How do you figure out what a good idle_min value is? LONG STORY SHORT -- take your idle%, convert it to RPM using the Motors tab, then set idle_min to 80% of that value. That is a guideline that has come out since this video was made, and it's one of the biggest questions people ask.
I know im off subject... My pitch and roll are switched on setup display in cleanflight. Tried setting board alignment off by 90 degrees but still no change. Also the elevator toggles backward. Box model rtf "RAPID" by world tech elite with an spf3
How to: Take Props off Go into Motors tab Connect a LiPo Set the "I understand the risk slider" Spin the Motors up to your idle value which is 10XX where XX is your %. So 5,5% would be 1055. Then have a look at the actual RPM above each motor slider Multiply it by 0.8 and you have your idle_min_rpm
Never call yourself "not smart" Josh! The smartest people can take complex material and explain it simply to others who do not understand. Thanks for all you do! Much love!
Worth mentioning that if you are running 4.3 Betaflight today (quote from BetaFlight): "When Dynamic Filtering is active the static idle value is disregarded, because the idle value is continually adjusted to maintain the configured minimum rpm on the slowest motor." It warns you of this now and they have a nice little field under the PID tab to adjust (alongside some other handy motor/throttle settings), you don't have to use the CLI unless you want to. They also recommend min value of 3000-3500 RPM now instead of 14000. Love the videos, these have helped me out immensely!
holy muck, your left handed and you tilt your page just like you are supposed to. i am left handed and i never learned the right way to tilt the page and i write like crap. my daughter is left handed and we are trying to make sure she tilts the page correctly just like you do. way to go man. doing it right :)
JB- if I asked a question about fc’s to the devs, I wouldn’t know what language they were responding in. You do. You are my interpreter. I thank you for having the education and experience to ask the right questions. And then the really important part, remember the answer and put an eloquent video together to let me learn something today! Esoteric at my stage of flying, but hopefully will need in the future.
Brand new to droning. I have a traxxas aton that im cutting my teeth on. I finally purchased a build kit from rotorriot.com with dji hd fpv. Take into mind i am so new that im still wet behind the ears. I enjoy your videos, they have taught me so much. Looking forward to fpv’ing!
Why is there no double like button! This is one of your best videos in ages( I'm not saying your other videos aren't awesome, I'm praising this one). This is the kind of videos that got me hooked in the first place! Thank you!
Awesome rundown JB!! If folks want to get to "zero" prop wash, tune your stuff and raise idle to 8 or 9 percent. Dynamic idle takes that improvement one (little?) step further.
'Zero' propwash AND the ability to chuck into massive rainbow-shaped inverted yaw spins? Yeah. Realistically, I only got into this because I had a pesky 7" 6S rig that would hover at 9% throttle, but when I tried racing it would desync at values below half that.
UAV Tech what do you recommend if i run dshot idle value on 5.5 percent with 2600 rpm. What should i set on idle min rpm? Is 12 ok which JB has in the Video?
Interesting as advertised! Thanks JB, these kinds of tips don't just help pilots get better, it advances the entire hobby. Shoutout to the BF devs for their work!
I feel I may be missing some of the usefulness. Take the "Vanover" case where your using a higher idle. Isn't activating dynamic idle removing some of that benefit? If the motor loses torc at low speeds, you are essentially allowing for the lower torc on those motors again. Is it that the lower torc is outweighed by the smaller range the other motors need to move? I think seeing the math would help. Thinking more, I guess it's redistributing a small amount of the work required to a slower but parallel process. In this way, I can see a nominal benefit that most people wouldn't notice and probably not worth the effort. Thanks for this interesting video!
BOOM!!!! Finally an answer, thank you Bardwell. I've been lowering idle min in order to grind and slide. Keep getting death rolls when smash technical flips, particularly with yaw. A tip! When mid death, roll you can shake it free by jamming yaw back and forth, then drawing it out and catch it.
What you describe is almost literally what airplane pilots do to get out of a "spinning fall" (not english native, not sure how to call it) I had to higher my idle speed as a beginner at first I had no clue why my kwad death rolled, I might have a weak motor but I learned a lot with JB's videos, and I didn't get a death roll today when I did yesterday, time to tune PIDs and lower filters for prop wash now :) (already didn't get that much today with rpm filtering via a beta BLHeliS release but I didn't test it that much bc I was scared of motor desync)
ESC temp display and warning are another obscure useful features, especially if you're running absolute minimum filtering. ESC and motor temperature are linearly correlated and can indicate whatever you're about to burn a motor due to excessive noise. These settings can only be accessed through CLI.
Very true, although with most of the dynamic filter and RPM filtering capability, ESCs tend to stay pretty cool, even being pushed to the point where motors are getting hot again in my experience... although I tend to encounter these conditions after I 'encounter' solid objects with my props
Do you really need that? My ESC can do 60A and the motors can't do much less. The battery, even a GNB 130C, only manages half. So nothing can smoke up.
It depends, @@BBFPV. If you're running default filter, then temp display won't help much. I run very minimal filter so the temp display is a must have because whatever I clip gates or things where my prop bends to a point where it generates noise which gets amplified in the PID loop and my motors will get real hot real quick. It's not about whatever you're going to overload one of your system, but if you're getting noise into your motors, the motors and ESCs will heat up until the enamel on the winding gets burned and you could get short circuit which may burn ESC as well. Just play it safe.
@@Asu01 Therefore you should use RPM filters, even with bl_S because then a notch filter is on the motor rpm and even with bent propellers the pid didn't amp motor/prop out of balance. I now have the filters on 1.4 and the motors stay cold. For testing its ok but every day use i don't like OSD overload.
my esc will remain stone cold even while literally burning up all 4 motors at once what you are really looking for is motor current, which you can roughly see in motor throttle in blackbox. i wish betaflight had some automatic detection of a motor sticking out in terms of thrust vs throttle because its the quickest way to spot a bad motor or prop.
Hi Joshua! As you pointed out, it is bad to let the motor speed drop too low. The motors are powerful when they spin fast, but they are surprisingly weak at starting up especially when faced with oncoming air. That is, of course, the main reason why we all use the idle speed to keep our motors from ever spinning that slow. So, here's my question... If I set my main idle speed to be just enough to prevent the low RPM problems, then why would I want to let idle_min_rpm spin the props slower under any circumstance? I may tell the quad to do a hard roll at zero throttle, and it may want to lower the speed of the motors on one side, but a hard low-throttle roll still isn't a good time to have motors stuttering because of low RPMS. So, I think it's a really interesting feature but I don't use it for myself. In your video, you claimed that idle_min_rpm might help Vanover bring down his high idle speed, but I think you got that part backwards. If you use idle_min_rpm to prevent low rpm problems, then you are free to increase the main idle speed to your desired feel. Some people might intentionally increase the main idle speed above idle_min_rpm to make it feel more floaty. Some people like Vanover might increase the main idle just to increase resolution on the throttle stick (because that lower part of the throttle would be unused in a race anyway). Before idle_min_rpm there was more of a downside to having a really high main idle speed, but now you can tune the feel of your idle separate from idle_min_rpm which is just there to set the absolute minimum RPM you're willing to use. Still an interesting feature. Just a different way to think about it. Cheers!
I didn't understand how this work and what it's used for at all from this video, nice for getting attention that this feature exists though. Would greatly suggest everyone to just read what's written in betaflight documentation about dynamic idle, it's well written and super easy to understand.
It's really not at all, and even the people in the JESC group think that the Wiki page is terrible at communicating what the feature does and how to tune it. Please click through to the FB group link in my video description if you want to read more about it.
