything from zero upwards. When it comes to soft mounting the FC you have two competing requirements - you want the FC's attitude to accurately track the attitude of the airframe that, in rapid roll, pitch and yaw accelerations (OK, we old heli guys will tell you quads don't really do decent yaw acceleration but I digress 8-)=) requires a low compliance to the mount while the need to isolate low frequency vibrations requires a high compliance to reduce the resonant frequency of the mount.
By isolating the motors from the frame you are only needing to isolate the relatively high frequencies of each individual motor and not the much lower beat frequencies that exist in the frame with hard mounted motors. I think this easily explains why, despite the motor being quite firmly mounted, you are getting a significant reduction in the low-frequency components in the gyro traces.
There is another effect which I think explains why you are seeing the relatively slow meanderings of the raw gyro traces (especially the yaw axis). This arises from the fact that, although the sensors are 3-axis, each element of the sensor is actually only a single axis device - sensitive only to rotation about the axis it is currently pointing. The effect of this is that simultaneous vibrations in pitch, roll and yaw can cause a single axis sensor to have a non-zero mean output when there is no net rotation about its nominal sensitive axis. To see how this comes about consider the following movement of the quad and consider what the yaw gyro actually sees.
Tip quad 90 degrees nose down
Roll quad 90 degrees right
Tip the quad 90 degrees nose up (back to level)
Yaw the quad 90 degrees left
The quad is pointing in its original direction but the yaw axis gyro only saw the last of these motions and its mean output through this sequence is thus displaced to the left!
The important thing is to realise this happens to a degree even when the angles of motion in the three axes are very small. So when you have the sensor experiencing simultaneous vibrations in all three axes its output will generally be displaced (the degree and direction depending on the amplitude and relative phases of the pitch, roll, and yaw vibrations and this accounts for the stochastic nature of the disturbances). The result is to get small amounts of drift in the yaw axis (reduced if there is a magnetometer to give another heading reference to add to the Kalman 'mix'). Twitches in pitch and roll are also to be expected as it takes time for the accelerometers to take gyro errors out via the Kalman filter.
The final important point is that this vibrationally induced drift effect occurs even if the individual gyro sensors are “perfect” i.e even if they don't have any g-sensitivity or out-of-axis rotation sensitivities.
I watched this video the same day it was posted but never soft mounted my motors out of laziness... I soft mounted them today and just WOW!! It really maid a difference. I had some problems with vibrations and I had soft mounted my FC but it wasn't enough and soft mounting the motors really got the problem fixed.
This has transformed my quad video footage. Had some similar tape at home, used it on the wizard which had suffered vibrations after a crash and now the view through the runcam eagle is silky smooth. Thanks for the vid!
Glad it helped. Your gyro is probably happier too.
Mr. Bardwell, I've been experiencing crazy noise in my FPV goggles, especially when at high throttle punch outs. I've ripped my quad apart so many times in an attempt troubleshoot the issue but to no avail. I switched cameras rerouted wires and still nothing. I was beginning to believe that it was my PID settings. At the time I was using Cleanflight but recently switched to Betaflight. I finally got my quad to fly super buttery and locked in, but the damn FPV goggles still had a isht ton of noise.
I finally soft mounted my motors and now my video feed is solid. Thank you. I also like to thank you for many of your other videos as well. So thanks again! I'm in Sherman Oaks and would love to rip with you one day.
Best regards,
-The First African American Quad Racer -
QUADristotle
Excellent. Glad it made a difference for you. You can also add capacitors to your ESCs which will probably fix the noise, if you didn't do that already.
I do in fact have nasty vibrations in my Tarot 130. 2 motors came buttery smooth, 2 sound and feel like the bearings have rusted out! Will try this and report back. Thanks for taking the time to test this and report back to the community.
Better to replace the bearings, but this will do in the mean time.
Joshua: Thanks for sharing. The world is better thanks to people like you.
Hello Josh!
Saw this right now here in 2018 and i do have some technical knowhow in filterdesign to share that might help understand why there are still noises in the Gyro Signal (also an idea on how to reduce vibrations even further).
