New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

MATEK FC F405 AIO yaw twitch and drift #3912

Closed
Preathor opened this Issue Aug 21, 2017 · 27 comments

Comments

Projects
None yet
@Preathor

Preathor commented Aug 21, 2017

I'm using Matek FC F405 AIO and I'm experiencing some yaw twitches and uncontrolled yaw spins.
After reverting back to 3.1.7, the problem seems to go away. I've tried this at 8k/8k, 8k/4k, dshot600, antigravity at default values and airmode turned on. Everything with different standoffs, softmounting motors and softmounting pdb. Since reverting back to older version, I'm just thinking (but not claiming) that it must be something in new target version. I haven't done anything special with configuration either... everything is basically just "stock". I've also turned on active notch filter and turned off all filters and used pt1

My particular build is:

  • armattan chameleon 5''
  • armatan 30amp esc
  • emax 2305s red bottoms
  • matek fc f405 aio
  • frsky xsr
  • tbs unify race

I don't have any black box data, but this youtube example shows what I mean (not my video, but fairly similar issue): https://youtu.be/tSvOhG7PK5M?t=63
I've also experienced slow drifts in yaw direction, when I had to compensate with my sticks to go forward normally.

Problem is visible on video at 1:06 and I hope I'm not violating any licences by posting this video here.

Best regards,
Link

@tigert

This comment has been minimized.

Show comment
Hide comment
@tigert

tigert Aug 22, 2017

This is a shot in the dark, but I had this also on 3.2RC3 on a "ISHAPEDF3" (A Banggood clone of "Lulz F3" board they sold a while ago).

It would also show as yaw twitches in the 3D quad preview in the initial configurator screen.

I am not sure what it was, but I reflashed it with 3.2RC3 and 3.2.1 configurator and rebuilt my config from scratch and it seems to have gone away now. I assumed it might have been a damaged gyro chip, but seems like it is stable now, and the cause could be something in 3.2 or perhaps some combination of configuration options that made things behave weirdly.

tigert commented Aug 22, 2017

This is a shot in the dark, but I had this also on 3.2RC3 on a "ISHAPEDF3" (A Banggood clone of "Lulz F3" board they sold a while ago).

It would also show as yaw twitches in the 3D quad preview in the initial configurator screen.

I am not sure what it was, but I reflashed it with 3.2RC3 and 3.2.1 configurator and rebuilt my config from scratch and it seems to have gone away now. I assumed it might have been a damaged gyro chip, but seems like it is stable now, and the cause could be something in 3.2 or perhaps some combination of configuration options that made things behave weirdly.

@xlla

This comment has been minimized.

Show comment
Hide comment
@xlla

xlla Aug 23, 2017

I am using that board too, matek f405 aio with bf 3.1.7 firmware.
It will yaw spin and yaw twitch when I quick raise throttle, or do a quick yaw spin.
It fly smooth on hover.

Another quad with same device fly fine. I think that was caused by a bad gyro on FC.

xlla commented Aug 23, 2017

I am using that board too, matek f405 aio with bf 3.1.7 firmware.
It will yaw spin and yaw twitch when I quick raise throttle, or do a quick yaw spin.
It fly smooth on hover.

Another quad with same device fly fine. I think that was caused by a bad gyro on FC.

@Preathor

This comment has been minimized.

Show comment
Hide comment
@Preathor

Preathor Aug 24, 2017

What's interesting to me is that in my case, it works fine on 3.1.7, smooth flight, no problems at all whatsoever, but when I upgraded it to 3.2 it started to show problems. When I reverted back... worked okay again.

I've been following the FB page for this particular board, and I'll try some of their "hacks"... moving from uart2 to uart5 (sbus and telemtry) then report back how it goes (next week)

Preathor commented Aug 24, 2017

What's interesting to me is that in my case, it works fine on 3.1.7, smooth flight, no problems at all whatsoever, but when I upgraded it to 3.2 it started to show problems. When I reverted back... worked okay again.

I've been following the FB page for this particular board, and I'll try some of their "hacks"... moving from uart2 to uart5 (sbus and telemtry) then report back how it goes (next week)

@rossonado

This comment has been minimized.