@@JoshuaBardwell Ok, I do agree that it can be improved, but nonetheless it's quite clear what it does from reading it (not in a detailed manner (this I wish was explained somewhere, because it's quite unclear how exactly idle values are changing on what conditions), but in general). It's stated there that using bidirectional dshot rpm data some external conditions can be ignored for idle_min_rpm floor and thus motors can spin quite slowly without desync issues. And I totally agree that then it can get a bit tricky to think of its usefulness, but thinking about rotations it's quite obvious that slowing down a bit more some of the motors would help to make a certain moves more precise and efficient. Anyways, thanks for the video, wouldn't have known about this feature without it! :)
My understanding of dynamic idle was that it used the rpm data to make sure the motors never stalled whilst allowing a much lower minimum idle overall. This is probably useful for making last minute adjustments when flying as your quad should be less floaty.
This is one approach, but the smart folks I talked to seemed to think this was actually not the way to get the best results out of dynamic idle. That a quad will generally fly better with a higher idle speed, and the value of dynamic idle was allowing a higher idle speed without a net increase in idle rpm due to the PID loop needing to speed up motors.
Wow. I started out as a LOS pilot and when I switched to FPV I needed a high idle value and didn't know why. I just took my first rip with dynamic idle settings adjusted and this is how I should have been flying all along. I don't know why this isn't a hot topic in FPV conversations but Thank you Joshua!
Electric motors actually have max torque when starting from stall and the torque goes downwards as the RPM increases because magnetic field cannot be modulated fast enough due inductance. However, the current generation of brushless motors do not have required sensors to actually push the motor hard from stall so that's why you need minimum RPM for max performance.
I don't even have a quad yet but when I can afford a decent set up I will (hopefully) have a better idea as to what I'm doing thanks to your videos. Keep up the good work and beginners thank you.
Wonder when someone starts using Sensored motors on a quad. In conjunction with current control they would make all those features obsolete. The motor would start at maximum torque and no death rolls whatsoever. Yes it requires 5 signal wires, but maybe a racer seeking every little advantage?
Fabian Rudzewski race drones are supposed to be raw and best power to weight ratio but when you add complexity more things can go wrong and that’ll add extra weight slowing racers down
@@flyingmonkey8245 yet racers prefer predictability and consistency over the last bit of performance. There are even features in betaflight, to limit the motors to an empty batterys performance in order to have consistent behaviour for the whole pack. 5 signal wires and 3 Hallsensors would make a less than 10g difference for the whole quad. I'd only be worried about robustness.
Josh this vid may be answer to my dsync I had. I check mine and idle min rpm was on default 0. Definitely will give this a try to as I get it rebuilt! Thank you
Motors have weird non-linear behaviors at both extremes, it can be a bit finnicky. This approach should solve that though. I only had to bother learning this because I was getting desyncs on my 6S 7" setup, and it wants to hover at 9% throttle. Not a guarantee to fix the problem, but if you are already having issues, definitely run the slightly higher values over the calculated ones to make sure (and then test at worst case high speed split-s entries and such)
I find if your ds are too low or your ps are too high.... I was getting flip of death.. messed with the pids. Idk how that could make the difference but it did..! No more fod! Try lowering ps and see if it goes away..
Right on the spot! It was a topic with a friend while checking his configuration done by someone and his idle was set to 2. I was surprised to see such low value and asked if the person who set it knows what he was doing... Now the idle value has more clearer meaning to me, definetrly learned a big thing today. Instantly checked one of my quad and idle_min_rpm is set to 0 as default. As I understood in order to activate the dymanic idle, We have to put some value, right. I know it may seen a dumm question, yet I just need the clarification. Thanks Boss!
I'm running a raging droner with t motor f60 pro 4 1950kv. I left the stock 5.5 idle and put 20 in dynamic idle. I'm using configurator 10.7 RC1 and 4.2 RC1. It's definitely a great feature that improves flight performance. I'm a huge fan after trying it out. Thanks JB Edit. I ended up getting much better results doing proper math. It's definitely something that needs to be followed to a T for optimal results
Um at around 5:00 can it really spin around the center of mass? The pivot point should be somewhere between the right motor and the center of mass if not on the right motor imho, just because the right motor won't reverse thrust.
an object will always rotate around its center of mass, no matter where the force is applied from. if a force is applied at the center of mass, or the forces are net 0 at the center of mass, than the object wont rotate
@@DirtyLew42 I agree somewhat. It will rotate around the center of mass, but another vector is introduced which induces a change of the position of the center of mass, as it is not compensated by an equal force on the opposite side of the center of mass.
Just had a desync at zero throttle with my 3 inch build resulting in a crash. Enabling bidirectional dshot and dynamic idle sure enough fixed the issue. Also feel that enabling rpm filter and disabling first gyro filter resulted in much more responsive flight characteristics.
Always enjoy watching your vids, always learning something new , question - will this work on smaller quads for example say a tiny whoop , tiny 7 , quads with brushed motors ? keep well brother and keep teaching us...
Been using this for a while now, it's based off max rpm, rpm filters have to be enabled, and transient throttle has to be off. I use idle min rpm of 15 for 6s 2000kv.
JB, have you ever thought about exporting the audio from all your videos and posting them up as podcasts? I do a lot of driving and can't have UA-cam open but would love to follow with all you post. I know you need video to get the full experience and see some of the differences between cameras and such, but it would be great to listen to everything and control it in the car. Just a thought as you have great content. I could even listen to stuff I'm catching up on while riding a plane this way and use a variety of podcast apps.
I assume that the min idle depends a lot on the prop ? As there can be large counterforce against the motor rotation during some maneuvers. Also a full or an empty pack could probably make a difference ? and what about the BLHELI Settings for motor timing and start up power?
I think this is the issue I've been having with my beta85. I got one motor that will stall out on snappy maneuvers. I have motor stalls on one of my motors that dethrolls me out of the air as I'm coming out of a maneuver.
My first bit of tuning advice is always to up the min_throttle, I'm up to 16% on some builds! This helps out massively with my inverted yaws dropping like bricks :)
I always have the problem with yaw washouts when doing building dives with my Mobula7 and I found out that raising the idle to about 15%, where the quad is just about to lift off by it self, helps with the yaw washouts. I just recently got into dynamic idle but I'm still testing. (Yea, I'm commenting on a 4 year old Video)
Wow, what a coincidence this is exactly what I was looking for. I like to do a lot of max throttle runs, and towards the end I go to zero throttle, then quickly turn around, and then as the quad starts slowing down, it will occasionally do a half roll and fall a bit I believe because of motor stall from low voltage and slow voltage recovery? Most of the time I can quickly recover, but occasionally if I’m flying too low, it will stall, do a half roll, and crash into the ground upside down, and I think this might fix that.
The way i understood, it had more to do with the throttle percentage being able to go below your normal idle percentage. So dynamic Idle will allow you to go below your normal idle percentage of lets say 5% throttle, in cases where reverse flow conditions are already providing torque on the motor (due to reverse flow on your props). So in your example for people who have high trottle percentages, should their min_rpm values correspond to the RPM they aim at to have a lot of response (which are high). The benefit there as i understand it, would be that the PID loop is allowed to lower throttle to 0% as long as the props keep spinning at the minimum desired speed. What am i missing here?
Very informative Josh as always. I have been using this feature for quite some time now and I agree with you this is a very useful tool when used properly you can definitely get better performance as always thank you for the awesome video keep up the good work.
So I got confused by the idle min section. You went through the exercise of getting the lowest possible rpm, (300) and then it seemed like you "winged it" and put in 1200 in the CLI. Do we use 4x the number we come up with?
I didn't wing it. I came up with a value of about 300 for three motors and then 600 for one (probably damaged) motor. I flew with a value of 600 and got an arm-drop on sharp moves. So I ended up raising it until the arm-drop disappeared. 1200 is where I ended up. That's the procedure I suggest you use.
Ok I see what you saying now. Update: it wasnt the dynamic idle change I made that caused the quad to die, it was loose positive wire coming out of the xt60 plug. Dynamic idle is working great, thank you Josh also informative.