Filters in genereal:
There are basically two kinds of filters, HIGH PASS Filters which cut out only low frequencies from a signal, and LOW PASS Filters which cut out only high frequencies. You can also combine those two Types (but that doesn't matter here).
As you can imagine, what is used in the gyro-application is a LOW PASS filter, as we want to reduce high frequencie noises that mess around with our gyro-feed.
The reason why you still have noise after the filters are applied:
However, the term "CUT OUT a frequencie" doesn't really apply to what a filter is able to do. Filters, no matter if analogue or digital (what is used in every gyro application for quads) can only dampen a noise down to a level where it doesnt matter anymore, and they do it gradually from a fixed frequencie on. (See: upload.wikimedia.org/wikipedia/commons/thumb/6/60/Butterworth_response.svg/350px-Butterworth_response.svg.png After the fixed Filter Frequencie (cutoff frequencie in the graph), the damping is not straigt, it is gradually)
The gradual decrease of the noise intensity above this fixed frequencie is the real reason you can still see noise after the filter applied, because it is simply too close or even before the fixed frequencie.
Other Way to reduce vibrations (not tested, since i never flew anwhere else than in the sim lol):
To do that, you could also try to softmount the arms that hold the motors. This approach is not possible with every frame, since you would need the arms to be mounted individually, but it should work on the wizard.
Thanks Joshua. I'll try this tonight. My quad flew away yesterday but lucky it went directly straight up to the sky since I was hovering. My setup is Emax Eco 2306 2400kv, Spedix GS45 4in1, Hyperlite F4 OSD v2 flight controller and 6" props. I could not believe what I saw and heard in the span of 2 seconds. I was hovering the quad and it was about 10ft away from me when suddenly, it flew to 150ft high and I was so friggin scared since I'm a newbie to this. I wasn't aware that flight controllers should be soft mounted. I was told yesterday in one of the Discord channels. I'll fix it later. I'll try soft mounting the motors too, Thank you Joshua!
you fixed my oscillation problem with the x220s. i had the exact same double sided tape and it flies so stable now!
Great idea Joshua, people in general DO forget the physics behind many things in the RC world.
Great advice, even if it's old news.
Thank ya sir and...See you in the air.
Wayne
Love your videos man - you've got a rare combination of brains and humor - keep em coming.
Interesting Joshua! I have some similar tape laying around. I think I'll try this. I love the data driven analysis!
Don't let theory ruin a good experiment! AWESOME! I get excited when my hypothesis is wrong! Most people try to force their hypothesis into the result. Not good science folks. Your killing it as always Brother! Keep it up.
I learned this years ago when I soft mounted my wife's thunder stick 2000 to the headboard! I have slept like a baby since lol good stuff man thanks for sharing this information with us 👍🏽👍🏽
Don't worry, pal. This comment didn't go unnoticed 🏆
Thanks for taking the time to do this experiment and post your results. I think I may give the soft mount a try with the same tape and see if I can tell a difference. I'm running the TBS Powercube on my builds and for the most part they fly smooth, but curious to see if this makes things even better.
Hey Joshua! Great video as always! We featured this video in our favorite drone videos of January! Thanks for making such awesome content.
Awesome! Thank you!
Very nice. Thanks for speaking on this. I soft-mounted mine and thought I could notice a difference, but also wasn't sure if it was just in my head. haha ... thanks again!
Your videos are amazing! Thank you for being so scientific in your processes. Too many multirotor channels lean hard on their opinions vs objective results. Subscribed!
Josh, what can I say.... Love you for doing this video. Finally putting all the doubters to rest. The graphs cant be denied. Great vid!
Finally a video with some objective data and proofs. Thanks Joshua.
Instead of electrical tape I'm planning in using silicone backing sheets, 1mm thickness.
Softer, cheap, doesen't leave a sticky mess and is high-temperature resistent.
Also, I think is better using only 2 screws for motor (A' la FGA or MrSteele), it will give more wiggle room.
