A friend and I have noticed an unusual high amount of burned motors since we changed the ESC settings to 16kHz (low) and ByRPM (high). It happend to us after hitting race gates or small branches not to hard. The problem disappeared as quickly as it appeared after switching back to 24kHz for high and low PWM frequency. This issue was confirmed by other discord users on the UAV tech server also.
@@29jvskate we lost 5 motors while flying with 16kHz (low PWM) for 2 days. After switching back to 24kHz we flew around 40 packs so far and had a few hits too but all motors are still working.
@@Zentriefugal yeah I was getting some bad high throttle issues (vibrations) with it set on " by rpm" as well. Someone better at tuning might have been able to get rid of them. But I couldn't. Running the Mamba f7 dji 128kHz stack, with the fpvcycle imperials, apex frame.
I remember 3 years ago being a total noob i would run away from your channel cuz it was soo complicated for me. Right now im checking all of your videos to see how to improve my flight performance, see what new gears are out. Very helpful as always thank you JB !
Thank you for not falling out of the drone world. Because I do.. and when I come back.. you right there. Thanks man. I have some old bl heli 32 from like 4 years ago I'm totally gonna try this on. Been wanting to built a trash drone to practice flying low. I only do high and in fields. I can't control my altitude worth shit.
I am so glad you made this video and it's a happy coincidence that you released it today, when I really needed it. Thanks, from myself and all pilots who find this video helpful. You really are a valuable asset in the FPV community.
You the man man I just upgraded to this test code day before yesterday and it fixed an issue my quad was having with excessive noise on the pitch axes only due to using 32.8 It would cause a bad twitch on fast throttle moves which ultimately led to crashing a brand new build and ruining a new vista cam lens and fpv cycle motor
Great video! I had some troubles with that using 32.8 version. I put low pwm at 24 and high at 48 and the sound was horrible. The bad thing is that I did not realized it was because of that and I disassembled all the quad several times, changed some screws before realizing that the problem was the pwm low/high different values. After set then at 24 the sound was good again... Honestly I was lost and frustrated about that, even if I didn't realize that bad sound came after changing pwm frequency... Thanks Joshua!!
Flashed on Hobbywing XRotor (30x30 60A), I rolled back to 32.7. 32.83 was flying OK at fixed PWM feq., but not as good as 32.7. Enabling variable PWM Freq (16-By RPM) gave terrible results : bad noise, hot motors. Currently, it is still test code, might not work as well with every ESCs, be careful.
@@CesiumSalami Even if it was pretty OK at fixed 48kHz it wasn't as smooth as with 32.7 and it is audible. I'll give an other try when final release will come out, right now, with these ESCs, 32.7 flies better.
I made the mistake to buy a "xrotor 60 amp" ESC and I can only warn from those line of ESCs. Aside of being > 100$ priced, the pinout in the manual is wrong (GND and 5V pins are intermixed with actual sensor pin locations) The "telemetry" pin is only available in the ribbon cable (which has 2 undocumented pins) and it actually carries 5.5 volt and can destroy an FC. The CRT amp monitor pin is supposed to have a voltage based on current consumption, it has an inverse voltage of 2mV that drops to 0 when a bit current is drawn, so it's flawed. And on top, if you connect that thing to BLheli 32 suit then it can connect but it can not show any data.
I can’t really explain it, but after I flashed to 32.8.3 and using the “By RPM” feature I definitely felt the difference on my first pack. Low end throttle feels significantly more “in control” and the all out flight performance feels SUPER smooth, even the motors sound better. I definitely recommend this and I cnt wait for the official release.
D term response is tighter at lower PWM. You can still achieve that with fixed 48kHz PWM using thrust_linear, but I'm with you - I can't wait for 32.9, because that will finally present an upgrade over my BLHeli_S ESC's
@@erlandfpv The stock 25 value does really well for most, if it's a very under-authority craft you can go as high as 35. For lightweight (100g) racers on 2S packs that are really 3S quads (GTB339), I'll use a value of 15 to clean up a lot of the low throttle bobbles where a bit more P&D gain can settle things out better than having too much gain at high throttle.
I just followed this exact procedure step by step using the exact links and the exact procedure listed in the video. I found the exact one that was already on my quad on blheli 32 and matched it up to the list from GitHub I downloaded that and literally as soon as I hit reset up after it was done flashing three of my motors completely smoked and my ESC popped. I literally just got those motors and I'm completely devastated I don't have enough money to just oops and redo it.
Just a heads up for noobs like me... (I'm here to set PWM frequency to fix Jello in my o3 air unit)... BLHeliSuite32 won't be able to find your ESCs unless your battery is plugged in to your quad AND your USB cable is plugged in. As always, remove the props and use a little fan on the o3 air unit to avoid overheating ;) Much love! Thanks so much Joshua for all the great vids!!!
One other thing... I couldn't get rid of my jello even after adjusting the PWM to 48, soft mounting, AND gluing the o3 lens barrel. Turned out it was the props I was using. I went from Gemfan Floppy Proppy biblade props to Dal Fold Tri blade props and the Jello is gone now. Hope this helps somebody out there.
With regards to the different hex files I believe the 4IN1_Multi hex is 48kHz where as the 4IN1_96K_Multi hex means up to 96kHz modulation. Only helpful for smaller quads anyway though that spin at high enough RPM's to need it. (3/4in and below I think)
Version 32.8 on "airbot wraith32" ESCs gave me lot of trouble as long as I had selected different PWM freqs for "low" and "high", like 24 and 48. Motors were stuttering. They run smooth now however with 48/48.
@Alexander Dräbenstedt : and is low:24 high: by-rpm any good for you on 32.8.3 testcode with these Wraith32 ESCs? (and is 48/48 what you are running on 32.8.3 or on 32.8?)