@@JoshuaBardwell So if we have to add the 4 motor idle values from the bench test (that wasn't quite clear tbh), why didn't you then come up with 15 (1500) or put it even higher as you suggested?
I have MOTOR_STOP enabled and airmode controlled by a switch. If I during flight switch airmode off and lower the throttle to 0% the motors stop but they never stall, as soon as I increase the throttle all the motors start spinning again.
In new 10.7 configurator, Dynamic Idle RPM is on PID tuning page. It is set to 0 at default. This is why the wobble is there in 4.2. Hope they fix or have it “slide” with increase in throttle percent.
So what is each motor has a different rpm in which it spins without hesitating while in the motors tab? For example, motor 1 is 350, 2 is 620, 3 is 900, and 4 is 900. Do I just take the motor with the highest rpm and use that?
So to just distill it slightly further.it helps your quad rotate around the center axis properly(or better than normal),rather than just lift one side of the quad to rotate.its the best we can do with the design of quads at present.ideally you would have a second motor&prop on each arm opposite mounted to the motor on top of the arm.so as one side spins up,the motor on the opposite arm facing down,would pull down.therefor having an equal push to pull&properly pushing the quad around its center point.but then youd pay a weight penalty&drag.so this dynamic idle&air mode etc. Is doing the best it can with the hard/&software configuration.theres been many incremental benefits added to bflight over the years.many of which we might have thought at the time were just adding complexity.which we now consider the norm&wouldn't be without.im just wondering when/if these incremental tweaks are gonna result in diminishing returns.although I do agree with this one as d syncs & motors being able to respond and spool up quicker from a certain point of idle.is allowing for spool up lag&the lag that the pid controller/rc system as a whole introduces to be partially negated.
Great informative vid as always! Would you have a link to a complete setup video for all these new features in the new betaflight? Or are they spread out.👌
10:42 is it posible that this minimum depends from battery voltage? lets say cel voltage is 4.2 and i set minimum idle, and batery drops to 3.9 can my motor stal from this?
Yes, it very much does depend if you're changing cell count without this. Across battery discharge, the old way would have fairly minor changes, but this does RPM... and it's usually best to run VBAT PID Compensation anyway (technially unrelated, but it does help with the same problems). While it sounds odd, this feature removes those changes based on input cell voltage because it's using RPM data, instead of just blinding sending a DSHOT command and the RPM you get is what you get.
I usually use a 6.5 to 8.5 min idle depending how new or old my motors are. Reason being I like heavy props and hate de-sinks, but inverted I fall like a brick🤣 This could help, and reckon this is worth a go for sure. Thanks...
When you took the base value of the stall speed of the motor by slowing it down to see the min value, you did this without a propeller on. If motor stall speed is a function of resistance like you said, you would want to have a propeller on during this to get the right base value...correct?
It kind of doesn't matter because the number will always be much higher in the air, but I thought it was a good exercise to get the lowest possible value.
Why would some people run a higher idle rather than just using air mode? I guess I don't understand is disadvantage of air mode, other than the inversion time like you mentioned.
I thought motors had more toque at lower speeds... at least on my CNC machine with stepper (brushless) motors. Are quads different or is it something a bit different (time to spin up from very low rpm). I thought the desync was the command to push not being in sync with the motor (magnet) position due to the lack of torque from the last command. This being motor stall, not areodynamic stall. Curious if I've got this wrong. None of which changes the usefulness of dynamic idle and your awesome video (as always).
I'm a jackass... Had I watched 5 seconds more before posting I would have heard "motor stall", "motor stall", "motor stall". But I did catch the problem with the bench test... You need a lot more torque with motors attached. That was the "no load" stall speed.
It's not truly an issue of torque, but that's a short-hand way of expressing what's happening. One thing is that the position sensing of the ESC becomes less accurate at low rpm, because the back-EMF voltage is weaker. Therefore the chances of a stall are greater. Another thing is that when the prop is spinning slowly, the motor makes less thrust and therefore the PID loop has less authority.
Always learning from you Joshua. Just a clarification, in BF motor control do you consider the motor stalled when it stops rotating or when it starts to stutter? Yes, I'll watch the video again.
Hi Joshua, I've got a problem with my Protek25 and my Betafpv 95 V2, when I fly between tries, and heppened I have to cut throttle to pass throw sometimes motor idel same speed for few seconds, I've tried to rise and again slow the rpm using the stick but nothing it slow down after few seconds and I crash on the tree.... It seems it happened when there is wind but when I fly them not beetween tries in the wind it doesn't heppened. Could it be a dynamic Idle adjustment to do? Many thanks
I'm going to pose a question and see what your answer is. Why don't we use a sensor based esc in FPV? My background is heavy in surface and in racing in particular they use a sensor based esc. There is a cable that runs from the motor to the esc that tells the esc where the rotor is in its rotation allowing the esc to make more precise drive inputs giving the driver a better "feel". This is mainly for lower rpm and usually is turned off at higher rpm as the sensor actually hinders hi rpm performance. I see where this could be beneficial, especially for dynamic idle.
So, pole crossings are very specifically how the ESC knows to initiate the next switching event. What your'e describing is core to how an electronic speed controller knows when to dump DC current into the motor windings, and switch it back off with the next one coming. Getting this data that's already onboard the ESC (this is why Motor_Poles is a parameter) back to a flight controller via bidirectional DSHOT is the magic that enables RPM Filtering and Dynamic Idle.
@@xNmMmAx No problem - this does a much better job of explaining it - the back-EMF sensing is integratl to how it performs (and why some ESC settings like motor timing to med-high/23° affects performance so much) howtomechatronics.com/how-it-works/how-brushless-motor-and-esc-work
So why would we ever want the pid controller to have access to the min_idle. If the quad flies so much better with higher percentage and if response times are slower when coming out of that super low rpm then aren't you recommending everyone runs higher idle and to not run min_idle? In what scenario would the PID controller getting to use min_idle actually be beneficial?
@@MewoFPV You should have more authority lower down. As assuming an ideal case you will have to spin the motor up only half as much as you now have 4 forces to play with not just 2. Might not be the case in the realworld though, maybe it takes the same time to spool them up more, or braking is to slow this low in their speed curve.
@@MewoFPV MewoFPV dive gates? A reduced hang time will result in less control? In a competitive racing environment you would probs be flying with consistantly enough power that PID will never access min throttle anyways. However by increasing the min idle, when operating at the bottom of the power band, you effectively shift the center of rotation in the direction of the turn. An example of where this might be a problem is In prop wash, as the center of rotation Is moved rapidly any additional rotation command by the pilot will result in an ossilating track through the sky. Yes it becomes more rotationally more stable but for racing having a consistant track may be valued more than a consistent than attitude.
A while ago , I lowered the idle speed down as far as I could , on some kwads as low as 1.5 % and stopped using air mode . But will be interested in trying this out .
Hey man. I just hopped into fpv and now I'm here cause I was wondering about dynamic idle. In 4.5.1 , I just set the value for dynamic idle in the pid tab and that's it? motor idle (static) is greyed out in the motors tab after setting a value for dynamic idle in the pid tab. Or do I have to do the CLI thing too? Thx for all the time you've been putting into these videos. They help so much!! Cheers m8
hey joshua i have a pavo25 and when punch out and drop the throttle the motors start twitching when recover the altitude and latter have this same twitches in fly! any suggestion? you are help a lot over this fpv world!
The point is that with dynamic idle you can raise that so your quad flies better at zero throttle, without giving up the ability of the motors to go below that point if they need to.
@@JoshuaBardwell Hi never flown just got mobula7 hd can't set it up in betaflight just about got it binded I'm almost giving up on it if you could give us some info many thanks 👍
Hi Joshua, can you please confirm on the updated firmware of DJI's goggles? I want to get goggles after the lockdown (no shipping to SA at the moment), and was almost going for the newest Fatshark's because they display all info on the OSD which is important for me. Now I've read in a comment on another video that DJI brought out a big update that adds all this OSD info to their goggles, which was holding me back from getting them..
