KISS FC v1.03 + KISS ESC 32a - Motors out of control

Raste

New member
Hi,

I am writing here because I am out of ideas on how to solve this issue. I built my first quad with the following hardware:

Chameleon 5 + mini pdb v3
KISS FC V1, RC34
KISS ESC 32a racing, 1.20x voice stable
Emax RS2306 white edition

When I went for my first flight, I managed to lift up and do a bit of hovering, and when starting to put a bit more throttle it did flips of death and crashed. This behaviour is systematic, it happens everytime (check around 0min50s):


As you can see it flies perfectly well, so props are not inversed, motors spin in the correct direction, FC is oriented correctly. After seing this, I added anti vibration pad under the motors and put a capacity of 1000uFahrad on the battery line, as I thought either vibrations or spikes of current would trigger this. NO LUCK!

To go one step further, I flashed the KISS FC with betaflight, and not the KISS firmare. Turns out the hover lasted longer before my quad, nevertheless, ended up in uncontrollable flips of death.

At this point I ordered a new FC, installed it and went to fly today. You guessed it right, still no luck... The very same behaviour happens over and over. I have removed the props, plugged into the KISS GUI, and when I just put it on my hand and arm, the motors spin correctly. As soon as I go closer to mid throttle, all the motors just go crazy. Here is the video showing the data of the gyro, and the rpm on the motors.


When doing this test, I had the same behaviour as shown here: https://www.youtube.com/watch?v=7H-xKuFlxOQ When I use the KISS GUI to test the motors, one by one, they all work perfectly. Even if I try to test all 4 at the same time, perfect response to the throttle stick.

I have absolutely no idea what causes this, and since this is my first quad, I am starting to think about throwing it away and buying something ready out of the box. I spent over 600 bucks to get good hardware and this thing is just getting worse and worse.

I looked all around the web and found nothing. Please help, I'm going crasy here...
 
Last edited:

PsyBorg

Wake up! Time to fly!
I see you have the new 32 amp esc's not the 24 amp ones. I also see you are using a v1 FC not the V2. I think what may be going on here is you have chosen a dshot over the dshot 600 capabilities of the V1 FC.

I also notice that you have not done proper channel set up and got your channel centers to 1500. that little bit of offset on all three axis can add up to a problem at higher throttle as the Fc is trying to balance itself using parameters that are not level with the attitude of the actual craft.

I would go back into set up and make sure you are not running dshot above 600 ( I can not see in the video you made as that part is too far out of focus for me to read) and do propser set up for channel settings including centers and end points. Then after that redo throttle calibration and try that. Hopefully that solves the issue if not get back to us with any changes that have happened. Oh also make sure to set fail safe. That looked like it went into full throttle bye bye mode and luckily the FC shut itself down before it got away.

Edit: the video you linked with the guy on the test bench that showed the quad rocking back and forth is either wrong motor orientation or Fc is not facing forward and was not adjusted for in the gui.
 
Last edited:

Raste

New member
I see you have the new 32 amp esc's not the 24 amp ones. I also see you are using a v1 FC not the V2. I think what may be going on here is you have chosen a dshot over the dshot 600 capabilities of the V1 FC.

Hi,

Thank you for your answer. I have checked and I am running D-Shot600, and cannot go higher with the FC since the v1 does not support more than that. The ESCs confirm with voice that they run on D-Shot600 when I plug the battery.

I also notice that you have not done proper channel set up and got your channel centers to 1500. that little bit of offset on all three axis can add up to a problem at higher throttle as the Fc is trying to balance itself using parameters that are not level with the attitude of the actual craft.

What do you mean by proper channel set up ? My end points are 1000 and 2000 stable on all sticks. Could you detail a bit more ?

Oh also make sure to set fail safe. That looked like it went into full throttle bye bye mode and luckily the FC shut itself down before it got away.

Good point! Will do.

Thanks a bunch for the pointers. I will try that and report the results here
 

PsyBorg

Wake up! Time to fly!
In the video your sticks sit at 1494 ish at rest. That will cause drift once in a stable hover. Not sure how far out that can be before the fC fights the gyros thinking that is neutral above hover rpm.

After thinking about this I had the same issue with TAZ (my big build) I had rotated the FC so I could access the usb at the rear of the quad due to wire harnesses for the VTX in the way. I had forgot to tell the gui I had rotated it and got a very similar reaction that the end of your video showed. The only difference is that I could not get it to do anything but roll on the ground at even the slightest bit of throttle. It took me a good three days of scratchin my head till I remembered that little detail.

So physically check and recheck your motor orientation matches your FC orientation.. IE fc facing forward and motor order in proper places. OR if you have rotated the FC you still have the original wiring order AND changed the orientation in the gui in the proper direction. That makes a huge difference if you went positive and should have gone negative.(It kind of looks like the FC orientation numbers are not at 0 in the video but too blurry to see for sure.) make sure all of that is matched properly. Hopefully you have not tried a different wiring order and rotational changes in the gui as that would get really confusing.
 

Raste

New member
In the video your sticks sit at 1494 ish at rest. That will cause drift once in a stable hover. Not sure how far out that can be before the fC fights the gyros thinking that is neutral above hover rpm.

After thinking about this I had the same issue with TAZ (my big build) I had rotated the FC so I could access the usb at the rear of the quad due to wire harnesses for the VTX in the way. I had forgot to tell the gui I had rotated it and got a very similar reaction that the end of your video showed. The only difference is that I could not get it to do anything but roll on the ground at even the slightest bit of throttle. It took me a good three days of scratchin my head till I remembered that little detail.

So physically check and recheck your motor orientation matches your FC orientation.. IE fc facing forward and motor order in proper places. OR if you have rotated the FC you still have the original wiring order AND changed the orientation in the gui in the proper direction. That makes a huge difference if you went positive and should have gone negative.(It kind of looks like the FC orientation numbers are not at 0 in the video but too blurry to see for sure.) make sure all of that is matched properly. Hopefully you have not tried a different wiring order and rotational changes in the gui as that would get really confusing.

I checked the sticks middle position and did a new calibration on the Taranis. I ended up giving two of trims on each axis.

My FC has the arrow pointing to the right when you look at the drone from above. So I put 0, 0, 90 in the GUI. I checked all the controls and it works perfectly, the quad goes in the direction I ask it to go to, otherwise I would have had a bad time trying to keep it sable while hovering. The motors are in the right order and when I test them out they spin in the correct direction. I followed the original wiring of the ESCs on the FC, so that should be good.

I will go for a test later today, will report what happens! Again, thx for the pointers and advice
 

Raste

New member
Just back from the test flight, and I am happy to report that for two full batteries I had no flip of death happening. The quad is responding as it should, I can fly with it.

However, I noticed a very weird noise while flying, just like it would actually have something resonating very strongly. I tried to change the props, and the noise was still there. When arming and very little throttle is applied, it does not appear. When I get to hover, it starts. I have started the quad with the propellers on, armed it while holding it from the battery just 10cm above the ground, and started playing with the throttle. I noticed that when this noise happens, at least 2 motors are hesitating... they are not running smoothly, as if they were unsure at which speed to go and would constantly oscillate between two speeds. When this happens in flight, I notice the quad has small twitches happening, and it directly recovers from it.

I have installed the 1.20x beta FW for the ESCs yesterday as I wanted to get the latest version in case something got fixed, but I also noticed it with 1.20l stable. On the maiden flight I did not had that, and the only things I changed are (1) it crashed into some dirt, I tried to clean up the motors as best as I could and (2) I added anti-vibration pads under the motors.

Any idea ?

Edit (I will try again but I will remove the dampeners under the motors to see if that causes the issue) - EDIT: The anti-vibration pads are not the cause, it is the behaviour of the motors that make the frame resonate. Could the PID do that ?
 
Last edited:

PsyBorg

Wake up! Time to fly!
Glad to hear you have it in the air mate.

The sound you are getting is like a grindy, chattery type sound like there may be dirt in the motors kind of? Listen to the maiden of my latest build from lift off to like 1:40. Probably have to use headphones to hear it but I am reasonably sure that is the sound you hear. This is in fact due to pids. It can also make a motor desync and fall from the sky which is what may have happened. Raise pitch and roll P .50 and that should change it enough to tell if this is infact the issue. Don't fly too long before checking motor temps. I do the same as you can see in the maiden. Flew for a min or so then landed to see how motors felt and again after the flight.

 

Raste

New member
Glad to hear you have it in the air mate.

The sound you are getting is like a grindy, chattery type sound like there may be dirt in the motors kind of? Listen to the maiden of my latest build from lift off to like 1:40. Probably have to use headphones to hear it but I am reasonably sure that is the sound you hear. This is in fact due to pids. It can also make a motor desync and fall from the sky which is what may have happened. Raise pitch and roll P .50 and that should change it enough to tell if this is infact the issue. Don't fly too long before checking motor temps. I do the same as you can see in the maiden. Flew for a min or so then landed to see how motors felt and again after the flight.