Edit: [On a side note: If you find the perfect silicone backing sheets already in your kitchen, better ignore them and order some new one from Amazon. Your significant other may disapprove.]
Great explanation. When you put it in terms of surface area contact it makes scene. Also, Data does not lie. Thanks,
Thanks for this vid. I like the fact that you take the time to show us these theories and whether they work or don't work, rather than random pilot on the internet saying "it works" or "it's b.s." Without any real evidence.
Joshua, you are, and continue to be a tremendous asset to the hobby. Thank you for these investigations you do.
Finally! Somebody did a proper video on this! Gonna have to give it a try on my quad as well now. Might not see any change right now, but as my motors wear out and get noisy, this should help keep it flying the same.
THANK YOU! Finally, someone with a voice mentioned the surface area factor. Now, back to our regularly scheduled arguing on the Internet.
Thanks for going through the trouble to test this theory. I really like your channel, keep up the great work.
Thanks for the info. heading down to the table to soft mount the motors.
Great job!! I was dubious as well. I am going to try orings around each screw under motor.
LoL at the beginning.
I have similar shakes at around mid throttle, but the gyro traces are really clean. Haven't been able to tune it out and have been at it for months. It was buttery smooth with my earlier setup 20A little bees and 2204 Cobras. After switching to Xrotor 20A micros and Emax RS 2205 the shakes started.
Willard, Johnny and Kevin all had vids about the same time on this - good flight videos after doing it but no empirical data.
Again, thanks for performing the test and publishing Joshua, this is exactly why vlogging _ISN"T_ killing the hobby but instead educating and making everyone a better pilot.
Vlogging with information is fine. Just because a guy is on camera talking doesn't make it vlogging. I am personally bored by vlogging about vapid stuff, but to each their own.
agreed, everyone has their own reasons for watching what they do watch
I've been curious what the black box data looks like with soft mounting. Thanks for doing this experiment!
I asked you the other day about soft mounting motors . . . I got the rubber one's off amazon. . . . Currently not using them because I'm one of the people who aren't really having that problem. . . And they are a HUGE pain to get on correctly . . Then you run the risk of motors flopping off . . . And were talking about super glue . .. my quad flies fine . . Thanks for the advice !
great work Joshua, really awesome way to provide tangible evidence using BB. Love it!
Thanks Joshua! I have a friend who soft mounted his motors. His hd footage before showed blurry leaves on trees but afterward the leaves were crisp. Thanks for the black box backing to support his experience!
That's a benefit I don't see mentioned much in the discussions of soft mounting: less jello in HD camera footage.
Thanks, Joshua, this is very compelling data. You are the Galileo to all the Aristotelian pundits out there.
Dan Milliron that's why I actually started following discussions on the topic. Otherwise I might've just thought it another trending hashtag
3m 4010 clear mounting tape works great. When a similar discussion went on in the rx helicopter world it was settled(mostly) that this is the best tape to mount gyros and flybarless units. I've been using it to soft mount my motors and it works well for me. Like the grey stuff it's a bit thicker than 4 layers of electrical tape but it compresses without loosing its ability to dampen vibrations. I am curious to see how the 3D printed mounts work since they would be much cleaner and easier.
Dude. I haven't even listened to it all the way yet and the beginning is awesome.
Ok now that I have watched the whole thing, that's really telling. I suspected that we would see results like this after thinking it through a bit. It's the same reason electricians put rubber grommets on big transformers when they hang them. It dampens the vibrations and keeps the "hum" down.
Thanks Joshua. I was anxiously waiting for your test, but was thinking "Joshua wouldn't touch this topic with a ten foot pole, would he?". Anyway, great results! A couple of us did post similar controlled tests on Facebook, which confirm your observations. Most who payed any attention to it were convinced instantly, but the vast majority were laughing too hard to notice.
More great information, thanks for another worth while video! Love to see stuff that really helps.