Show comment
Hide comment
@rossonado

rossonado Aug 26, 2017

I have the same issue. New build 3.2. Quad is OK, but if punch out and throttle back, quad drifts into a spin on its own. Same at most points when inverte. When flying normally it's relatively OK. I've upped timing, raised, p, I and antigravity and messed with filters and nothing changes It. I've changed escs with no difference. It feels like airmode is not on. System is clean, cool motors etc. Rebuilt quad to add more softmounting. Will try 3.17 and see if it helps. Seems like this is an amazing fc if you're lucky and awful if not. Regroups is full of it.

rossonado commented Aug 26, 2017

I have the same issue. New build 3.2. Quad is OK, but if punch out and throttle back, quad drifts into a spin on its own. Same at most points when inverte. When flying normally it's relatively OK. I've upped timing, raised, p, I and antigravity and messed with filters and nothing changes It. I've changed escs with no difference. It feels like airmode is not on. System is clean, cool motors etc. Rebuilt quad to add more softmounting. Will try 3.17 and see if it helps. Seems like this is an amazing fc if you're lucky and awful if not. Regroups is full of it.

@robert-hoffmann

This comment has been minimized.

Show comment
Hide comment
@robert-hoffmann

robert-hoffmann Aug 28, 2017

I had some slight yaw twitches when i put in a (very) tired battery, put in ta new one and twitches went away ...probably not related though

robert-hoffmann commented Aug 28, 2017

I had some slight yaw twitches when i put in a (very) tired battery, put in ta new one and twitches went away ...probably not related though

@Peter-JanGootzen

This comment has been minimized.

Show comment
Hide comment
@Peter-JanGootzen

Peter-JanGootzen Aug 28, 2017

I have flown 15 batteries with my new build with the Matek F405-AIO. I have had zero issues whilst running 8k/8k. No twitching and very little noise. I have flown with 3.1.7 and 3.2 RC4, both been great.

But, when I try to run 16k/16k my motors don't arm at the same time, the motors also sound way different and when trying to take off it just tumbles into the ground. Very weird stuff. I think this is some weird configuration issue, I am running DShot600. This issue is happening on both 3.1.7 and 3.2 RC4.
If someone wants to help me trouble shoot, I can post all my BLHeli and BF relevant settings.

@Preathor This sounds a lot like you do not have airmode turned on. If you are currently using airmode as a feature, you might wanna try using it as a mode. There have been people who have been having problems with that before.
https://www.youtube.com/watch?v=6GcsRXr9gzQ
https://www.youtube.com/watch?v=EQ-qxBsMl4w&t=333s

Peter-JanGootzen commented Aug 28, 2017

I have flown 15 batteries with my new build with the Matek F405-AIO. I have had zero issues whilst running 8k/8k. No twitching and very little noise. I have flown with 3.1.7 and 3.2 RC4, both been great.

But, when I try to run 16k/16k my motors don't arm at the same time, the motors also sound way different and when trying to take off it just tumbles into the ground. Very weird stuff. I think this is some weird configuration issue, I am running DShot600. This issue is happening on both 3.1.7 and 3.2 RC4.
If someone wants to help me trouble shoot, I can post all my BLHeli and BF relevant settings.

@Preathor This sounds a lot like you do not have airmode turned on. If you are currently using airmode as a feature, you might wanna try using it as a mode. There have been people who have been having problems with that before.
https://www.youtube.com/watch?v=6GcsRXr9gzQ
https://www.youtube.com/watch?v=EQ-qxBsMl4w&t=333s

@Preathor

This comment has been minimized.

Show comment
Hide comment
@Preathor

Preathor Aug 29, 2017

I have airmode turned on correctly. I don't think this is the problem.
@Peter-JanGootzen, how did you connect your RX and which one are you using? Did you use diagrams from Matek site?

I've been seeing more and more people posting videos of this twitching on Matek FB page. It's a shame, because it's a nice board, and so many people reporting problems with it.

Preathor commented Aug 29, 2017

I have airmode turned on correctly. I don't think this is the problem.
@Peter-JanGootzen, how did you connect your RX and which one are you using? Did you use diagrams from Matek site?

I've been seeing more and more people posting videos of this twitching on Matek FB page. It's a shame, because it's a nice board, and so many people reporting problems with it.

@robert-hoffmann

This comment has been minimized.