Joshua, I am pretty sure that a 5" quad with "by rpm" enabled will only ever get to about~ 35-40khz no matter how high the ESC itself actually goes. I can post a comment below (im guessing you will have to approve it) with the link to a github post from sskaug with a very easy to understand graph as to how the feature works, but basically, thats as high as the second fundamental resonance of the erpm will go (its ~32khz in extremely simple chart in his post of a 2500kv 4s 5"). Basically the "by rpm" rides just above the second fundamental. At least that is my understanding. SO, a 24-96k esc would sort of be wasted on that feature. I am not sure how a 48-96k esc would handle that. I assume it would just sit on the min...
This is why I think an F4-based ESC with full range 24-128 or 16-128 whatever it is is the simplest answer. It'll give super low PWM at low throttle and go as high as the motor RPM demands, regardless of kv and prop size. But if you can do the math, you may find a certain F3-based ESC can do the job.
@@JoshuaBardwell I think so too. 24-128 sounds like it would be amazing. The "by rpm" can be useful on 16/24-48 ESCs, but even then.... The braking loss at 48k isn't that bad, so it's probably worth it just to use 48k straight up, rather than by-rpm which is going to be more like 24-35 (on F1 ESCs or whatever), which won't be as smooth. Idk. 16-128 sounds amazing though for sure.
it looks like JB hasnt approved my comment with the link (happens). If you want to see what I am talking about, go to the blheli github, look at issue 546, and scroll down to the comment from sskaug on aug 28
@@JoshuaBardwell found out that esc only supports 96kHz at most except for Fettek and Kiss even if esc itself supports and you install Rev32.8.3 testcode. mosfets themselves are not able to physically dial more than 96k hertz. what do you think about this? it’s not enough just to buy an esc with 16-128k, you also need to read the manual on the mosfets installed on it at the hardware level
"Don't let the fact that I cut my hair confuse you" 😅 That sentence confused me more than why you thought I should've been confused. Rewound the video to double-check. From one baldy to the other: THERE'S NO DIFFERENCE!
Ho due 5 pollici, uno con stack Speedybee f7 v3 e uno skystars+ ESC 60a ma non funzionano questi parametri. I motori tremano come affetti da Parkinson... (Ho provato by RPM ma è lì stesso). Ho dinamico idle attivo a 35, può essere che un'altra cosa dinamica vada in conflitto? Grazie
I hope this helps my fpv cycle motors. They start to make a sound around 70% like something is in the motor. They’re brand new with one flight on them…
I am Going to Build an FPV Fixed Wing Airplane, Should I Put the PWM Frequency High or low for Efficeny? Can I youse 2 Identicel BL Motors exactly Paralel Conected Motor Windings with only One ESC? I mean with not Twistet Motor Wires.
Hello Joshua I updated my Hobbywing ESC to 32.89 using loaded startup test code. The motors were making weird sounds and also were getting really really hot.
Hi, on esc-configurator I kept getting an error "Failed reading ESC 1, Failed reading ESC 2, Failed reading ESC 3, Failed reading ESC 4". Do you know how i can fix this?
Bro, I just got a Evoque F5d, and, first flight, went to gyroflow, and surprise!! Just can't use any of his O3 videos, cause it's flickering when smoothing by gyroflow. So the software says to use 48khz or more at ESC pwm, that's OK. I have changed it. But it is a 5" Drone, it will handle it or burn my motors? Setup in min pwm 48khz and maximum By RPM. Thank you!
Thank you Joshua. That is all nice and scientifically proven by guys like UAV Tech, Chiris Rosser,.... But does it really feel it is flying better that ''just'' having clean build, using bidirectional D-shot, minimize the filters and do some PID tuning?
Thanks for the video. This is my experience with this: I use BF 4.3 rc3 with RPM filtering ON. I use t motor p50a esc, with f80 pro motors. I use esc fw version 32.8.3. I tried motor timing auto, then 16 to 22, with demag medium and high. Whenever I set any combination of variable pwm, 16 to by rpm. 24 to 48, I get this metallic ringing sound pronounced when hovering. Motor temp slightly increased after micropunches at home (storm for days outside). But when I set PWM to be fixed, 24 24, 36 36, 48 48, I get supercold motors and perfect sound. Now, my question is: Is it normal for sound to change to a bit metallic one, at 20-30 throttle (cant use more at home). The frame is apex 6 inch. BF filter sliders were at default 1.0 position. Thanks for any info on this.
@@artemsloytsev1052 Hello. On one quad the sound disappeared when I used rushfpv extreme 60a secs, motor timing 16 and 24 to by-rpm setting. I am not sure what made sound go away on this build though. You can try these settings i mentied above.
@@nakmax nothing special. now using 23 timing on all quads, fixed 48 48 pwm, demag low. And some preset tunes. Motors cold and no issues. So, for me, variable pwm is not an option for now.
Well shit, blheli 32's cloud servers are dead, I think I'm on the version that has a chance of spooling up when you plug in. What should we do if there is anything we can do since the files are unavailable? I was halfway through updating my fleet when this hit.
These videos are like reading stereo instructions. And they never made sense either lol. I'm getting decent at flying but aside from soldering a new motor on my Nazgul, I have no idea what's what inside the frame. I really want to build my own now bur doubt I'll do it properly. Any advide on a kit or something? Thanks J.
@joshuabardwell - I went to flash my Foxeer Reaper 65A 128k ESC with 32.8.3 fix. It shipped with 32.8 but U only see a 96k version for both 32.7 and 32.8/8.3, Am I missing something stupid? I notice that on the shipped 32.8, as per your reccomendation, I can change my min/max - to 24k/128k so it is definitely not the 96k version.
I've been trying for 2 days to get my kwad to connect in blheli but I just get an hourglass and lockup. Close it and open again trying different things but nothing is working. Idkwtf to do.