Im a freestyle pilot and usualy up my idle speed a bit more than default. I found it helps on propwash. Thanks for the explanation... going to try in the nexr flight.
I noticed that, at least for my setup, the RPM value in BF is 50% that of the RPM value in BLHeli motors tab, for the same PWM. Any idea why? Basically BLHeli only lets me step from 1015 to 1017 and reflects 850RPM while BF shows 450 RPM for what appears to be the same PWM value.
Thank you JB, On top of that, what is the behavior if you change value in BLheli like Throttle Min into a lower or upper value than the default in ESC firmware configuration ?
I have a Mobula 8 and the Dynamic Idle is greyed out in Betaflight (i have firmware 4.5 with configurator 10.10.0), does this mean that Dynamic Idle isn't supported on that whoop?
Dammit Bardell, you're so smart, you're answering the question about deathrolls that I didn't go online and ask you about this weekend. The same quad, deathrolled twice lately, while descending into the next field (low throttle) where one motor on my (plus) frame was probably at idle. Without even checking, I bet my idle % is too low and you're a mind reading genius!
Thank you for the briefing, Joshua! Have you gathered any feedback/opinion from Captain V.? Also, what is/are the parameters that cause the idle value to vary between the two endpoint settings? For example, if idle is only adjusted downward in response to differential prop rate (e.g., rolls or flips), then inverted hang time would still be lost if we raise our dshot_idle_value, right?
That is a question that I am still trying to understand but the answer seems to be no. That right now, your motors are not spinning at idle when you are at zero throttle, because the PID loop must increase throttle to stabilize the quad. So if you enable dynamic idle, the actual rpm at zero throttle will be lower, and you can actually raise idle percent with no net change in hang time or speed of descent. Obviously at a certain point, the idle is so high that you feel a difference.
@@JoshuaBardwell Reading TFM (github.com/betaflight/betaflight/wiki/Tuning-Dynamic-Idle) helped me to understand your explanation ;-) It is still working against inverted hang time to raise our dshot_idle_value. Enabling dynamic idle allows a motor's drive to go to zero as long as its prop is still spinning at or above idle_min_rpm. The PID controller can thus command reduced drive (i.e., below the dshot_idle_value) on motors that need braking to maintain orientation (e.g., when inverted). The reduced net drive among all motors while inverted increases hang time. We can then raise our dshot_idle_value to consume as much of the the newly-acquired margin as we like.
I am in firmware 4.2 and using 10.7 beta flight configurator. I tried changing idle min both thru CLI and in the tuning tab and neither approach is saving. It’s currently set to zero and I want to change it to 14. What am I doing wrong ??? I have used CLI before to make changes and it’s always worked for other features. Please help !!! Thanks for the great videos
I Run a 6 inch 6s build with brotherhobby avenger v2 2507 1200kv .. I find at low throttle I wobble just a tiny bit if in coming down into dirty air.. would bumping up the idle percent make more sense in my application rather then dynamic ?
That's extremely interesting!!! I had never imagined that CaptainVani had 8% idle! Personally i had it down to 3% because 1) i find difficult to do inverted yaw spins because tends to fall faster 2) the quad takes forever to get down to the ground. I definitely learn something today. Question: Propellers pitch affect the idle ratio or dynamic idle?
Prop weight, pitch, and relative chord thickness all affect it, usually in that order. Heavy aggressive props will have you wanting to run higher values, thinner lighter ones with less pitch let you run lower values.
No no no Completely overwhelmed and caught in the ceiling fan can we try this again. OK I just need to re watch this and then review the FB post then watch it again I guess
small correction: For BLDC's you shouldn't lose torque at lower speeds, torque is dependent entirely on magnetic flux in the core, which is governed by current. The motor certainly cannot _maintain_ high torque at low speed because inadequate cooling. I suspect this is more of a fundamental problem with sensorless control of said motors. desync happens because the ESC losses track of where the rotor is relative to the stator and the motor losses all torque.
There is more to the torque claim, you're right. The desync is because at low speed, the BEMF that is used to detect pole crossings is very weak and is lost in the noise. The lack of authority is in part because when the prop is moving slowly, it isn't making much thrust. But from what I am told by people developing the feature, the motor is less responsive coming from low rpm, which I approximate as "lack of torque" when explaining it. The motor will take longer to get from 2% throttle to 5% than from 50% to 53% for example. Once you get out of the low-rpm region, this problem goes away.
@@JoshuaBardwell Its interesting observation noneheless! No doubt that the practical wins out over this theory. I think I read somewhere that the sensorless ESCs often switch between algorithms at low speeds for this reason. Keep up the great work!
Anyone know if there's a way to override betaflight's max idle % to go beyond 20%? I'm doing some cinewhoop experimenting and this would be very helpful. Thanks.
How do you figure out what a good idle_min value is? LONG STORY SHORT -- take your idle%, convert it to RPM using the Motors tab, then set idle_min to 80% of that value. That is a guideline that has come out since this video was made, and it's one of the biggest questions people ask.
So for clarification, if your dshot idle calculates to 2200rpm then 80% would 1660rpm and that is what you would set you idle_min_rpm?
I know im off subject... My pitch and roll are switched on setup display in cleanflight. Tried setting board alignment off by 90 degrees but still no change. Also the elevator toggles backward. Box model rtf "RAPID" by world tech elite with an spf3
@@trixshop5188 Im pretty sure JB has a video about that - this might be helpful..... ua-cam.com/video/1azl1f8qJ84/v-deo.html
How to:
Take Props off
Go into Motors tab
Connect a LiPo
Set the "I understand the risk slider"
Spin the Motors up to your idle value which is 10XX where XX is your %.
So 5,5% would be 1055.
Then have a look at the actual RPM above each motor slider
Multiply it by 0.8
and you have your idle_min_rpm
@@dominikthone512 if dshot idle value is 5.5 shouldn’t that be 5.5% of 2000? I.e 1110?
Never call yourself "not smart" Josh! The smartest people can take complex material and explain it simply to others who do not understand. Thanks for all you do! Much love!
Joshua, you are by definitions a true intellectual , thats what u do, make complex subjects understandable to the average person. Thanks much!
Try to understand everything when you're French with a small English 🤯🤯🤯.
I WILL LEAR SOMETHING TODAYYYY 😂🤣
@@enriquemaillard-mora9193 when u understand more than one language your pretty smart too.🌝
Worth mentioning that if you are running 4.3 Betaflight today (quote from BetaFlight): "When Dynamic Filtering is active the static idle value is disregarded, because the idle value is continually adjusted to maintain the configured minimum rpm on the slowest motor." It warns you of this now and they have a nice little field under the PID tab to adjust (alongside some other handy motor/throttle settings), you don't have to use the CLI unless you want to. They also recommend min value of 3000-3500 RPM now instead of 14000. Love the videos, these have helped me out immensely!
Yup just enabled rpm filtering and it enabled dynamic idle automatically, and it fixed my desync at 0 throttle issue.
Cheers. The quick video was the least I could do for how many times I’ve used your tutorials.
Right? Same boat, I'm still using them now anytime my first braille method approach to something doesn't work out the first time.
Such a nice and humble guy, not taking all the credit for other people’s hard work but also helping the fpv community to understand these features
holy muck, your left handed and you tilt your page just like you are supposed to. i am left handed and i never learned the right way to tilt the page and i write like crap. my daughter is left handed and we are trying to make sure she tilts the page correctly just like you do. way to go man. doing it right :)
JB- if I asked a question about fc’s to the devs, I wouldn’t know what language they were responding in. You do. You are my interpreter. I thank you for having the education and experience to ask the right questions. And then the really important part, remember the answer and put an eloquent video together to let me learn something today! Esoteric at my stage of flying, but hopefully will need in the future.