Show comment
Hide comment
@robert-hoffmann

robert-hoffmann Aug 29, 2017

@Preathor I used the exact setups (except voltage pads) as shown on matek site with: f405 AIO (softmounted + furnished cap), tbs unify v2, e-xsr (smartport on softserial), blheli_32, emax 2306, eagle 2, chameleon frame

With rx on 4.5v, vtx on 5v, and cam on 9v

It flew great out of the box, but since i can't leave stuff alone, i'm now on dshot 600, 8k/4k gyro/pid loop, no notch filters at all, but both gyro/dterm filters are set to biquad, with dynamic filters on. Airmode & Anti-Gravity as features on, changed dterm weight/threshhold, vbat compensation on, and messed around with rc interpolation (and forced my taranis to 8ch), but pids a bit lower than the defaults ..i think like 15 on D right now.

The only instance i got some twitches was when i put a bad battery in that was rapidly dropping voltage anytime it goes above 20-30amps.

robert-hoffmann commented Aug 29, 2017

@Preathor I used the exact setups (except voltage pads) as shown on matek site with: f405 AIO (softmounted + furnished cap), tbs unify v2, e-xsr (smartport on softserial), blheli_32, emax 2306, eagle 2, chameleon frame

With rx on 4.5v, vtx on 5v, and cam on 9v

It flew great out of the box, but since i can't leave stuff alone, i'm now on dshot 600, 8k/4k gyro/pid loop, no notch filters at all, but both gyro/dterm filters are set to biquad, with dynamic filters on. Airmode & Anti-Gravity as features on, changed dterm weight/threshhold, vbat compensation on, and messed around with rc interpolation (and forced my taranis to 8ch), but pids a bit lower than the defaults ..i think like 15 on D right now.

The only instance i got some twitches was when i put a bad battery in that was rapidly dropping voltage anytime it goes above 20-30amps.

@Peter-JanGootzen

This comment has been minimized.

Show comment
Hide comment
@Peter-JanGootzen

Peter-JanGootzen Aug 29, 2017

@Preathor
I am using the Flysky X6B and it is connected to uart2.
Yes, it really is a shame. This is an amazing FC, when it works...

Peter-JanGootzen commented Aug 29, 2017

@Preathor
I am using the Flysky X6B and it is connected to uart2.
Yes, it really is a shame. This is an amazing FC, when it works...

@TimothyGold

This comment has been minimized.

Show comment
Hide comment
@TimothyGold

TimothyGold Aug 30, 2017

I also have used the FC in a new build and have struggled with this form the beginning. No matter what I try I could not get the yaw issue under control. I originally built with the black included vibration dampers and to be honest I never saw any vibration issues in Black Box that seemed too severe. In fact Pitch and Roll was very good. Yaw seemed to be worse than pitch and roll but didn't look worse than other examples i have seen. I eventually lost control of the copter and could not recover and now have to rebuild. In my rebuild I was going to try softer isolation to see if something would change. I am starting to wonder if I just need to settle on a different FC maybe using a more reliable IMU not the ICM20602 that seems to be plagued with issues in 3.2.

TimothyGold commented Aug 30, 2017

I also have used the FC in a new build and have struggled with this form the beginning. No matter what I try I could not get the yaw issue under control. I originally built with the black included vibration dampers and to be honest I never saw any vibration issues in Black Box that seemed too severe. In fact Pitch and Roll was very good. Yaw seemed to be worse than pitch and roll but didn't look worse than other examples i have seen. I eventually lost control of the copter and could not recover and now have to rebuild. In my rebuild I was going to try softer isolation to see if something would change. I am starting to wonder if I just need to settle on a different FC maybe using a more reliable IMU not the ICM20602 that seems to be plagued with issues in 3.2.

@s700ky

This comment has been minimized.

Show comment
Hide comment
@s700ky

s700ky Aug 30, 2017

Put my 405 AIO into a Crusader 175.
Violent yaw twitches.
All filters on and BiQuad set and it calms down a little, but still unflyable.
Reflashed Board and ESC’s still the same.
Tried all different ESC settings still the same.
Running 16K 16K still the same.
Need to try 8K 8K as suggested above.
Do not have this problem on the Matek board with separate PDB.

s700ky commented Aug 30, 2017

Put my 405 AIO into a Crusader 175.
Violent yaw twitches.
All filters on and BiQuad set and it calms down a little, but still unflyable.
Reflashed Board and ESC’s still the same.
Tried all different ESC settings still the same.
Running 16K 16K still the same.
Need to try 8K 8K as suggested above.
Do not have this problem on the Matek board with separate PDB.