I have a set of Airbot Wraith32 mini st blheli32 escs on my v1 Kopis. I think they would have the f0 processor. Would I still be able to run 32.8.3 testcode for some benefit, or would byRPM not work at all on these? -i'm trying to gauge if should update to 32.8.3 or rollback to 32.7 on these (they were recently updated to 32.8 from 32.2 in order to enable the rpm filter/bidirectional dshot, but never tested out in flight)
Hi JB, I fly with PWM @48kHz, with thrust linear set at 25. Do you imagine this variable PWM will have a detrimental effect with the throttle curve from thrust linear?
Hobbywing 60a 4 in 1 only goes up to 48khz so do I use "by rpm" for the high and 24 for the low? I'm also unsure as if it can go to 16khz but it will do Dshot 1200 don't know if that factors in this setup
DShot 1200 is what the FC tells the ESC to do. In current versions of BF, the highest DShot you can do is DShot 600. They removed the option of DShot 1200 after BF was released. BF 3.5.7 is the last one I now of that can support DShot 1200, but that was before we could do RPM filtering since BF 4.0 was the first that could do that.
@@BikeGuyFPV oh ok I see that makes sense why I couldn't find it in BF, do you think the " by rpm" will just go up to 48khz on escs that have the max as such?
@@Q1320 not sure as it wasn't really explained in the video on what the advantage was with "By RPM", only what the low and high PWM was doing. I'm taking a guess on this, but setting your low PWN as low as it would go handles low throttle stuff. Then as you increase throttle, the PWM frequency would ramp up to whatever is deemed necessary on the fly. 48kHz being the max but it might just hover just under it depending on what is going on based on the eRPM of the motor. Only then would it push the frequency higher until it hits it's limit (48kHz, 96kHz, or 128kHz depending on your ESC).
Joshua, do you know if F4 ESCs are downwards compatible with same model's F3 version, meaning can I basically just swap an F3 with an F4 or would there be anything to watch out for? And what's the advantage of the F4, can they just run higher Dshot or is it just future-proofing?
I ended up putting my frequency back down to 24khz on my long range 8" with small motors. It flew a little worse. Think it needed that braking performance. I think I did get more flight time but seemed to wobble more. Slow wobble. Idk. Maybe it's in my mind
@@zachcarrizales5038 32.8. What ever was the latest a few months ago. Was in process of tuning the quad. Figured I'd try again if I got it flying great and see if then I could just tweak it some
I don't see on my windows computer anything about a raw file so I got stumped from the very beginning. What program are you opening the file with. All I have is word and notepad
Do you really need 100% ramp up power? Uavtechs has his around 12-13% and chris has around 20% for 5” quads. Ive smoked 4 motors now and i dont know where i went wrong with stock filters and relatively low D gains. My master multiplier is at 1.3
Smoking motors can be caused by a Lot of things.. trying to arm when the motors cant spin, for one. If they smoke while flying, IT can be a problem with filter settings, but stock filters shouldnt be a problem on a 5". Also, trying to fly with a mechanical dirty build, ie old wonky motors, a loose FC flapping around, etc, can make bad things happen. Or loosing a blade on a prop, can make the motors really hot in seconds. Make sure everything is right with the build, motors doesn't have a bent shaft, and props arent bent too bad, and try again.
I’m not sure about 100% ramp up - it was 50% prior to JB updating from 32.8 (and I think I’ve seen 50% being mentioned elsewhere on either a uavtech or Chris rosser vid -given JB doesn’t explicitly mention that ramp up value change in this vid, I would exercise caution until he can confirm).
I do not set rampup to 100%. That most be a decision by the maker of this bind n fly quad. There are little benefits to increasing rampup past a relatively low number and increased chance of smoking motors.
I'm gonna try this on my DJI Diatone Roma F5. I just switched it from Betaflight to Emuflight cause I like the tuning on there better not to mention Bidirectional keeps giving me error no matter if I change 8k to 4K and all that. Anyways I put my GoPro 9 on the Roma F5 and the weight was very noticable so I took it off and took up my Bandolero with a KISS f7 FC and I couldn't even notice it. Do you think that running on 24 PWM frequency could have been the problem and that the upgrade could actual help? Sorry for the very long winded question. Just wanted to explain as much as possible.
A quick test would be to plug in your Emuflight board and (if you haven't already) go to the Rate tab and set 'Feathered PIDS' to 0. This makes it so that your PID controller runs completely off the 'Error' value like the KISS PID controller does rather than the 'Measurement' value that BF uses. This should let you make a more apples-to-apples comparison between the two quads. At that point I would try raising PWM frequency on the Roma, and see if that improves the performance compared to the KISS quad. At 24HZ its definitely possible that you 're not getting the motor efficiency you need to move a heavier quad the way you want to. Hope this helps. BTW I run Emu and much prefer the KISS style PID loop (Feathered PIDS at 0). If you haven't tried this yet you're in for a treat.
How much efficiency (or breaking power) is actually gained here, given we indeed constantly change motor speed and thus, I guess, only use higher PWM frequencies less than 10% of the time? Every time we lower the throttle, a lower PMW frequency kicks in, no?
You can watch Chris Rosser's video linked in the video description for numbers. To me theain advantage of the higher PWM frequency is avoiding the motor harmonic. The extra efficiency is secondary at least in 4" and up quads. The increased braking performance is substantial comparing 48k or 96k to 24k or 16k.
Is there a way to do this without usb? With Wi-Fi erls, basically how you can on betaflight . Reason asking , I saved a quad I had down because of a bad usb
Hi future Joshua, I have a question: Why does a high PWM frequency hurt breaking power? I've seen the graphs of Chris but I cannot see any physical reason why that should be.
@@tehllama42 Sorry, but I could not find anything from Ryan relating to breaking. I understand why a high PWM frequency can hurt torque on low RPMs but I fail to see why that effect is so much higher on breaking.
I was able to follow up to 2:30, When i "Read Setup" it comes back with NOTHING! 1-4 esc's do not show any data and an error says "No valid ESC configuration". Where did i go wrong?