Brand new to droning. I have a traxxas aton that im cutting my teeth on. I finally purchased a build kit from rotorriot.com with dji hd fpv. Take into mind i am so new that im still wet behind the ears. I enjoy your videos, they have taught me so much. Looking forward to fpv’ing!
Why is there no double like button! This is one of your best videos in ages( I'm not saying your other videos aren't awesome, I'm praising this one). This is the kind of videos that got me hooked in the first place! Thank you!
Awesome rundown JB!!
If folks want to get to "zero" prop wash, tune your stuff and raise idle to 8 or 9 percent. Dynamic idle takes that improvement one (little?) step further.
'Zero' propwash AND the ability to chuck into massive rainbow-shaped inverted yaw spins? Yeah.
Realistically, I only got into this because I had a pesky 7" 6S rig that would hover at 9% throttle, but when I tried racing it would desync at values below half that.
I was wondering about that! thx
UAV Tech what do you recommend if i run dshot idle value on 5.5 percent with 2600 rpm. What should i set on idle min rpm? Is 12 ok which JB has in the Video?
9% idle will kill your inverted hang time but help your propwash. Does it really take that much idle to fix your propwash?
@@user-rs8zg8ey2b , prop was is never "fixed".
Interesting as advertised! Thanks JB, these kinds of tips don't just help pilots get better, it advances the entire hobby. Shoutout to the BF devs for their work!
I feel I may be missing some of the usefulness. Take the "Vanover" case where your using a higher idle. Isn't activating dynamic idle removing some of that benefit? If the motor loses torc at low speeds, you are essentially allowing for the lower torc on those motors again. Is it that the lower torc is outweighed by the smaller range the other motors need to move? I think seeing the math would help. Thinking more, I guess it's redistributing a small amount of the work required to a slower but parallel process. In this way, I can see a nominal benefit that most people wouldn't notice and probably not worth the effort. Thanks for this interesting video!
Thanks!
BOOM!!!!
Finally an answer, thank you Bardwell.
I've been lowering idle min in order to grind and slide. Keep getting death rolls when smash technical flips, particularly with yaw.
A tip! When mid death, roll you can shake it free by jamming yaw back and forth, then drawing it out and catch it.
What you describe is almost literally what airplane pilots do to get out of a "spinning fall" (not english native, not sure how to call it)
I had to higher my idle speed as a beginner at first I had no clue why my kwad death rolled, I might have a weak motor but I learned a lot with JB's videos, and I didn't get a death roll today when I did yesterday, time to tune PIDs and lower filters for prop wash now :) (already didn't get that much today with rpm filtering via a beta BLHeliS release but I didn't test it that much bc I was scared of motor desync)
ESC temp display and warning are another obscure useful features, especially if you're running absolute minimum filtering. ESC and motor temperature are linearly correlated and can indicate whatever you're about to burn a motor due to excessive noise. These settings can only be accessed through CLI.
Very true, although with most of the dynamic filter and RPM filtering capability, ESCs tend to stay pretty cool, even being pushed to the point where motors are getting hot again in my experience... although I tend to encounter these conditions after I 'encounter' solid objects with my props
Do you really need that? My ESC can do 60A and the motors can't do much less. The battery, even a GNB 130C, only manages half. So nothing can smoke up.
It depends, @@BBFPV. If you're running default filter, then temp display won't help much. I run very minimal filter so the temp display is a must have because whatever I clip gates or things where my prop bends to a point where it generates noise which gets amplified in the PID loop and my motors will get real hot real quick.
It's not about whatever you're going to overload one of your system, but if you're getting noise into your motors, the motors and ESCs will heat up until the enamel on the winding gets burned and you could get short circuit which may burn ESC as well. Just play it safe.
@@Asu01 Therefore you should use RPM filters, even with bl_S because then a notch filter is on the motor rpm and even with bent propellers the pid didn't amp motor/prop out of balance. I now have the filters on 1.4 and the motors stay cold.
For testing its ok but every day use i don't like OSD overload.
my esc will remain stone cold even while literally burning up all 4 motors at once
what you are really looking for is motor current, which you can roughly see in motor throttle in blackbox. i wish betaflight had some automatic detection of a motor sticking out in terms of thrust vs throttle because its the quickest way to spot a bad motor or prop.
Thanks Josh from Australia! This is why my heavy 7 inch 6S always flys so unpredictable 😁
Hi Joshua! As you pointed out, it is bad to let the motor speed drop too low. The motors are powerful when they spin fast, but they are surprisingly weak at starting up especially when faced with oncoming air. That is, of course, the main reason why we all use the idle speed to keep our motors from ever spinning that slow. So, here's my question... If I set my main idle speed to be just enough to prevent the low RPM problems, then why would I want to let idle_min_rpm spin the props slower under any circumstance? I may tell the quad to do a hard roll at zero throttle, and it may want to lower the speed of the motors on one side, but a hard low-throttle roll still isn't a good time to have motors stuttering because of low RPMS.
So, I think it's a really interesting feature but I don't use it for myself. In your video, you claimed that idle_min_rpm might help Vanover bring down his high idle speed, but I think you got that part backwards. If you use idle_min_rpm to prevent low rpm problems, then you are free to increase the main idle speed to your desired feel. Some people might intentionally increase the main idle speed above idle_min_rpm to make it feel more floaty. Some people like Vanover might increase the main idle just to increase resolution on the throttle stick (because that lower part of the throttle would be unused in a race anyway). Before idle_min_rpm there was more of a downside to having a really high main idle speed, but now you can tune the feel of your idle separate from idle_min_rpm which is just there to set the absolute minimum RPM you're willing to use. Still an interesting feature. Just a different way to think about it. Cheers!
I didn't understand how this work and what it's used for at all from this video, nice for getting attention that this feature exists though. Would greatly suggest everyone to just read what's written in betaflight documentation about dynamic idle, it's well written and super easy to understand.
It's really not at all, and even the people in the JESC group think that the Wiki page is terrible at communicating what the feature does and how to tune it. Please click through to the FB group link in my video description if you want to read more about it.
@@JoshuaBardwell Ok, I do agree that it can be improved, but nonetheless it's quite clear what it does from reading it (not in a detailed manner (this I wish was explained somewhere, because it's quite unclear how exactly idle values are changing on what conditions), but in general). It's stated there that using bidirectional dshot rpm data some external conditions can be ignored for idle_min_rpm floor and thus motors can spin quite slowly without desync issues. And I totally agree that then it can get a bit tricky to think of its usefulness, but thinking about rotations it's quite obvious that slowing down a bit more some of the motors would help to make a certain moves more precise and efficient. Anyways, thanks for the video, wouldn't have known about this feature without it! :)
Dam you Bardwell, that’s another 17mins off my nightly sleep quota!
As always good info and kind of got me thinking about how to get more hang time…..
My understanding of dynamic idle was that it used the rpm data to make sure the motors never stalled whilst allowing a much lower minimum idle overall.
This is probably useful for making last minute adjustments when flying as your quad should be less floaty.
This is one approach, but the smart folks I talked to seemed to think this was actually not the way to get the best results out of dynamic idle. That a quad will generally fly better with a higher idle speed, and the value of dynamic idle was allowing a higher idle speed without a net increase in idle rpm due to the PID loop needing to speed up motors.
Wow. I started out as a LOS pilot and when I switched to FPV I needed a high idle value and didn't know why. I just took my first rip with dynamic idle settings adjusted and this is how I should have been flying all along. I don't know why this isn't a hot topic in FPV conversations but Thank you Joshua!
You may not be the smartest guy, but you're a really freaking good teacher.
I'm not the smartest guy, or the funniest, but I'm also not the best looking.
Thanks Josh! I would have loved to see a before and after but I appreciate the video none the less!
Electric motors actually have max torque when starting from stall and the torque goes downwards as the RPM increases because magnetic field cannot be modulated fast enough due inductance. However, the current generation of brushless motors do not have required sensors to actually push the motor hard from stall so that's why you need minimum RPM for max performance.