I tried changing the props, removing the anti vibration pads under the motors, different PID settings, and I still get those vibrations from 35% to roughly 60% throttle. Any clue ? Here is a video, you can really hear the sound it makes, and see the twitches these vibrations are causing:

 

PsyBorg

Wake up! Time to fly!
That is the same noise I got from Taz on the maiden. Mine was not as strong and predominant as that though. Could you screen shot the pids you used when you made that video for us? Sounds like it is on the pitch axis as if changes for a brief instant when you change direction forwards and back in that video. Slowly raise Pitch P up a little more. Stock pids are too low for most quads so there is a ways to go before you hit higher P oscillations.

Watch these two videos and see if tuning clicks for you a bit more. the skitzo one is what made it work for me. Stingy has a slightly different method that works too.


 

Raste

New member
I will watch the videos, thx. I have experienced these noises with both the default PIDs (on the video) as well as the ones from Mr Steele (wanted to try other PIDs, even if not fitted to my drone, to see if that was causing it). I will try to fiddle with the PIDs and work a P that works on all the axis. Hopefully that solves the issue.

I really don't see what else could cause this, but again I am starting in this field so I don't have that much experience. For that it does not help to have 12+ years of flying experience with planes sigh
 
Last edited:

Raste

New member
That is the same noise I got from Taz on the maiden. Mine was not as strong and predominant as that though. Could you screen shot the pids you used when you made that video for us? Sounds like it is on the pitch axis as if changes for a brief instant when you change direction forwards and back in that video. Slowly raise Pitch P up a little more. Stock pids are too low for most quads so there is a ways to go before you hit higher P oscillations.

I tried raising the P, no luck, still the same vibrations. It somewhat got worse as now from barely lifting off to almost half throttle I get terrible loss of power, I have to go above half throttle to get the quad to gain altitude. Just my raising 0.5 on P roll and pitch.

I will go back to the stable firmware of the ESCs, and try one more time with brand new propellers. If the problem is still there after that I will contact directly flyduino... I'm out of options here
 

PsyBorg

Wake up! Time to fly!
I was flying RC 28 on my Alien and it was super smooth til I wore out the bearings on my motors..I replaced the motors and did a complete rebuild to clean it all up as well as swap out the PDB. While I was doing all that I flashed the firmware to 1.2 from 1.14 and RC9 for the esc's and I am now having troubles getting mine to tune properly as well.


I think I may have found the issue... changes from 2.0.5 to 2.0.6 fixes a bug where the output, TPA and rates tab dont worked with FC FW versions before 1.3RC34 is the release notes for gui verion 2.06 found here... I am using the original 1.15.6 standalone gui from the bundle. The firmware for the FC and the ESC seem to be the current good release.

https://github.com/flyduino/kiss-gui/releases

I am going to update to that and see if the pid changes I do after actually change something.
 

Raste

New member
I was flying RC 28 on my Alien and it was super smooth til I wore out the bearings on my motors..I replaced the motors and did a complete rebuild to clean it all up as well as swap out the PDB. While I was doing all that I flashed the firmware to 1.2 from 1.14 and RC9 for the esc's and I am now having troubles getting mine to tune properly as well.


I think I may have found the issue... changes from 2.0.5 to 2.0.6 fixes a bug where the output, TPA and rates tab dont worked with FC FW versions before 1.3RC34 is the release notes for gui verion 2.06 found here... I am using the original 1.15.6 standalone gui from the bundle. The firmware for the FC and the ESC seem to be the current good release.

https://github.com/flyduino/kiss-gui/releases

I am going to update to that and see if the pid changes I do after actually change something.

I am using RC34 on the FC, GUI standalone v2.0.6. My motors are brand new... they just got some dirt in there once from one of those flips of death. I cleaned it up, it runs smoothly when I do test motors from the GUI.

I however noticed that when you flash the firmware of the ESCs, and download the firmware remotely via the KISS GUI, they used to have 1.20l, 1.20x, and a beta version. Now the 1.20x has become the beta version. So I might have been flying with a beta version pretty much all along without even knowing it because Flyduino failed to name them properly.