Okay, that's cool. But, what is 'By RPM' really achieving? Is it supposed to be working along the same basis as how the variable PWM frequency works in AM32? Or better yet, is there a GitHub wiki on what 'By RPM' is actually doing?
@@JoshuaBardwell okay cool. It wasn't really explained what 'By RPM' did in the video when enabled, versus what BLHeli32 was trying to achieve with the changing frequency based on throttle only which was explained. That's why I asked.
Thanks, Joshua! Apparently my Tekko32 F3 goes from 48 to 96khz (IIRC) and my T-Motor F4 goes from 16 to 48khz (IIRC) Is it still safe/useful to set BLHELI Max PWM to "by RPM"?
You can, but it may be depending on your motor rpm that the optimal PWM frequency is not available. In that case, the ESC will do the best it can, and you won't be worse off than with fixed PWM.
I have the exact same iFlight ESC in the video. I downloaded the exact same iFlight file, and put it in the right folder, but in BLHeliSuite I don't see the hex file, it just shows 32.6, 32.7 and 32.8. I even updated BLHeliSuite to the latest version. Did I miss something??
It must mean you downloaded the wrong hex file, or maybe you accidentally download the file incorrectly due to github confusing way of managing file downloads.
@@JoshuaBardwell Yea, I'm a dumbass... after rebooting and trying it again, I decided to download the iFlight WITHOUT 96k, and that one worked... 🤷♂️ (just smack me now...lol...)
@@JoshuaBardwell hi jb just a update .on my new hd h7 kakute I had errors in the motors tab when trying to bidirectional dshot .after I flashed back to 32.8 no errors .maybe 4.3 when released it might work with the h7. Only sending this jb fpr feedback lol not to bother you .happy flying
BLHeli 32.9 has released. You do not need to download the testcode any more. You can simply flash a normal version of blheli32 version 32.9 now.
A friend and I have noticed an unusual high amount of burned motors since we changed the ESC settings to 16kHz (low) and ByRPM (high). It happend to us after hitting race gates or small branches not to hard.
The problem disappeared as quickly as it appeared after switching back to 24kHz for high and low PWM frequency.
This issue was confirmed by other discord users on the UAV tech server also.
@@Zentriefugal any updates to this? have you tried changing back to 16KHz to see if the problem persists after switching back to 24khz?
@@29jvskate we lost 5 motors while flying with 16kHz (low PWM) for 2 days. After switching back to 24kHz we flew around 40 packs so far and had a few hits too but all motors are still working.
@@Zentriefugal wow very very important info thank you for sharing! Lots of money could get burned up if this isn't fixed.
@@Zentriefugal yeah I was getting some bad high throttle issues (vibrations) with it set on " by rpm" as well. Someone better at tuning might have been able to get rid of them. But I couldn't. Running the Mamba f7 dji 128kHz stack, with the fpvcycle imperials, apex frame.
Finally, a VTEC kinda technology on an electric motor!! 🤩
I remember 3 years ago being a total noob i would run away from your channel cuz it was soo complicated for me. Right now im checking all of your videos to see how to improve my flight performance, see what new gears are out. Very helpful as always thank you JB !
Thank you for not falling out of the drone world. Because I do.. and when I come back.. you right there. Thanks man. I have some old bl heli 32 from like 4 years ago I'm totally gonna try this on. Been wanting to built a trash drone to practice flying low. I only do high and in fields. I can't control my altitude worth shit.
I am so glad you made this video and it's a happy coincidence that you released it today, when I really needed it. Thanks, from myself and all pilots who find this video helpful. You really are a valuable asset in the FPV community.
was just looking for 32.83 last night... 32.8 has been terrible for my quads.
Perfect timing per usual.
Danko
for me too. I was getting near fly-aways. Huge glitches when hovering.
You the man man I just upgraded to this test code day before yesterday and it fixed an issue my quad was having with excessive noise on the pitch axes only due to using 32.8
It would cause a bad twitch on fast throttle moves which ultimately led to crashing a brand new build and ruining a new vista cam lens and fpv cycle motor
Great video! I had some troubles with that using 32.8 version.
I put low pwm at 24 and high at 48 and the sound was horrible. The bad thing is that I did not realized it was because of that and I disassembled all the quad several times, changed some screws before realizing that the problem was the pwm low/high different values.
After set then at 24 the sound was good again...
Honestly I was lost and frustrated about that, even if I didn't realize that bad sound came after changing pwm frequency...
Thanks Joshua!!
wow - you have a video on frigin' EVERYTHING !!!
This tutorial works with a DJI Air Unit O3, and the recommended settings for now are a minimum of 48kHz and a maximum of 96kHz.
HOW DID I MISS THIS VIDEO????!!!! 🤦🏻♂️🤦🏻♂️🤦🏻♂️
Thanks again, as always, JB!!
So glad you did this! BLHELI_32 3.83 is phenomenal. The combination of BF 4.3 and this firmware feels amazing.
Flashed on Hobbywing XRotor (30x30 60A), I rolled back to 32.7. 32.83 was flying OK at fixed PWM feq., but not as good as 32.7. Enabling variable PWM Freq (16-By RPM) gave terrible results : bad noise, hot motors. Currently, it is still test code, might not work as well with every ESCs, be careful.
you can always choose 24khz as the low - if you were at fixed 48khz before and switch to a lower pwm (even if 'By RPM') you may need to retune.
@@CesiumSalami Even if it was pretty OK at fixed 48kHz it wasn't as smooth as with 32.7 and it is audible. I'll give an other try when final release will come out, right now, with these ESCs, 32.7 flies better.
Same thing for me, hot motors and Bad noise... Return to 32.7 ans it's ok !
My esc's are aikon ak32 55A v3
@@samuelaugry478 Your ESCs are STM32 F0 based, like mine. Maybe it has something to do with our common issue.