Bless you for putting that time stamp in for everyone who cant sit still for more than 7 seconds....
I don't even have a quad yet but when I can afford a decent set up I will (hopefully) have a better idea as to what I'm doing thanks to your videos. Keep up the good work and beginners thank you.
Wonder when someone starts using Sensored motors on a quad.
In conjunction with current control they would make all those features obsolete. The motor would start at maximum torque and no death rolls whatsoever.
Yes it requires 5 signal wires, but maybe a racer seeking every little advantage?
Fabian Rudzewski race drones are supposed to be raw and best power to weight ratio but when you add complexity more things can go wrong and that’ll add extra weight slowing racers down
@@flyingmonkey8245 yet racers prefer predictability and consistency over the last bit of performance.
There are even features in betaflight, to limit the motors to an empty batterys performance in order to have consistent behaviour for the whole pack.
5 signal wires and 3 Hallsensors would make a less than 10g difference for the whole quad.
I'd only be worried about robustness.
Great video! You would be a great teacher/professor! You are great at explaining everything. Thanks for all you do!
Now we have BetaFlight 4.5 just released 😊
Josh this vid may be answer to my dsync I had. I check mine and idle min rpm was on default 0. Definitely will give this a try to as I get it rebuilt! Thank you
I'm Learning...
I learned that maybe my de-syncs are from setting my idle percentage too low... never dawned one me
Yeah I thinks that's what's been happening to me
Me too...
Motors have weird non-linear behaviors at both extremes, it can be a bit finnicky. This approach should solve that though.
I only had to bother learning this because I was getting desyncs on my 6S 7" setup, and it wants to hover at 9% throttle.
Not a guarantee to fix the problem, but if you are already having issues, definitely run the slightly higher values over the calculated ones to make sure (and then test at worst case high speed split-s entries and such)
same here^^
What a relief, I was really down because I couldn't really find the reason. This is a new thing to look into :)
Thanks JB!
I find if your ds are too low or your ps are too high.... I was getting flip of death.. messed with the pids. Idk how that could make the difference but it did..! No more fod! Try lowering ps and see if it goes away..
Right on the spot! It was a topic with a friend while checking his configuration done by someone and his idle was set to 2. I was surprised to see such low value and asked if the person who set it knows what he was doing... Now the idle value has more clearer meaning to me, definetrly learned a big thing today. Instantly checked one of my quad and idle_min_rpm is set to 0 as default. As I understood in order to activate the dymanic idle, We have to put some value, right. I know it may seen a dumm question, yet I just need the clarification. Thanks Boss!
I'm running a raging droner with t motor f60 pro 4 1950kv. I left the stock 5.5 idle and put 20 in dynamic idle. I'm using configurator 10.7 RC1 and 4.2 RC1. It's definitely a great feature that improves flight performance. I'm a huge fan after trying it out. Thanks JB
Edit. I ended up getting much better results doing proper math. It's definitely something that needs to be followed to a T for optimal results
Um at around 5:00 can it really spin around the center of mass? The pivot point should be somewhere between the right motor and the center of mass if not on the right motor imho, just because the right motor won't reverse thrust.
an object will always rotate around its center of mass, no matter where the force is applied from. if a force is applied at the center of mass, or the forces are net 0 at the center of mass, than the object wont rotate
@@DirtyLew42 I agree somewhat. It will rotate around the center of mass, but another vector is introduced which induces a change of the position of the center of mass, as it is not compensated by an equal force on the opposite side of the center of mass.
HoverFly incorrect. Center of mass will never change, unless mass is physically moved. Please do some research of your own.
Just had a desync at zero throttle with my 3 inch build resulting in a crash. Enabling bidirectional dshot and dynamic idle sure enough fixed the issue. Also feel that enabling rpm filter and disabling first gyro filter resulted in much more responsive flight characteristics.
Always enjoy watching your vids, always learning something new , question - will this work on smaller quads for example say a tiny whoop , tiny 7 , quads with brushed motors ? keep well brother and keep teaching us...
I expect it would work well on any quad, but quads like micros, where the idle speed is often much higher, may particularly benefit.
Hey man, just wanted to say I enjoy your content as someone who is looking to get into drones!
Been using this for a while now, it's based off max rpm, rpm filters have to be enabled, and transient throttle has to be off. I use idle min rpm of 15 for 6s 2000kv.
JB, have you ever thought about exporting the audio from all your videos and posting them up as podcasts? I do a lot of driving and can't have UA-cam open but would love to follow with all you post. I know you need video to get the full experience and see some of the differences between cameras and such, but it would be great to listen to everything and control it in the car. Just a thought as you have great content. I could even listen to stuff I'm catching up on while riding a plane this way and use a variety of podcast apps.
I assume that the min idle depends a lot on the prop ? As there can be large counterforce against the motor rotation during some maneuvers. Also a full or an empty pack could probably make a difference ? and what about the BLHELI Settings for motor timing and start up power?
I think this is the issue I've been having with my beta85. I got one motor that will stall out on snappy maneuvers. I have motor stalls on one of my motors that dethrolls me out of the air as I'm coming out of a maneuver.
the same
My first bit of tuning advice is always to up the min_throttle, I'm up to 16% on some builds! This helps out massively with my inverted yaws dropping like bricks :)
magicalpencil I don’t understand. When you’re inverted and motors are spinning faster you should fall faster, not slower...
@@prottentogo Yea, makes zero sense
Thanks, Josh Bardwell!
I always have the problem with yaw washouts when doing building dives with my Mobula7 and I found out that raising the idle to about 15%, where the quad is just about to lift off by it self, helps with the yaw washouts. I just recently got into dynamic idle but I'm still testing. (Yea, I'm commenting on a 4 year old Video)
Wow, what a coincidence this is exactly what I was looking for. I like to do a lot of max throttle runs, and towards the end I go to zero throttle, then quickly turn around, and then as the quad starts slowing down, it will occasionally do a half roll and fall a bit I believe because of motor stall from low voltage and slow voltage recovery? Most of the time I can quickly recover, but occasionally if I’m flying too low, it will stall, do a half roll, and crash into the ground upside down, and I think this might fix that.
The way i understood, it had more to do with the throttle percentage being able to go below your normal idle percentage. So dynamic Idle will allow you to go below your normal idle percentage of lets say 5% throttle, in cases where reverse flow conditions are already providing torque on the motor (due to reverse flow on your props). So in your example for people who have high trottle percentages, should their min_rpm values correspond to the RPM they aim at to have a lot of response (which are high). The benefit there as i understand it, would be that the PID loop is allowed to lower throttle to 0% as long as the props keep spinning at the minimum desired speed. What am i missing here?
A mounted prop will affect stall, a lot more weight to get/keep rotating, so higher torque via rpm is needed.....as you found out😎
Very informative Josh as always. I have been using this feature for quite some time now and I agree with you this is a very useful tool when used properly you can definitely get better performance as always thank you for the awesome video keep up the good work.
So I got confused by the idle min section. You went through the exercise of getting the lowest possible rpm, (300) and then it seemed like you "winged it" and put in 1200 in the CLI. Do we use 4x the number we come up with?
yea im confused also i just went through this whole thing and now my quad just dies at the lightest blip of a roll.
I didn't wing it. I came up with a value of about 300 for three motors and then 600 for one (probably damaged) motor. I flew with a value of 600 and got an arm-drop on sharp moves. So I ended up raising it until the arm-drop disappeared. 1200 is where I ended up. That's the procedure I suggest you use.
Ok I see what you saying now.
Update: it wasnt the dynamic idle change I made that caused the quad to die, it was loose positive wire coming out of the xt60 plug. Dynamic idle is working great, thank you Josh also informative.
@@JoshuaBardwell So if we have to add the 4 motor idle values from the bench test (that wasn't quite clear tbh), why didn't you then come up with 15 (1500) or put it even higher as you suggested?