@s700ky

This comment has been minimized.

Show comment
Hide comment
@s700ky

s700ky Aug 30, 2017

Also even though I installed the board with the arrow facing forward I had to yaw the board 270 to get its orientation right.

s700ky commented Aug 30, 2017

Also even though I installed the board with the arrow facing forward I had to yaw the board 270 to get its orientation right.

@rossonado

This comment has been minimized.

Show comment
Hide comment
@rossonado

rossonado Aug 30, 2017

rossonado commented Aug 30, 2017

@Preathor

This comment has been minimized.

Show comment
Hide comment
@Preathor

Preathor Aug 31, 2017

So.. I did some testing. I don't know if it's my particular board that just might be bad, or is this a general thing for this board. I'm still reading about same problem on FB page. Some have managed to solve it (with some sort of luck and leprechaun magic), but I'm not actualy sure....

I've replaced the reciever with X4R-sb (instead of XSR)... resoldered wires on it so that i have uninverted sbus and smartport (telemetry) using Project blue falcon and Joshua Bardwells instructions. That gave me the option to move these two things to a different uart. I've disabled soft serial, disabled blackbox, i've been using 8k/8k, I've enabled airmode as a feature (in configuration tab, so that it's always on), I've disabled accelerometer, tried different tpa values and antigravity values... pt1 without any notch filters, enabled new active notch filter, tried different soft mounts... I've replaced my emax motors with mr. steele silk motors... the white ones... (because emax ones can be a little noisy) did new pid tunnings... aaaaaand...

... same deal :(

Thing flies okay for half a pack... but one I do a couple of hard manevers, it starts to drift to the right... I always have to compensate with my yaw input... this does for about 30 seconds (slow drift)... then it "let's go", makes a twitch... and it's back to semi normal. Until i do another manever.. split-s for example...then it's starts to drift again.

I don't have any black box logs because I've been having some troubles with sd-cards... and I don't have any high-res/low-res footage of it. I'm not putting my gopro on that thing.

I'll be reverting to BF 3.1.7 tomorrov and see what happens again.

I don't mind staying on 3.1.7, but I was liking that active notch filter thing... Shame really. Might aswell switch back to my Flip32, or maybe buy Kakute... If weather won't allow it, I'll start doing Matek diffs on 3.1.7 version with 3.2 :D

Anywho... still best kuddos to BF developers. Keep on doing the great work.

Best regards,
Link

Preathor commented Aug 31, 2017

So.. I did some testing. I don't know if it's my particular board that just might be bad, or is this a general thing for this board. I'm still reading about same problem on FB page. Some have managed to solve it (with some sort of luck and leprechaun magic), but I'm not actualy sure....

I've replaced the reciever with X4R-sb (instead of XSR)... resoldered wires on it so that i have uninverted sbus and smartport (telemetry) using Project blue falcon and Joshua Bardwells instructions. That gave me the option to move these two things to a different uart. I've disabled soft serial, disabled blackbox, i've been using 8k/8k, I've enabled airmode as a feature (in configuration tab, so that it's always on), I've disabled accelerometer, tried different tpa values and antigravity values... pt1 without any notch filters, enabled new active notch filter, tried different soft mounts... I've replaced my emax motors with mr. steele silk motors... the white ones... (because emax ones can be a little noisy) did new pid tunnings... aaaaaand...

... same deal :(

Thing flies okay for half a pack... but one I do a couple of hard manevers, it starts to drift to the right... I always have to compensate with my yaw input... this does for about 30 seconds (slow drift)... then it "let's go", makes a twitch... and it's back to semi normal. Until i do another manever.. split-s for example...then it's starts to drift again.

I don't have any black box logs because I've been having some troubles with sd-cards... and I don't have any high-res/low-res footage of it. I'm not putting my gopro on that thing.

I'll be reverting to BF 3.1.7 tomorrov and see what happens again.

I don't mind staying on 3.1.7, but I was liking that active notch filter thing... Shame really. Might aswell switch back to my Flip32, or maybe buy Kakute... If weather won't allow it, I'll start doing Matek diffs on 3.1.7 version with 3.2 :D

Anywho... still best kuddos to BF developers. Keep on doing the great work.