I just tried 32.89 today and i still had the same issue with the Hobbywing X rotor esc
Thanks Joshua! You make the most detailed and easy to follow FPV tutiroals out there. You are appreciated :)
Just rocking my Mobula 6 with the BLheli_S ESC but damn if this isn't a great video. Loved the explanation on the PWM. Thanks Joshua!
I made the mistake to buy a "xrotor 60 amp" ESC and I can only warn from those line of ESCs.
Aside of being > 100$ priced, the pinout in the manual is wrong (GND and 5V pins are intermixed with actual sensor pin locations)
The "telemetry" pin is only available in the ribbon cable (which has 2 undocumented pins) and it actually carries 5.5 volt and can destroy an FC.
The CRT amp monitor pin is supposed to have a voltage based on current consumption, it has an inverse voltage of 2mV that drops to 0 when a bit current is drawn, so it's flawed.
And on top, if you connect that thing to BLheli 32 suit then it can connect but it can not show any data.
Thank you josh, for a beginner es very difficult, you explainin very clear
Really, really interesting! 😃
Thanks, Joshua!!!
Stay safe there with your family! 🖖😊
Thanks JB. Just went through the process.. will test it out tomorrow.
Should be all in one video. You just learned something today.
I can’t really explain it, but after I flashed to 32.8.3 and using the “By RPM” feature
I definitely felt the difference on my first pack. Low end throttle feels significantly more “in control” and the all out flight performance feels SUPER smooth, even the motors sound better.
I definitely recommend this and I cnt wait for the official release.
Wow, thats nice
D term response is tighter at lower PWM. You can still achieve that with fixed 48kHz PWM using thrust_linear, but I'm with you - I can't wait for 32.9, because that will finally present an upgrade over my BLHeli_S ESC's
@@tehllama42 thrustlinear, what number did you using?
@@erlandfpv The stock 25 value does really well for most, if it's a very under-authority craft you can go as high as 35. For lightweight (100g) racers on 2S packs that are really 3S quads (GTB339), I'll use a value of 15 to clean up a lot of the low throttle bobbles where a bit more P&D gain can settle things out better than having too much gain at high throttle.
@@tehllama42 nice sharing, im using "25" for my setup, reference from uavtech presets, even for my 7 inch, i think need to raise the number?
I just followed this exact procedure step by step using the exact links and the exact procedure listed in the video. I found the exact one that was already on my quad on blheli 32 and matched it up to the list from GitHub I downloaded that and literally as soon as I hit reset up after it was done flashing three of my motors completely smoked and my ESC popped. I literally just got those motors and I'm completely devastated I don't have enough money to just oops and redo it.
Joshua bardwell
So which quad, motors etc do/did you have?
Yikes.
Just a heads up for noobs like me... (I'm here to set PWM frequency to fix Jello in my o3 air unit)... BLHeliSuite32 won't be able to find your ESCs unless your battery is plugged in to your quad AND your USB cable is plugged in. As always, remove the props and use a little fan on the o3 air unit to avoid overheating ;) Much love! Thanks so much Joshua for all the great vids!!!
One other thing... I couldn't get rid of my jello even after adjusting the PWM to 48, soft mounting, AND gluing the o3 lens barrel. Turned out it was the props I was using. I went from Gemfan Floppy Proppy biblade props to Dal Fold Tri blade props and the Jello is gone now. Hope this helps somebody out there.
always the best content!! Thanks @JB
DUDE can't wait to try this right now
Excited to try this, got it on my 5in if all goes well will try on my 7in
The braking strength between 16khz and 48khz is marginal, but the motor heating up is significantly more at 16khz.
Thanks JB, u rock!
Thanks for the heads-up, man!
Weird that you drop this video the DAY after I finish hooking and flashing up my new tekko32
The simulation ain’t even trying to hide itself anymore 🤣
Excellent 💯👍👍
You save my life. Again, yes. 😁😇🤒
With regards to the different hex files I believe the 4IN1_Multi hex is 48kHz where as the 4IN1_96K_Multi hex means up to 96kHz modulation. Only helpful for smaller quads anyway though that spin at high enough RPM's to need it. (3/4in and below I think)
Version 32.8 on "airbot wraith32" ESCs gave me lot of trouble as long as I had selected different PWM freqs for "low" and "high", like 24 and 48. Motors were stuttering. They run smooth now however with 48/48.
@Alexander Dräbenstedt
: and is low:24 high: by-rpm any good for you on 32.8.3 testcode with these Wraith32 ESCs? (and is 48/48 what you are running on 32.8.3 or on 32.8?)
Joshua, I am pretty sure that a 5" quad with "by rpm" enabled will only ever get to about~ 35-40khz no matter how high the ESC itself actually goes. I can post a comment below (im guessing you will have to approve it) with the link to a github post from sskaug with a very easy to understand graph as to how the feature works, but basically, thats as high as the second fundamental resonance of the erpm will go (its ~32khz in extremely simple chart in his post of a 2500kv 4s 5"). Basically the "by rpm" rides just above the second fundamental. At least that is my understanding.
SO, a 24-96k esc would sort of be wasted on that feature. I am not sure how a 48-96k esc would handle that. I assume it would just sit on the min...
This is why I think an F4-based ESC with full range 24-128 or 16-128 whatever it is is the simplest answer. It'll give super low PWM at low throttle and go as high as the motor RPM demands, regardless of kv and prop size. But if you can do the math, you may find a certain F3-based ESC can do the job.