Let me know how your tests go
 

PsyBorg

Wake up! Time to fly!
I did not get to fly today as it was cold and rainy and really windy. I took the day to do some clean up stuff and to recheck the Alien. I found that the extra wires from the new pdb to the FC coupled with the buzzer wire and Receiver wires I ran under the board were tucked in pretty snug. I am thinking this pretty much killed the rubber stand offs ability to move and was transmitting vibrations directly to the FC. Hopefully that or the reflashing of my ESC's has done the trick.

Its supposed to be 50+ degrees tomorrow but cloudy with sprinkles on and off. I will see if I get time between rain drops to test it and get back with results.
 

Raste

New member
I did not get to fly today as it was cold and rainy and really windy. I took the day to do some clean up stuff and to recheck the Alien. I found that the extra wires from the new pdb to the FC coupled with the buzzer wire and Receiver wires I ran under the board were tucked in pretty snug. I am thinking this pretty much killed the rubber stand offs ability to move and was transmitting vibrations directly to the FC. Hopefully that or the reflashing of my ESC's has done the trick.

Its supposed to be 50+ degrees tomorrow but cloudy with sprinkles on and off. I will see if I get time between rain drops to test it and get back with results.

Got a reply from Flyduino, they told me to try to remove all the softmounts (motors + FC) and to turn off the LPF. They said explicitly that nothing should be softmounted with KISS hardware. Though I find this weird, I will try that out and report the results.

I might also try to tinker with the TPA if I cannot get good results
 

PsyBorg

Wake up! Time to fly!
Hehe I have been vindicated!!!! I been saying since I built the Alien all that is needed is the O rings under the FC. That is all I used on my first rendition. I swapped to the rubber standoffs because I actually only have one set of the O rings and liked it on my last build.

When the Alien was at its best for me I has set TPA lower as It was backing off too much on full throttle punch outs. I also dropped that first ramp down to 0 so it was flat til 55%. I actually wish they had a five point curve for that and not three.

KISS is Rock solid to me even at 1k loop times. Yes it can be finicky to get that last few percent tuned but I found it more of a prop issue them the FC. I started with them crappy gemfans. They would break if you even looked at em funny. I quickly swapped to the DAL T5045bn tri blades for the durability and never looked back. Once the Cyclones came out I quickly learned that profile prop is the best out there and I will never go back to a straight blade prop again. As soon as I tried those I was easily able to get a stable tune across the full throttle range with that little change in TPA.

My only thing left to look at if I still can not get this tuned is the Mr.Steel pdb and how that sits directly on the frame vs the original that had only four small hard points with the press nuts. I may end up using my four O rings under that to see what that changes.

Well here is hoping for success for both of us in getting things properly set up so we can get back to ripping holes in the universe!
 

Raste

New member
Hehe I have been vindicated!!!! I been saying since I built the Alien all that is needed is the O rings under the FC.

Well here is hoping for success for both of us in getting things properly set up so we can get back to ripping holes in the universe!

Haha, I just finished hardmounting everything, set the correct parameters, and I'm just waiting for a decent weather to go test it. I really hope this fixes it, even though the FC will eat a load of vibrations now.

But hey, at least if it does not work, I will have solid arguments to go back to the flyduino guys.
 

PsyBorg

Wake up! Time to fly!
Be sure to default everything and start from a fresh set up to make sure that the problem was not something you changed during previous set up and tuning attempts does not carry over.

I am going thru my radio now and doing that very thing so I have a totally fresh start.
 

Raste

New member
Be sure to default everything and start from a fresh set up to make sure that the problem was not something you changed during previous set up and tuning attempts does not carry over.

I am going thru my radio now and doing that very thing so I have a totally fresh start.

I only adapted a little bit the P for the roll and pitch, should not be an issue. I tried by holding the drone by the battery, with props on, and it seems to have done the trick somehow. I still want to have real flight conditions to validate it worked, but looks promising as far as I can tell. I even used nylon mounts for the FC, so it should be getting all the freakin vibrations in the world now.

Still waiting on a decent weather, probably early next week though.
 

Raste

New member
I only adapted a little bit the P for the roll and pitch, should not be an issue. I tried by holding the drone by the battery, with props on, and it seems to have done the trick somehow. I still want to have real flight conditions to validate it worked, but looks promising as far as I can tell. I even used nylon mounts for the FC, so it should be getting all the freakin vibrations in the world now.

Still waiting on a decent weather, probably early next week though.

Tried it, still not working... wrote again to flyduino, waiting for an answer. I think I will get a betaflight controller, been 6 weeks now I try to make this drone fly with KISS and colleague of mine bought a betaflight, flew perfeclty out of the box.