Best regards,
Link

@robert-hoffmann

This comment has been minimized.

Show comment
Hide comment
@robert-hoffmann

robert-hoffmann Sep 1, 2017

Still no twitches here.

Did have motors a bit warm, so bumped down gyro/dterm soft lpf filtes by 10 points each. Now they are barely warm.

Also soft mounted the motors (fc was already soft mounted), and seems to fly much smoother. I already went with PT1 on gyro when i bumped the lpf down, and tomorrow i'll see if i can go back to PT1 on dterm too.

Still running 8k/4k gyro/loop on dshot 600

robert-hoffmann commented Sep 1, 2017

Still no twitches here.

Did have motors a bit warm, so bumped down gyro/dterm soft lpf filtes by 10 points each. Now they are barely warm.

Also soft mounted the motors (fc was already soft mounted), and seems to fly much smoother. I already went with PT1 on gyro when i bumped the lpf down, and tomorrow i'll see if i can go back to PT1 on dterm too.

Still running 8k/4k gyro/loop on dshot 600

@1337flite

This comment has been minimized.

Show comment
Hide comment
@1337flite

1337flite Sep 4, 2017

Works for me.

My config is nearly the same as Preathor other than Wraith 32 35A ESCs and Red bottom 2205S and I have no problems.
I have 220uf 25 volt caps on my ESCs and the Matek provided cap on the battery leads and its flys beautifully on 3.2 and has been since RC2.

I have a similar set up but with the Matek F4 and PDB and that flies just as well on 3.2 since before the release candidates. That's a 6" Cham and has different motors (BoltRC 2207 2450s).

Motors on both are very soft mounted and FC uses matek supplied rubber mounts.

1337flite commented Sep 4, 2017

Works for me.

My config is nearly the same as Preathor other than Wraith 32 35A ESCs and Red bottom 2205S and I have no problems.
I have 220uf 25 volt caps on my ESCs and the Matek provided cap on the battery leads and its flys beautifully on 3.2 and has been since RC2.

I have a similar set up but with the Matek F4 and PDB and that flies just as well on 3.2 since before the release candidates. That's a 6" Cham and has different motors (BoltRC 2207 2450s).

Motors on both are very soft mounted and FC uses matek supplied rubber mounts.

@robert-hoffmann

This comment has been minimized.

Show comment
Hide comment
@robert-hoffmann

robert-hoffmann Sep 5, 2017

Check this out guys
https://youtu.be/uINN6AXVhFE

I didn't have any problems to begin with... But settling motor pwm to 48khz (if your on blheli32) does bring the global noise envelope down.

Now on zero notches (except dynamic) gyro/pid 8k/4k with both on PT1, and motors barely warm. Motors/FC soft mounted, and 470 uf cap on battery.

Gonna try 8k/8k and dshot 1200 and see if I can bump the lpf back up some soon.

robert-hoffmann commented Sep 5, 2017

Check this out guys
https://youtu.be/uINN6AXVhFE

I didn't have any problems to begin with... But settling motor pwm to 48khz (if your on blheli32) does bring the global noise envelope down.

Now on zero notches (except dynamic) gyro/pid 8k/4k with both on PT1, and motors barely warm. Motors/FC soft mounted, and 470 uf cap on battery.

Gonna try 8k/8k and dshot 1200 and see if I can bump the lpf back up some soon.

@chriswils

This comment has been minimized.

Show comment
Hide comment
@chriswils

chriswils Sep 5, 2017

Here is my blacbox https://drive.google.com/open?id=0B9Lnsj9U0mK8TkYyRmRpUEJnYmM with bf 3.1.7.

A lot of twitching and sometimes yaw spin. 8k/4k. 3% CPU load in idle.
RCX 2206 2450kV
Racerstar 30A LiteS
DSHOT 600

I have the exact same setup, but with Emax RS2205 2300kV, and no such problems. Also, I saw that my CPU load spiked to 100% when acc was enabled. tasks showed 32%.
When going in to turns, the PID's got really aggressive. This is default PID's.