@@JoshuaBardwell I think so too. 24-128 sounds like it would be amazing. The "by rpm" can be useful on 16/24-48 ESCs, but even then.... The braking loss at 48k isn't that bad, so it's probably worth it just to use 48k straight up, rather than by-rpm which is going to be more like 24-35 (on F1 ESCs or whatever), which won't be as smooth. Idk. 16-128 sounds amazing though for sure.
it looks like JB hasnt approved my comment with the link (happens). If you want to see what I am talking about, go to the blheli github, look at issue 546, and scroll down to the comment from sskaug on aug 28
@@JoshuaBardwell found out that esc only supports 96kHz at most except for Fettek and Kiss even if esc itself supports and you install Rev32.8.3 testcode. mosfets themselves are not able to physically dial more than 96k hertz. what do you think about this? it’s not enough just to buy an esc with 16-128k, you also need to read the manual on the mosfets installed on it at the hardware level
I think advantages have been demonstrated when the FET is driven at higher frequency, even if the FET itself cannot switch at that frequency.
Thanks JB
Watched the video and immediately performed the upgrade last night. Just ran 2 packs thru my 5", and it was definitely worth it! Thank you @joshua
"Don't let the fact that I cut my hair confuse you" 😅 That sentence confused me more than why you thought I should've been confused. Rewound the video to double-check. From one baldy to the other: THERE'S NO DIFFERENCE!
Thank you 👍👍
Now this was the video ppl need to see. 32.8 was trash. Little ppl knew about 32.8.3 until now and this is what we need ya for lol
I wonder if this is something that Bluejay will be able to emulate on BLHeli_S ESC's
I just test it i smoke motor 4 without props doing full throttle test
Can you make a new explenation video of BLHeli32 suite and BLheli_S/Bluejay configurator?
Joshua. Can you make a tutorial on setting up Blheli 32 for low kv motors spinning larger props. Perhaps a 530kv spinning 12 inch props
I tested it today, motor gets warm and sounds very bad, does not fly better.
Nice!
Ho due 5 pollici, uno con stack Speedybee f7 v3 e uno skystars+ ESC 60a ma non funzionano questi parametri. I motori tremano come affetti da Parkinson... (Ho provato by RPM ma è lì stesso). Ho dinamico idle attivo a 35, può essere che un'altra cosa dinamica vada in conflitto? Grazie
I hope this helps my fpv cycle motors. They start to make a sound around 70% like something is in the motor. They’re brand new with one flight on them…
I am Going to Build an FPV Fixed Wing Airplane,
Should I Put the PWM Frequency High or low for Efficeny?
Can I youse 2 Identicel BL Motors exactly Paralel Conected Motor Windings with only One ESC? I mean with not Twistet Motor Wires.
You should buy different ESCs than "software-controlled". Buy by hardware, not software programmable. Two separate for two motors.
I'm waiting for when cyborg Joshua timewarps in from the far future, with BLheli256 ESCs
Hello Joshua I updated my Hobbywing ESC to 32.89 using loaded startup test code. The motors were making weird sounds and also were getting really really hot.
I had the same issue with 32.8 with the hobbywing esc. I put 32.7 back on with no issues. I'm going to try 32.9 tomorrow. hopefully it's fixed?
@@BurgerPlanetLive yeah I updated to 32.9. It seems to be OK
Hi, on esc-configurator I kept getting an error "Failed reading ESC 1, Failed reading ESC 2, Failed reading ESC 3, Failed reading ESC 4". Do you know how i can fix this?
Heck yeah thank you Joshua! I have been wondering, does it matter what the pwm rate setting in Betaflight is?
Bro, I just got a Evoque F5d, and, first flight, went to gyroflow, and surprise!! Just can't use any of his O3 videos, cause it's flickering when smoothing by gyroflow. So the software says to use 48khz or more at ESC pwm, that's OK. I have changed it. But it is a 5" Drone, it will handle it or burn my motors? Setup in min pwm 48khz and maximum By RPM. Thank you!
We will see just made the move. Will an older version of beta flight be any issues with this upgrade?
🙌🙌🙌
Hi Josh, can you explain some turning skills with blheli_32 for fixed-wing users? Like how to get more efficiency from the new firmware etc. Thanks!
I already hit the like button my dude, i was the first LOL
How about Motor Timing setup - its ok to leave 24 degree ? Into By RPM mode on...
Thank you Joshua. That is all nice and scientifically proven by guys like UAV Tech, Chiris Rosser,.... But does it really feel it is flying better that ''just'' having clean build, using bidirectional D-shot, minimize the filters and do some PID tuning?
It's all small improvements but it adds up.
sure. check the shorter delay in pidtoolbox
@@JoshuaBardwell that makes sense, thank you.
Thanks for the video. This is my experience with this: I use BF 4.3 rc3 with RPM filtering ON. I use t motor p50a esc, with f80 pro motors. I use esc fw version 32.8.3. I tried motor timing auto, then 16 to 22, with demag medium and high. Whenever I set any combination of variable pwm, 16 to by rpm. 24 to 48, I get this metallic ringing sound pronounced when hovering. Motor temp slightly increased after micropunches at home (storm for days outside). But when I set PWM to be fixed, 24 24, 36 36, 48 48, I get supercold motors and perfect sound. Now, my question is: Is it normal for sound to change to a bit metallic one, at 20-30 throttle (cant use more at home). The frame is apex 6 inch. BF filter sliders were at default 1.0 position. Thanks for any info on this.
I have the same issue with multiple Blheli_32 quads, I can't find why it's happening. Have you solved it ?
@@artemsloytsev1052 Hello. On one quad the sound disappeared when I used rushfpv extreme 60a secs, motor timing 16 and 24 to by-rpm setting. I am not sure what made sound go away on this build though. You can try these settings i mentied above.
Appart from fixed min and max pwm, what are the other settings you are usually using on a 5 inch that gives you cool motors? thnx
@@nakmax nothing special. now using 23 timing on all quads, fixed 48 48 pwm, demag low. And some preset tunes. Motors cold and no issues. So, for me, variable pwm is not an option for now.
Well shit, blheli 32's cloud servers are dead, I think I'm on the version that has a chance of spooling up when you plug in. What should we do if there is anything we can do since the files are unavailable? I was halfway through updating my fleet when this hit.