I have MOTOR_STOP enabled and airmode controlled by a switch. If I during flight switch airmode off and lower the throttle to 0% the motors stop but they never stall, as soon as I increase the throttle all the motors start spinning again.
Thanks a lot no one is talking about this, excellent explanation.
Man!!! Great informative video once again. You break it down so nicely always!! Big thx
I wonder what sensored brushless esc and motors would do for flying qualities?
We'll find out soon, maybe later this year.
In new 10.7 configurator, Dynamic Idle RPM is on PID tuning page. It is set to 0 at default. This is why the wobble is there in 4.2. Hope they fix or have it “slide” with increase in throttle percent.
So what is each motor has a different rpm in which it spins without hesitating while in the motors tab? For example, motor 1 is 350, 2 is 620, 3 is 900, and 4 is 900. Do I just take the motor with the highest rpm and use that?
So to just distill it slightly further.it helps your quad rotate around the center axis properly(or better than normal),rather than just lift one side of the quad to rotate.its the best we can do with the design of quads at present.ideally you would have a second motor&prop on each arm opposite mounted to the motor on top of the arm.so as one side spins up,the motor on the opposite arm facing down,would pull down.therefor having an equal push to pull&properly pushing the quad around its center point.but then youd pay a weight penalty&drag.so this dynamic idle&air mode etc. Is doing the best it can with the hard/&software configuration.theres been many incremental benefits added to bflight over the years.many of which we might have thought at the time were just adding complexity.which we now consider the norm&wouldn't be without.im just wondering when/if these incremental tweaks are gonna result in diminishing returns.although I do agree with this one as d syncs & motors being able to respond and spool up quicker from a certain point of idle.is allowing for spool up lag&the lag that the pid controller/rc system as a whole introduces to be partially negated.
Great informative vid as always! Would you have a link to a complete setup video for all these new features in the new betaflight? Or are they spread out.👌
Here is my 4.1 playlist: ua-cam.com/play/PLwoDb7WF6c8kH_vLskPnY5xjuT284rhmj.html
@@JoshuaBardwell Thanks much appreciated.👍
15:16 LoL.... JB, searched for hours but I couldn't find "Nydamic" anywhere in the Betaflight CLI.....
10:42 is it posible that this minimum depends from battery voltage? lets say cel voltage is 4.2 and i set minimum idle, and batery drops to 3.9 can my motor stal from this?
Probably not because idle min appears to be based on motor RPM not percent throttle
@@Art1f4ct this have sense, my min was 0 and idle 300 I think I'll make some changes :)
I tested with empty and full battery and found the stall rpm to be the same. I think this is why idle_min is an rpm and not a percent.
Yes, it very much does depend if you're changing cell count without this. Across battery discharge, the old way would have fairly minor changes, but this does RPM... and it's usually best to run VBAT PID Compensation anyway (technially unrelated, but it does help with the same problems).
While it sounds odd, this feature removes those changes based on input cell voltage because it's using RPM data, instead of just blinding sending a DSHOT command and the RPM you get is what you get.
I usually use a 6.5 to 8.5 min idle depending how new or old my motors are. Reason being I like heavy props and hate de-sinks, but inverted I fall like a brick🤣
This could help, and reckon this is worth a go for sure.
Thanks...
When you took the base value of the stall speed of the motor by slowing it down to see the min value, you did this without a propeller on. If motor stall speed is a function of resistance like you said, you would want to have a propeller on during this to get the right base value...correct?
It kind of doesn't matter because the number will always be much higher in the air, but I thought it was a good exercise to get the lowest possible value.
No idea what any of that meant . What does dynamic idle do again ?
Why would some people run a higher idle rather than just using air mode? I guess I don't understand is disadvantage of air mode, other than the inversion time like you mentioned.
Nicely said Bardwell!
I thought motors had more toque at lower speeds... at least on my CNC machine with stepper (brushless) motors. Are quads different or is it something a bit different (time to spin up from very low rpm). I thought the desync was the command to push not being in sync with the motor (magnet) position due to the lack of torque from the last command. This being motor stall, not areodynamic stall. Curious if I've got this wrong. None of which changes the usefulness of dynamic idle and your awesome video (as always).
I'm a jackass... Had I watched 5 seconds more before posting I would have heard "motor stall", "motor stall", "motor stall". But I did catch the problem with the bench test... You need a lot more torque with motors attached. That was the "no load" stall speed.
It's not truly an issue of torque, but that's a short-hand way of expressing what's happening. One thing is that the position sensing of the ESC becomes less accurate at low rpm, because the back-EMF voltage is weaker. Therefore the chances of a stall are greater. Another thing is that when the prop is spinning slowly, the motor makes less thrust and therefore the PID loop has less authority.
Always learning from you Joshua. Just a clarification, in BF motor control do you consider the motor stalled when it stops rotating or when it starts to stutter? Yes, I'll watch the video again.
When it starts to stutter.
@@JoshuaBardwell Thank you.
Hi Joshua, I've got a problem with my Protek25 and my Betafpv 95 V2, when I fly between tries, and heppened I have to cut throttle to pass throw sometimes motor idel same speed for few seconds, I've tried to rise and again slow the rpm using the stick but nothing it slow down after few seconds and I crash on the tree.... It seems it happened when there is wind but when I fly them not beetween tries in the wind it doesn't heppened. Could it be a dynamic Idle adjustment to do? Many thanks
I'm going to pose a question and see what your answer is. Why don't we use a sensor based esc in FPV? My background is heavy in surface and in racing in particular they use a sensor based esc. There is a cable that runs from the motor to the esc that tells the esc where the rotor is in its rotation allowing the esc to make more precise drive inputs giving the driver a better "feel". This is mainly for lower rpm and usually is turned off at higher rpm as the sensor actually hinders hi rpm performance. I see where this could be beneficial, especially for dynamic idle.
So, pole crossings are very specifically how the ESC knows to initiate the next switching event. What your'e describing is core to how an electronic speed controller knows when to dump DC current into the motor windings, and switch it back off with the next one coming.
Getting this data that's already onboard the ESC (this is why Motor_Poles is a parameter) back to a flight controller via bidirectional DSHOT is the magic that enables RPM Filtering and Dynamic Idle.
@@tehllama42 thank you for that. Very clear answer.
@@xNmMmAx No problem - this does a much better job of explaining it - the back-EMF sensing is integratl to how it performs (and why some ESC settings like motor timing to med-high/23° affects performance so much)
howtomechatronics.com/how-it-works/how-brushless-motor-and-esc-work
#1, cost, #2 an aditional 20 wires per quad for sensors. But this would solve so much issues.
So why would we ever want the pid controller to have access to the min_idle. If the quad flies so much better with higher percentage and if response times are slower when coming out of that super low rpm then aren't you recommending everyone runs higher idle and to not run min_idle? In what scenario would the PID controller getting to use min_idle actually be beneficial?
a higher min idle results might result in less "hang time" when upside down 🙃.
@@georgedowning1803 I'm mostly speaking for racing and flight performance not for freestyle purposes
@@MewoFPV You should have more authority lower down. As assuming an ideal case you will have to spin the motor up only half as much as you now have 4 forces to play with not just 2. Might not be the case in the realworld though, maybe it takes the same time to spool them up more, or braking is to slow this low in their speed curve.
Altitude control will be better with a lower min idle.
@@MewoFPV MewoFPV dive gates? A reduced hang time will result in less control? In a competitive racing environment you would probs be flying with consistantly enough power that PID will never access min throttle anyways. However by increasing the min idle, when operating at the bottom of the power band, you effectively shift the center of rotation in the direction of the turn. An example of where this might be a problem is In prop wash, as the center of rotation Is moved rapidly any additional rotation command by the pilot will result in an ossilating track through the sky. Yes it becomes more rotationally more stable but for racing having a consistant track may be valued more than a consistent than attitude.
A while ago , I lowered the idle speed down as far as I could , on some kwads as low as 1.5 % and stopped using air mode . But will be interested in trying this out .
Hey man. I just hopped into fpv and now I'm here cause I was wondering about dynamic idle. In 4.5.1 , I just set the value for dynamic idle in the pid tab and that's it? motor idle (static) is greyed out in the motors tab after setting a value for dynamic idle in the pid tab. Or do I have to do the CLI thing too? Thx for all the time you've been putting into these videos. They help so much!! Cheers m8
I did notice the better braking from using it. The trick is finding the right idle speed to keep performance at its best
hey joshua i have a pavo25 and when punch out and drop the throttle the motors start twitching when recover the altitude and latter have this same twitches in fly! any suggestion? you are help a lot over this fpv world!
I need to check it myself too. But, on Mamba F405 MK2 stock min idle is 3.0 which is quite low already. Thanks for sharing Joshua!
The point is that with dynamic idle you can raise that so your quad flies better at zero throttle, without giving up the ability of the motors to go below that point if they need to.
@@JoshuaBardwell Hi never flown just got mobula7 hd can't set it up in betaflight just about got it binded I'm almost giving up on it if you could give us some info many thanks 👍
My brain is freezing... 😂 But a the third or sixth view Il will learn something for sure 👌
glad im not the only one who has to watch his videos more than once lol
Hi Joshua, can you please confirm on the updated firmware of DJI's goggles? I want to get goggles after the lockdown (no shipping to SA at the moment), and was almost going for the newest Fatshark's because they display all info on the OSD which is important for me. Now I've read in a comment on another video that DJI brought out a big update that adds all this OSD info to their goggles, which was holding me back from getting them..
Does this feature require bi-directional Dshot and BLHeli32 ESCs? Not mentioned in the video...
It requires bidirectional Dshot. You can use blheli32 or you can use JESC or JazzMaverick with blheli_s.
ua-cam.com/video/NgBJJG0slM4/v-deo.html
Should I round up or down? Also, all the motors don't spin exactly the same rpm, so should I go with the fastest, slowest, or somewhere in the middle?
Im a freestyle pilot and usualy up my idle speed a bit more than default. I found it helps on propwash. Thanks for the explanation... going to try in the nexr flight.
It's a bit of time and effort to get there, but if you can get the same awesome propwash performance, this can get you a bit better inverted throw
tehllama42 is that hard to tune dynamic idle ?
Cant you do this in the radio throttle curve setting also ? well some of it?
Thank u for teaching in to this. Hi from Latvia! Thank u for teach us more!
I noticed that, at least for my setup, the RPM value in BF is 50% that of the RPM value in BLHeli motors tab, for the same PWM. Any idea why? Basically BLHeli only lets me step from 1015 to 1017 and reflects 850RPM while BF shows 450 RPM for what appears to be the same PWM value.
This is the shiz you are good at. Freaking love it. Keep deep diving into this stuff.
thnx for the explanation! maybe you could make more video's about features we do not use much, e.g. thrust linearization
Thank you JB, On top of that, what is the behavior if you change value in BLheli like Throttle Min into a lower or upper value than the default in ESC firmware configuration ?
Quad66 knows his stuff peeps! the 1s magician! His latest tune video taught me more than any other tune vid.
I have a Mobula 8 and the Dynamic Idle is greyed out in Betaflight (i have firmware 4.5 with configurator 10.10.0), does this mean that Dynamic Idle isn't supported on that whoop?
Dammit Bardell, you're so smart, you're answering the question about deathrolls that I didn't go online and ask you about this weekend. The same quad, deathrolled twice lately, while descending into the next field (low throttle) where one motor on my (plus) frame was probably at idle. Without even checking, I bet my idle % is too low and you're a mind reading genius!
Can't find motor idle throttle value in betaflight's motors tab anymore
Thank you for the briefing, Joshua! Have you gathered any feedback/opinion from Captain V.? Also, what is/are the parameters that cause the idle value to vary between the two endpoint settings? For example, if idle is only adjusted downward in response to differential prop rate (e.g., rolls or flips), then inverted hang time would still be lost if we raise our dshot_idle_value, right?
That is a question that I am still trying to understand but the answer seems to be no. That right now, your motors are not spinning at idle when you are at zero throttle, because the PID loop must increase throttle to stabilize the quad. So if you enable dynamic idle, the actual rpm at zero throttle will be lower, and you can actually raise idle percent with no net change in hang time or speed of descent. Obviously at a certain point, the idle is so high that you feel a difference.
@@JoshuaBardwell Reading TFM (github.com/betaflight/betaflight/wiki/Tuning-Dynamic-Idle) helped me to understand your explanation ;-) It is still working against inverted hang time to raise our dshot_idle_value. Enabling dynamic idle allows a motor's drive to go to zero as long as its prop is still spinning at or above idle_min_rpm. The PID controller can thus command reduced drive (i.e., below the dshot_idle_value) on motors that need braking to maintain orientation (e.g., when inverted). The reduced net drive among all motors while inverted increases hang time. We can then raise our dshot_idle_value to consume as much of the the newly-acquired margin as we like.
I am in firmware 4.2 and using 10.7 beta flight configurator. I tried changing idle min both thru CLI and in the tuning tab and neither approach is saving. It’s currently set to zero and I want to change it to 14. What am I doing wrong ??? I have used CLI before to make changes and it’s always worked for other features. Please help !!! Thanks for the great videos
These are the videos I love
My motors tab does not shows RPM... and when setting the idle min rpm to any number and saving, after reboot it is 0 (zero) again... whats wrong?
I Run a 6 inch 6s build with brotherhobby avenger v2 2507 1200kv .. I find at low throttle I wobble just a tiny bit if in coming down into dirty air.. would bumping up the idle percent make more sense in my application rather then dynamic ?
Hi if the idle speed it high as to the point its still on the ground when in flight that would make your quad drop alt slower at 0 throtel correct
That's extremely interesting!!! I had never imagined that CaptainVani had 8% idle! Personally i had it down to 3% because 1) i find difficult to do inverted yaw spins because tends to fall faster 2) the quad takes forever to get down to the ground. I definitely learn something today. Question: Propellers pitch affect the idle ratio or dynamic idle?
Prop weight, pitch, and relative chord thickness all affect it, usually in that order. Heavy aggressive props will have you wanting to run higher values, thinner lighter ones with less pitch let you run lower values.
@@tehllama42 i think it's the opposite I run 4.8 pitch and it takes forever to get down to the ground on higher idle percentage
No no no Completely overwhelmed and caught in the ceiling fan can we try this again. OK I just need to re watch this and then review the FB post then watch it again I guess
Same here.
small correction: For BLDC's you shouldn't lose torque at lower speeds, torque is dependent entirely on magnetic flux in the core, which is governed by current. The motor certainly cannot _maintain_ high torque at low speed because inadequate cooling.
I suspect this is more of a fundamental problem with sensorless control of said motors. desync happens because the ESC losses track of where the rotor is relative to the stator and the motor losses all torque.
There is more to the torque claim, you're right. The desync is because at low speed, the BEMF that is used to detect pole crossings is very weak and is lost in the noise. The lack of authority is in part because when the prop is moving slowly, it isn't making much thrust.
But from what I am told by people developing the feature, the motor is less responsive coming from low rpm, which I approximate as "lack of torque" when explaining it. The motor will take longer to get from 2% throttle to 5% than from 50% to 53% for example. Once you get out of the low-rpm region, this problem goes away.
@@JoshuaBardwell Its interesting observation noneheless! No doubt that the practical wins out over this theory. I think I read somewhere that the sensorless ESCs often switch between algorithms at low speeds for this reason. Keep up the great work!
Anyone know if there's a way to override betaflight's max idle % to go beyond 20%? I'm doing some cinewhoop experimenting and this would be very helpful. Thanks.