Edit: Turned of PT1 (which was by default) and set BIQUAD. Didn't do anything for me. Turned of everything I could think of.
Blackbox with videos: https://drive.google.com/drive/folders/0B9Lnsj9U0mK8eDNhSmR0XzVxSkk after capping all ESCs with Panasonic FM 340uF 25V.

EDIT and solved: My bad (embarrassed); one of the stack screws was just a tad bit loose. Be sure to tighten things properly =)

chriswils commented Sep 5, 2017

Here is my blacbox https://drive.google.com/open?id=0B9Lnsj9U0mK8TkYyRmRpUEJnYmM with bf 3.1.7.

A lot of twitching and sometimes yaw spin. 8k/4k. 3% CPU load in idle.
RCX 2206 2450kV
Racerstar 30A LiteS
DSHOT 600

I have the exact same setup, but with Emax RS2205 2300kV, and no such problems. Also, I saw that my CPU load spiked to 100% when acc was enabled. tasks showed 32%.
When going in to turns, the PID's got really aggressive. This is default PID's.

Edit: Turned of PT1 (which was by default) and set BIQUAD. Didn't do anything for me. Turned of everything I could think of.
Blackbox with videos: https://drive.google.com/drive/folders/0B9Lnsj9U0mK8eDNhSmR0XzVxSkk after capping all ESCs with Panasonic FM 340uF 25V.

EDIT and solved: My bad (embarrassed); one of the stack screws was just a tad bit loose. Be sure to tighten things properly =)

@Hustleflyer

This comment has been minimized.

Show comment
Hide comment
@Hustleflyer

Hustleflyer Oct 4, 2017

I have gone through 3 different builds all with Matek F405 AIOs, and two out of three experienced this exact problem, mine was much worse with one I was debugging and troubleshooted all the way to the Gyro itself after rebuilding same machine three times, an Arris X220 with two different sets of motors and then different ESCs just to be extra double sure. I tried to band aid with a rubicon capacitor and it did seem to help a little, only other solution is a one capacitor per ESC. The other two had less noticeable gyro issues and more signal interference only when throttle was applied, signal noise would appear and I could not manage to make it disapper, no matter how I wired the power to cam Did not have to take logs to confirm the basic fact the gyro on these boards, though fast, has many noise issues coming from nearby emi -electro-mag interference.

Hustleflyer commented Oct 4, 2017

I have gone through 3 different builds all with Matek F405 AIOs, and two out of three experienced this exact problem, mine was much worse with one I was debugging and troubleshooted all the way to the Gyro itself after rebuilding same machine three times, an Arris X220 with two different sets of motors and then different ESCs just to be extra double sure. I tried to band aid with a rubicon capacitor and it did seem to help a little, only other solution is a one capacitor per ESC. The other two had less noticeable gyro issues and more signal interference only when throttle was applied, signal noise would appear and I could not manage to make it disapper, no matter how I wired the power to cam Did not have to take logs to confirm the basic fact the gyro on these boards, though fast, has many noise issues coming from nearby emi -electro-mag interference.

@jalao

This comment has been minimized.

Show comment
Hide comment
@jalao

jalao Oct 20, 2017

Please please help. . Is there going to be a solution in sight for the agreesive Yaw twitch of this board? If I slowly apply throttle while making turns, the quad flies well, but the moment you make aggressive turns it will twitch hard and it some instances a complete spiral crash. I have soft mounted the board. Soft mounted the motors in two places. Directly under the motor and at the screws that hold the motors. I have used the capacitor provided with the board. I am running the Matek AIO board, new f60 tiger motors , wraith32 escS. I am running 8k/8k. Running airmode, dynamic notch with pt1 enabled.

jalao commented Oct 20, 2017

Please please help. . Is there going to be a solution in sight for the agreesive Yaw twitch of this board? If I slowly apply throttle while making turns, the quad flies well, but the moment you make aggressive turns it will twitch hard and it some instances a complete spiral crash. I have soft mounted the board. Soft mounted the motors in two places. Directly under the motor and at the screws that hold the motors. I have used the capacitor provided with the board. I am running the Matek AIO board, new f60 tiger motors , wraith32 escS. I am running 8k/8k. Running airmode, dynamic notch with pt1 enabled.

@JMcDee

This comment has been minimized.

Show comment
Hide comment
@JMcDee

JMcDee Oct 22, 2017

I'm having the same issue, running 3.2.1 firmware. Violent yaw twitches, roll and pitch are perfect.