These videos are like reading stereo instructions. And they never made sense either lol. I'm getting decent at flying but aside from soldering a new motor on my Nazgul, I have no idea what's what inside the frame. I really want to build my own now bur doubt I'll do it properly. Any advide on a kit or something? Thanks J.
Womp 32.8 and the pre-release version were causing motor errors for me
@joshuabardwell - I went to flash my Foxeer Reaper 65A 128k ESC with 32.8.3 fix. It shipped with 32.8 but U only see a 96k version for both 32.7 and 32.8/8.3, Am I missing something stupid? I notice that on the shipped 32.8, as per your reccomendation, I can change my min/max - to 24k/128k so it is definitely not the 96k version.
I've been trying for 2 days to get my kwad to connect in blheli but I just get an hourglass and lockup. Close it and open again trying different things but nothing is working. Idkwtf to do.
I have a set of Airbot Wraith32 mini st blheli32 escs on my v1 Kopis. I think they would have the f0 processor.
Would I still be able to run 32.8.3 testcode for some benefit, or would byRPM not work at all on these?
-i'm trying to gauge if should update to 32.8.3 or rollback to 32.7 on these (they were recently updated to 32.8 from 32.2 in order to enable the rpm filter/bidirectional dshot, but never tested out in flight)
I download it then what? What folder do you I do what with? So confused
Hi JB, I fly with PWM @48kHz, with thrust linear set at 25. Do you imagine this variable PWM will have a detrimental effect with the throttle curve from thrust linear?
Спасибо за твою работу! скажи, а могу я пользоваться этими функциями если мой ESC поддерживает только 96khz? будет ли это хорошо?
Wonder if your bluejay video kicked them into gear lol
For me, it did not see the local hex files. But still, at least I've upgraded to the released version 32.8 from 32.7
Went fine after the downloading blheli suite from their github releases
Hobbywing 60a 4 in 1 only goes up to 48khz so do I use "by rpm" for the high and 24 for the low? I'm also unsure as if it can go to 16khz but it will do Dshot 1200 don't know if that factors in this setup
DShot 1200 is what the FC tells the ESC to do. In current versions of BF, the highest DShot you can do is DShot 600. They removed the option of DShot 1200 after BF was released. BF 3.5.7 is the last one I now of that can support DShot 1200, but that was before we could do RPM filtering since BF 4.0 was the first that could do that.
@@BikeGuyFPV oh ok I see that makes sense why I couldn't find it in BF, do you think the " by rpm" will just go up to 48khz on escs that have the max as such?
@@Q1320 not sure as it wasn't really explained in the video on what the advantage was with "By RPM", only what the low and high PWM was doing. I'm taking a guess on this, but setting your low PWN as low as it would go handles low throttle stuff. Then as you increase throttle, the PWM frequency would ramp up to whatever is deemed necessary on the fly. 48kHz being the max but it might just hover just under it depending on what is going on based on the eRPM of the motor. Only then would it push the frequency higher until it hits it's limit (48kHz, 96kHz, or 128kHz depending on your ESC).
Joshua, do you know if F4 ESCs are downwards compatible with same model's F3 version, meaning can I basically just swap an F3 with an F4 or would there be anything to watch out for? And what's the advantage of the F4, can they just run higher Dshot or is it just future-proofing?
Anyone know if the "Hobbywing - XRotor Micro 40A(20x20) 6S 4in1 ESC" supports this? I can't find anything about it'sm PWM range or it's processor.
Did this and my Quad is now jittering quite a lot, motors getting warm. Followed directions exactly. Any ideas why this might give my quad Tourette's?
can you confirm the blheli32 app on android is safe to use with the 32.8.3 testcode? i want to change setting in field. thanks joshua
Yes, it is. Also using it and no problems so far
@@MrSmithFPV nice
I ended up putting my frequency back down to 24khz on my long range 8" with small motors. It flew a little worse. Think it needed that braking performance. I think I did get more flight time but seemed to wobble more. Slow wobble. Idk. Maybe it's in my mind
What version of BLHeli firmware were you on and did you tune the kwad?
@@zachcarrizales5038 32.8. What ever was the latest a few months ago. Was in process of tuning the quad. Figured I'd try again if I got it flying great and see if then I could just tweak it some
I don't see on my windows computer anything about a raw file so I got stumped from the very beginning. What program are you opening the file with. All I have is word and notepad
Does anyone have any information on why the pwm freq affects braking this much?
Do you really need 100% ramp up power? Uavtechs has his around 12-13% and chris has around 20% for 5” quads.
Ive smoked 4 motors now and i dont know where i went wrong with stock filters and relatively low D gains. My master multiplier is at 1.3
Smoking motors can be caused by a Lot of things..
trying to arm when the motors cant spin, for one.
If they smoke while flying, IT can be a problem with filter settings, but stock filters shouldnt be a problem on a 5".
Also, trying to fly with a mechanical dirty build, ie old wonky motors, a loose FC flapping around, etc, can make bad things happen. Or loosing a blade on a prop, can make the motors really hot in seconds.
Make sure everything is right with the build, motors doesn't have a bent shaft, and props arent bent too bad, and try again.
I’m not sure about 100% ramp up - it was 50% prior to JB updating from 32.8 (and I think I’ve seen 50% being mentioned elsewhere on either a uavtech or Chris rosser vid -given JB doesn’t explicitly mention that ramp up value change in this vid, I would exercise caution until he can confirm).
I do not set rampup to 100%. That most be a decision by the maker of this bind n fly quad. There are little benefits to increasing rampup past a relatively low number and increased chance of smoking motors.
@@JoshuaBardwell thanks for confirming 👍👍👍
I'm gonna try this on my DJI Diatone Roma F5. I just switched it from Betaflight to Emuflight cause I like the tuning on there better not to mention Bidirectional keeps giving me error no matter if I change 8k to 4K and all that. Anyways I put my GoPro 9 on the Roma F5 and the weight was very noticable so I took it off and took up my Bandolero with a KISS f7 FC and I couldn't even notice it. Do you think that running on 24 PWM frequency could have been the problem and that the upgrade could actual help? Sorry for the very long winded question. Just wanted to explain as much as possible.
A quick test would be to plug in your Emuflight board and (if you haven't already) go to the Rate tab and set 'Feathered PIDS' to 0. This makes it so that your PID controller runs completely off the 'Error' value like the KISS PID controller does rather than the 'Measurement' value that BF uses. This should let you make a more apples-to-apples comparison between the two quads. At that point I would try raising PWM frequency on the Roma, and see if that improves the performance compared to the KISS quad. At 24HZ its definitely possible that you 're not getting the motor efficiency you need to move a heavier quad the way you want to. Hope this helps. BTW I run Emu and much prefer the KISS style PID loop (Feathered PIDS at 0). If you haven't tried this yet you're in for a treat.
How much efficiency (or breaking power) is actually gained here, given we indeed constantly change motor speed and thus, I guess, only use higher PWM frequencies less than 10% of the time? Every time we lower the throttle, a lower PMW frequency kicks in, no?
You can watch Chris Rosser's video linked in the video description for numbers. To me theain advantage of the higher PWM frequency is avoiding the motor harmonic. The extra efficiency is secondary at least in 4" and up quads. The increased braking performance is substantial comparing 48k or 96k to 24k or 16k.
@@JoshuaBardwell Thx, will look into that!
Hey could you help me find someone who tunes quads. i just built ur 2021 best freesytle quad and need some help getting it tuned. tks
Almost said fiasco I caught that one 😄
does the FC still talk dshot to the ESC?
Is there a way to do this without usb? With Wi-Fi erls, basically how you can on betaflight . Reason asking , I saved a quad I had down because of a bad usb
No, it's not possible to flash the ESC with wifi any way that I know.
@@JoshuaBardwell thank you
"My ESC only goes to 48KHz" *PUH* Who are you Pewdiepie??
Hi future Joshua, I have a question: Why does a high PWM frequency hurt breaking power? I've seen the graphs of Chris but I cannot see any physical reason why that should be.
Check out MiniQuadTestBench for that data, Ryan provides a fantastic explanation (or the JB-Ryan Harrell video discussing those)
@@tehllama42 Sorry, but I could not find anything from Ryan relating to breaking. I understand why a high PWM frequency can hurt torque on low RPMs but I fail to see why that effect is so much higher on breaking.
Your voice and talking somehow sounds different, more controlled. Have you been training your presentation skills?
I was able to follow up to 2:30, When i "Read Setup" it comes back with NOTHING! 1-4 esc's do not show any data and an error says "No valid ESC configuration". Where did i go wrong?
First guess is your ESC is blheli_s not blheli_32. In that case you won't be able to use this feature.
@@JoshuaBardwell you again are correct. thanks for the save.
I tried updating my ESCs to BLheli 32 version 32.9 and it smoked... No physical defects whatsoever. Anybody else had this issue?
Wasnt this available since last yr?
Okay, that's cool. But, what is 'By RPM' really achieving? Is it supposed to be working along the same basis as how the variable PWM frequency works in AM32? Or better yet, is there a GitHub wiki on what 'By RPM' is actually doing?
Yes it's similar to am32 variable PWM.
@@JoshuaBardwell okay cool. It wasn't really explained what 'By RPM' did in the video when enabled, versus what BLHeli32 was trying to achieve with the changing frequency based on throttle only which was explained. That's why I asked.
Thanks, Joshua!
Apparently my Tekko32 F3 goes from 48 to 96khz (IIRC)
and my T-Motor F4 goes from 16 to 48khz (IIRC)
Is it still safe/useful to set BLHELI Max PWM to "by RPM"?
You can, but it may be depending on your motor rpm that the optimal PWM frequency is not available. In that case, the ESC will do the best it can, and you won't be worse off than with fixed PWM.
Can we using BY RPM option with Kiss v2 fc ? Leave all 48khz by standart?
You can use by RPM with any FC as far as I know.
I have the exact same iFlight ESC in the video. I downloaded the exact same iFlight file, and put it in the right folder, but in BLHeliSuite I don't see the hex file, it just shows 32.6, 32.7 and 32.8. I even updated BLHeliSuite to the latest version. Did I miss something??
It must mean you downloaded the wrong hex file, or maybe you accidentally download the file incorrectly due to github confusing way of managing file downloads.
@@JoshuaBardwell Yea, I'm a dumbass... after rebooting and trying it again, I decided to download the iFlight WITHOUT 96k, and that one worked... 🤷♂️ (just smack me now...lol...)
One question. So motor timing is set to 24 degrees when using these new setting? or to Auto?
Check Chris Rosser's video for the recommendation he makes.
@@JoshuaBardwell Which video and you supply a link. Thank you in advance.
@@JoshuaBardwell found it thanks!
Guess I'll wait there's to many files for the tekko 32 f4 4 in 1 esc there's like 10 different flies. G,h ,multiple etc
Download all. Put them in the folder. BLHeliSuite will automatically pick the one that matches the current ESC.
@@JoshuaBardwell got it done going out to fly some packs now to try out the by rpm thanks again
It does feel smoother.flew 12 packs lol it kinda felt like my pids changed a little.liked, shared as always, I learn something everything lol🙏
@@JoshuaBardwell hi jb just a update .on my new hd h7 kakute I had errors in the motors tab when trying to bidirectional dshot .after I flashed back to 32.8 no errors .maybe 4.3 when released it might work with the h7. Only sending this jb fpr feedback lol not to bother you .happy flying
Any BLHeli devs want to guess when 32.83 will be released?