Follow up video, TPU, foam tape, electrical tape and hard-mount shootout ! We all need to know :D
Use a slightly longer screw and "soft mount" the screw side as well then you escape the contact of 1 axis of the screw. Bore out the holes in the frame and line it with mounting tape or plastidip, then it is fully isolated. Has anyone seen a spring setup? Having thoughts of the old car cd changers that still skipped. All this makes a difference but how much? Most of us are park flyers not aerial photography pros, so is it worth it for the hobbyist? FPV racer? Would be interested in seeing fpv with the data of each in a more controlled setting ie indoors with set parameters like a warehouse loop. Thank you for making great content, super informative. You Rock!
Quality!! 😁
Very much enjoyed watching. And a I am not only talking about the purely scientific part of "measuring".
Received it as more entertaining than potentially initially intended by you.
Very good!
Great vid! Very factual AND helpful! Thanks for taking to time to do this experiment. I'm gonna give it a try!
I beat the hell out of my quad, fly PC props and don't change them often, and I don't like swapping out motors once per month. I had great results from the electrical tape trick. However, after exposing my quad to moisture and mud and abuse, the tape started to peel and trap dirt underneath. Last night I applied 5 coats of liquid electrical tape to my mounts. I'm pretty happy with its adherence to the carbon fiber, and it even coated both sides of the screw holes. I'll be giving it a test fly today. Thanks for your comprehensive workup, Joshua!
+ATA FPV some people are using TPU mounts.
Joshua Bardwell I might try some on my next build. This frame is nearing the end of its life so I wasn't worried about applying an experimental goop.
Love your videos! Thank you so much for all of your work!
+1 on the soft mounting motors. Had a frame with yaw twitches I couldn't resolve. Soft mounted the motors with electrical tape after watching Johnny's video, and everything is right with the world again. Definite +1 on the need for loctite however. Only 3 packs through the quad and everything was coming apart.
Have a go at putting silicone washers under aluminium 'cone' washers on the bolt side of the arm to supplement what you've already done.
I've done so in the past while playing around with reducing vibration for an APM board (with mixed success). It would be very interesting to see what data comes out a black box log with modern hardware.
@Joshua Bardwell: I´m working as a engineer in the semiconductor manufacturing industrie and know "a bit" of PID, gyros, dampening, dynamics of vibrations etc. Because to produce wafers you realy need a good understanding how to keep your PID controllers dealing with the slightest vibrations. What do you think why it´s possible to do lithography in some nanometers resolution to produce computer chips? :-)
I´m realy glad you did that video. After all the haters and people knowhing nothing about this just laughing and shouting "hypetrain" you did a great job showing the results. You´re right with most of your assumptions and did a realy great video.
There are some other reasons why the connection of the screwheads are not a so big deal, not only the surface area, but your conclusion is damn right. Of course soft mounting is reducing the vibrations that your PID controler has to deal with. And of course, some people will feel a bigger effect that others, this is not only because of the motors, but also about the frame design. There are even enough quads with so less vibrations, that the gyro filters will output a clean enough signal.
The most important thing beside getting rid of the vibrations is, that you indeed (in some cases) can tune your PID´s much better, resulting in higher P and lower D and a better tuned quad.
I´m personaly sick and tired of discussing with haters who do not even listen and trying to understand, just trolling around and making jokes of others. Just look at all the jokes and memes in the Facebook groups. :-(
Nice Job Joshua Bardwell, you have one subscriber more.
Just one last hint: If you look at the spectrum of the vibrations without soft mounting, you may notice that the biggest amplitude varies in frequency between differend quads. If you would like to push the soft mounting further you could do that with selecting different hard soft mounting materials, because they will dampening better on specific frequencies. This can be achived with polyurethane elastomers in different shore hardness.
There is no vodoo, this is just known engineering in some industries. :-)
Keep on with your great videos!
Thanks for testing this. I liked your intro gag too =D
I love this, watched it too many times haha.. The next big thing for 2019 I hope you guys cover and stating soon, is frame resonance and vibration and how it impacts your flight experience. I recently tried to build a 6s using an iFlight battleaxe frame. I recon if you want a prime example of a bad frame that will burn your stuff on day one, its this one. I can send you blackbox data on this as well.
Great information. Time to cast some polybushes and drill out the mounting holes so even the screws don't touch! PU is available in a range of hardness grades.
I build my micro quads from acetal because that's what I have available and I can cut it on my laser. Many people have been sceptical of acetal because it is not as stiff as carbon. However for a micro I think it is stiff enough. The fact that it is more flexible than carbon might mean less vibration is transmitted to the FC.
As always your thorough work is appreciated!
Great video, Joshua! Thank you!
Mass vs. Vibrations; if the mass of the entire frame is isolated from the motors but hard mounted to the FC we have the mass working for us. Alternatively, when the entire frame of the quad is vibrating along with 99% of the mass and then we expect the mass of the soft mounted FC ( a few grams) to stabilize on a few o-rings we get very poor results. This makes me think of back in the day when David Windestal started making his Tricopter frames, the camera plate was suspended from the flight frame and included the mass of the battery. Now off to print some flexible motor mounts!
Thanks for bringing this to light Joshua, not all of us have time to lurk in the forums or the tolerance to wade though people's BS! :-)
finally thank you. I've been saying this since day one...not all gyros/quads/motors are gonna be the same and notice the same improvement gains!
I found that drop of oil (corossionx) on noisy bearings works very well on removing micro vibrations and mid throttle oscillations
fpvYou interesting use of Corrosion-X. Do you need to re-apply bearing oil afterwards? (Like how a more refined oil will act as a solvent on a less refined oil -- I don't know how Corrosion-X compares to bearing oil like Scorpion.)
no. They state they are "World's Greatest Lubricant"
GREAT WORK, GREAT VID! Thanks for sharing!
My quad flys so much better since i softmounted! really happy with the difference!
You are a master of tuning josh! I would kill to get a glimpse of your final PID's for the wizard
Mine came with 2 really bad motors as well :(
nice with blackbox that's the great proof with results! awesome job!
Off to go try my own experiment now - easy to do on most of my quads & compare the results.
Awesome video Josh. I like your comedic additions, namely your opening and homeopathic comments!
Not to mention your clear explanation of why soft mounting isn't BS. Sadly I think most of the soft mounting haters on the rotor riot Facebook group are not the kind of people that enjoy your nuanced and scientific approach to these kinds of questions.
Much easier to just make a meme than think critically
Thanks for teaching me how to soft mount
To prevent the heads of the screws touching the arm, you can apply a spring between the head and the arm. By choosing the strength of the spring, you can also define the force of pressing the motor against the foam. Maybe a "Tellerfeder" (sorry, I don't know the English word) can also do a good job.
I read somewhere before, yes 2 years ago, I think 3 years ago either about soft mounting these motors. This was when we still fly these 450mm hex, octo, all those bunch of big multirotor.
Awesome video man. we need to get Steele to try this. he doesn't believe it will make a difference
Well that was fast lol... Great video as always.. My tape arrives today!
I've been doing this for a long time, even before it was talked about like now, the benefits became clear very quickly. Now I can even buy pads specifically for this on eBay. Place them under the motors and with the motor protector under the arm the screws even tighten onto that, not the carbon. You can heard the motors sounding cleaner in the air, to me this is a must because of the benefits I achieve.
Now the verdict has been finalized. I guess people will use torque hex key on specification material(e.g. silicon, tape, washer, etc) to fine tune the softmount. Sooner or later, there will be specify products for the softmount
Nice vid thanks josh.
PS There's a lot of noise across the whole spek analyser- especially in areas low frequency which betaflight can't filter. I had a similar problem and it ended up being ground loop by connecting both vtx grounds might be worth double checking eaachine didn't balls it up?
Excellent presentation.
From the first time I saw the soft mount videos, it made sense.
Will I do it? Probably not. I'm not that good a pilot and I don't race, yet.
All my birds are low budget quads. So I don't really think any amount of insulation from vibration will make much difference.
But for those that have spent tons of money on their gear, this should probably be something for them to try.
There will always be naysayers. That's the way the internet rolls. :)
awesome. I was too lazy to do this myself. I just tested a 1.5mm TPU pad on my motors but I need to get longer screws to feel comfortable. So ideally, the motor screw holes should be outside the base of the motor because then you can use a lock nut and then you don't need to cinch them down so hard.
8:45 Thumbs up. Great experiment, nice results. applicable to lower quality and/or damaged motors.
Impressive how some videos can create a such big thing nowaday!
Have a great weekend Joshua!
JUICY!?!?!?!? Doing this on my F80 rig right now as I wait for my PDB . . thanks!
Yo great video bro 👍
I've just changed the bearing out on my x220 wizard as yeah they Was rattley lol! now there nice and tight and run nice and smooth. I could feel the vibration from the motors by just spinning the motors by hand and definitely could hear it when flying 😆
And I've also tpu soft mounted them too and wow I can feel the difference in them just by hand through the frame with spinning them 👍 great job
I'll have to fly it tomorrow as I have a esc to replace as that crapped out on me just over night 😆 so I'll give you a update on my findings 👍
Love your videos Joshua. I wonder what the results would look like if you had done this test with electrical tape. I'm assuming the 3M tape has a better effect due to it's thickness. I want a cleaner look, so I'm planning on using the rubber o-rings that I use for the FC, good idea you think?
Hi Josh! Thanks for this deep test! May I ask what kind of GoPro mount are you using in the video? Thanks in advance.
Thanks for sharing. Very interesting. I started FPV back in 2012 when things like the QAV400 were the go to FPV quads, which back then was actually described as a small aerobatic frame, ha. Most of the popular frames around then had soft mounting built into the design. At the time I just figured they stopped including it in designs because software and filtering had improved and 250 miniquads didn't need it because of the smaller scale.....obviously not entirely true.
I think maybe the Blackout 250 miniquad (now old too) moved the trend away from soft mounting and vibration dampening. When the Blackout 250 came out people were mostly using smaller 1800 sized motors on it. I guess now that motors have gradually moved back to larger sizes with bigger magnets and more torque; vibration issues are coming to attention again.
Wonder if frames will start to move back to including vibration dampening into the design.
Yes yes yes! I backed out the screws on my motors and bye bye vibration.
Ran the same motors on a different style frame, no vibration. Frame design must just resonate at a certain frequency.
Thanks for doing this Josh! Great video
Gonna try this on one of my quads that has vibration issues this weekend
Well said Josh! if only our FPV community has our own Nobel prize :) lol. awesome and very informative post as always.
thanks for bringing evidence to this juicy question !
Thanks for clearing that up!
Soft mounted my tv, inauguration is still on.
Myth busted.
yep, it won't filter that much noise
My god, I think I just snarfed a cheese sandwich.
GAPiT FPV Don't remind me of that overgrown baby.
This is bs😂...ha...if it doesnt work for tvs then definitely wont work for quadz
As always a fantastic video. Thanks.
Really good work. Thanks.
Very nice and technical video showing some difference in soft mounting. I do agree with the mounting tape and I would even try small o-rings because it is thicker and would dampen the vibration more. My question is how tight was the first try and how tight was the tightened test? Is there an objective guage on how to tighten the screws and have a difference and not have your motor fall off? Do we tighten all the way and back off 1/16? Your videos are always very informative. Thanks
+RCDR1000 I wouldn't tighten all the way because that would crush and deform the material.
What did you do the first time you soft mounted it because that made a big difference based on your testing? Thanks
Great video Sir!
Awesome idea! Have you thought about adding a bit of padding to the screw head side when mounting? It may create even more vibration isolation as it reduces the problem you brought up in the beginning (how the quad still has contact to the frame), as well as possibly allowing you to put more torq on the screws for added safety. Maybe it will make the ride that much more "juicy"? Haha
Or use an O-ring. :D Also, some adhesive tapes have just as much holding power as those screws, potentially eliminating them all together!
@@wanderlost.productions Are you suggesting using double sided adhesive tape to hold your motors on? Good luck with that!.. :D
Thanks for this video Joshua.