Frame : RED RD-025 TALON SVX 220 low side plates
FC: Matek F405 AIO
ESC: Racerstar Tattoo_S 35A 4in1 2-4S STM32F051/ARM Blheli_32 Dshot1200 Ready Dual BEC FPV Racing ESC
VTX: TBS Unify pro 5G8 V2 (SMA)
RC RX: TBS Crossfire Micro RX V2
MOTORS: EFAW 2407R 2500KV
CAM: Runcam Swift 2

JMcDee commented Oct 22, 2017

I'm having the same issue, running 3.2.1 firmware. Violent yaw twitches, roll and pitch are perfect.

Frame : RED RD-025 TALON SVX 220 low side plates
FC: Matek F405 AIO
ESC: Racerstar Tattoo_S 35A 4in1 2-4S STM32F051/ARM Blheli_32 Dshot1200 Ready Dual BEC FPV Racing ESC
VTX: TBS Unify pro 5G8 V2 (SMA)
RC RX: TBS Crossfire Micro RX V2
MOTORS: EFAW 2407R 2500KV
CAM: Runcam Swift 2

@rossonado

This comment has been minimized.

Show comment
Hide comment
@rossonado

rossonado Oct 22, 2017

rossonado commented Oct 22, 2017

@DimaVIII

This comment has been minimized.

Show comment
Hide comment
@DimaVIII

DimaVIII Oct 23, 2017

Hey Guys,
To solve this problem you need to place a low ESR Cap on each ESC.
They include now 4 Caps in the package inserted of one, if you are lucky to get the new batch.
Check out this: https://www.rcgroups.com/forums/showthread.php?2912273-Matek-Flight-Controller-F405-AIO/page33

DimaVIII commented Oct 23, 2017

Hey Guys,
To solve this problem you need to place a low ESR Cap on each ESC.
They include now 4 Caps in the package inserted of one, if you are lucky to get the new batch.
Check out this: https://www.rcgroups.com/forums/showthread.php?2912273-Matek-Flight-Controller-F405-AIO/page33

@Fuzzball27

This comment has been minimized.

Show comment
Hide comment
@Fuzzball27

Fuzzball27 Nov 28, 2017

Is this issue solved?
Is this fixed by placing caps on the ESCs, or by reverting firmware?

Fuzzball27 commented Nov 28, 2017

Is this issue solved?
Is this fixed by placing caps on the ESCs, or by reverting firmware?

@jalao

This comment has been minimized.

Show comment
Hide comment
@jalao

jalao Nov 28, 2017

jalao commented Nov 28, 2017

@martinbudden

This comment has been minimized.

Show comment
Hide comment
@martinbudden

martinbudden Nov 28, 2017

Contributor

Closing, it's a hardware issue.

Contributor

martinbudden commented Nov 28, 2017

Closing, it's a hardware issue.

@peterhk80

This comment has been minimized.

Show comment
Hide comment
@peterhk80

peterhk80 Nov 28, 2017

I had the same issues but solved them by turning OFF the dynamic filter and anti-gravity features. Once those were off - never had an issue and love the FC... l do not have any capacitors either (only had 1 bc it’s the older version)

IMO - using the same looptime for gyro/pid is not that beneficial and use a 4:1 ratio.

The F405 can go to 32k and I run 32/8k (NO antigravity or dynamic filter - that adds 10+% to the CPU)... 16/16 with all the features on will be too much.

Regardless of the looptime, try turning off the 2 features and if needed, lower your lpf dyerm and gyro down in increments of 10 (from default - 100, go to 90) and you should see improvements

peterhk80 commented Nov 28, 2017

I had the same issues but solved them by turning OFF the dynamic filter and anti-gravity features. Once those were off - never had an issue and love the FC... l do not have any capacitors either (only had 1 bc it’s the older version)

IMO - using the same looptime for gyro/pid is not that beneficial and use a 4:1 ratio.

The F405 can go to 32k and I run 32/8k (NO antigravity or dynamic filter - that adds 10+% to the CPU)... 16/16 with all the features on will be too much.

Regardless of the looptime, try turning off the 2 features and if needed, lower your lpf dyerm and gyro down in increments of 10 (from default - 100, go to 90) and you should see improvements

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment