• This site uses cookies. By continuing to use this site, you are agreeing to our use of cookies. Learn more.

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

PsyBorg

Wake up! Time to fly!
Mentor
#21
This would be a whole lot easier to help you if we had screen shots of your settings and maybe a short video of what it is doing when you try to fly
 
#22
This would be a whole lot easier to help you if we had screen shots of your settings and maybe a short video of what it is doing when you try to fly
Video is still the same one where you can hear those noisy vibrations throughout the whole frame. The motors are causing this and not because they run, but because the FC asks them to do weird speed changes all the time.

As for the settings, here they are:

Code:
{"G_P":[3.5,3.75,8],"G_I":[0.035,0.035,0.05],"G_D":[10,10,5],"ACCtrim":[0,0],"RC_Rate":[0.75,0.75,0.75],"RPY_Expo":[0.8,0.8,0.8],"RPY_Curve":[0.4,0.4,0.4],"ACCZero":[0,0,0],"TPA":[0.4,0.2,0.4],"RGB":[0,0,0],"CBO":[0,0,90],"AUX":[17,0,53,0,0,0,0,0,0,0,0,0],"DB":[0,0,0],"NFE":[0,0],"NFCF":[200,200],"NFCO":[100,100],"ver":111,"reverseMotors":0,"A_P":4,"A_I":0.04,"A_D":10,"RXType":8,"PPMchanOrder":0,"CopterType":2,"Active3DMode":0,"ESConeshot125":5,"MinCommand16":1000,"MidCommand16":1500,"MinThrottle16":1050,"MaxThrottle16":2000,"TYmid16":1500,"TYinv8":0,"maxAng":50,"LPF":0,"ESConeshot42":0,"failsaveseconds":2,"BoardRotation":0,"CustomTPAInfluence":0,"TPABP1":30,"TPABP2":50,"TPABPI1":30,"TPABPI2":0,"TPABPI3":0,"TPABPI4":100,"BatteryInfluence":0,"voltage1":12.8,"voltage2":14.8,"voltage3":16.8,"voltgePercent1":130,"voltgePercent2":100,"voltgePercent3":70,"loggerConfig":0,"vtxChannel":32,"vbatAlarm":0,"debugVariables":0,"mahAlarm":0,"lipoConnected":0,"lapTimerTypeAndInterface":0,"lapTimerTransponderId":0,"loggerDebugVariables":7,"YawCfilter":35,"vtxType":0,"vtxPowerLow":25,"vtxPowerHigh":25,"motorBuzzer":0,"loopTimeDivider":1,"yawLpF":0,"DLpF":0,"adaptiveFilter":1,"ledBrightness":100,"setpointIntoD":100}
I now have raised the TPA to <45, 15, 35, 100>, lowered the LPF, raised the yaw filter and I almost had a full flight without those vibrations. I tried turning off the LPF again and as soon as I armed the motors went full retard on the ground. I will get a lower value for the LPF and I hope it flies, but a lower value on the LPF means more latency in the sticks because there will be more filtering...

I am really starting to understand a tad more the KISS FC, and the more I watch comparisons between FCs the more I see people leaving KISS because of all the settings and tweaks it needs just to "fly".
 
Last edited:

PsyBorg

Wake up! Time to fly!
Mentor
#23
Ok from my experience once again... stock pids are too low and just basic average to get the most builds in the air so they can be tuned.On any KISS platform I have built I have never had one ether the Alien true X or Taz Stretched X have the p gains come in under 4 or higher.

What is happening is that you are starting with low PIDS to begin with. This is why the motors seem cranky and complain at hover. The base TPA curve (and this just baffles me) starts off lowering gains making the issue at idle to 35-40% throttle less stable. Then there is a short range (cruising) where it is back to a base low tuned pids. Once you start getting above 50% throttle they really get loose and that is why you are desyncing and falling out of the air.

So.. where to go now.. default everything. No filters, no yaw filter change, and defaulted TPA settings. Remove the props and go into the gui so you can watch the gyros. manually run each motor up one at a time with a smooth ramp up. Look for excessive vibration. Once you have a fair assessment they are not super noisy then we can move on.

Set Failsafe time to 0 seconds
Set P gains to R=4 P=6 Y=8.5 Set I gains to R=.04 P=.04 Y=.05 and set D gains to R=10 P=10 Y=0.
Set the TPA curve to be flat 0 til 55% then make a straight line to 100%
Set the TPA strength to .25 .2 .4

These are my current Alien settings. Where stock pids fail is they are too low and they do not compensate for longer body quads which require more effort on the pitch axis to control thus higher P gains for that axis. KISS also for some unknown reason has that screwy base TPA curve that makes zero sense to me.

Kconfig 4_9_18.jpg

Krates 4_9_18.jpg

KTPA 4_9_18.jpg

If your build is good and your motors show no sign of bent shaft or warped bell (noise) then this should get you off to a good start to be able to begin real tuning not that "set defaults and just use filters" FOTM everyone seems to like these days.
 
#24
I see you have tweaked TPA as well, and I really feel like there is very little resources to really understand how it works and what it does. Could you dive a bit into your TPA settings and explain ? That would be super helpful!
 

PsyBorg

Wake up! Time to fly!
Mentor
#25
Its pretty simple really.

TPA=Throttle Pid Attenuation. That means the higher the throttle goes to more it attenuates (lowers) pids. If you go into the gui on the TPA tab you can see the curve you produce. If you move your throttle stick you will see a marker that travels along the path of the curve you set by adjusting the TPA numbers. Now when you do this you can see what changes are made numerically in the right hand set of PID boxes as it travels that curve.

It can significantly drop your pids using stock settings. Add that to already low PID defaults and you are asking for problems. I think this is also part of the issues people have with mid throttle oscillations when using KISS.

So what I did was get rid of the default curve where it starts at like 30% influence.. No need for that at all and as I said that STILL baffles me why they set that. I tune my pids pretty tight from the start. I dont want that to change at all. BUT.. when the motors start hitting higher rpm and the resistance (drag) starts kicking in it makes the quad act as if they are tuned to high and oscillations begin to happen. So I use the TPA to slightly loosen the pids as the throttle gets higher. That is how I clear up the mid and upper throttle shakes.

It takes very little change if your gear is in good condition. Bent props don't really make much difference as long as there are no chunks missing and they have been bent back as straight as possible. Which also brings up a good point I have learned. Straight props are garbage. Like the airplane style ones. The varying pitch blades like the Avans and the Cyclones and others like them are much better for quads. They give a smooth throttle response across the range unlike flat blade styles where you can tune them for a "band" within the throttle range and can end up chasing a tune forever. They also tend to not flatten out at higher rpms like flat blades do.
 
#26
I solved my problem: I installed a Revolt FC V3. All problems solved instantly, flies perfectly with stock pids.
  • No mid-throttle oscillations
  • No flips of death
  • No losses of power
  • No issue with the setup
  • Cleaner build
  • Cheaper
Took me 6 weeks to get my drone to fly like real bad, very unusable, wasted money with Kiss, 15 minutes with Raceflight. Just burn your kiss FCs and get pretty much any other raceflight or betaflight board. Best advice possible
 
Last edited:

PsyBorg

Wake up! Time to fly!
Mentor
#27
OK mate several things...

First and foremost this is a family oriented forum and foul language is a no go.. EVER.. Please edit that other wise I am sure that will be reported and a mod will adjust your score card.

Second Just because YOU have had issues with KISS and may be missing something or not doing proper set up or what ever the unresolved issue may be does not mean the product is as bad as you think or state. KISS is one of the most stable and smooth FC out there and they have always been one of the top pioneers in the industry.

It probably boils down to a comparability issue between the old FC and the new ESC's which YOU made the choice to mismatch. Anyways I am glad you got your quad in the air and are happy with it. I wish you the best of luck with it.

Please don't forget to adjust the language in your previous post to something less 5 year old temper tantrumish and remember this is a place of all age membership and be respectful of that in the future.
 

Snarls

Gravity Tester
Mentor
#28
The noise you were experiencing sounds very distinctly like D term related oscillations. I would have tried lowering D, but since you were at stock I would have suggested even increasing D (yes Ive eliminated that sound by increasing D).

Yes, good to hear you got your quad working, but be mindful of your attitude towards this hobby. I know it is frustrating at times. KISS can be one of the hardest FCs to tune, especially for some components like Emax motors for some reason. I have yet to get my KISS quad to fly exactly like I want it, but even still it is the most locked in quad I have ever flown. And it is two years old with he V1 FC. My betaflight setup is smooth like butter in flips and rolls, but the KISS tracks like its on rails. That's why people race (and win) with it. Sorry you couldn't experience it, but build up your stick time (and have fun!) , and maybe one day you'll come back to it.
 
#29
Im just plain done with Kiss... tried two entire builds, followed everything to the spec. I tried tons of setup for the PIDs and just nothing worked. Im not sorry to say it but even for advanced pilots Kiss is just unbearable. I have experienced everything that could go wrong with Kiss, with two different FCs v1.

About the "you have an FC v1 and ESCs v2" thing, it is written nowhere that those are not compatible or may cause issues. Not even on the official site. Documentation is lacking, experience is over-complicated, and racers now go to raceflight because they invested where it is useful: a better gyro and not a more powerful processor. It took me 6 weeks to get in the air to have a drone that would not just fall out of the skies with kiss ; it took less than 3 hours to build and fly perfectly with Raceflight.

It is complicated enough already to tune a quad to your liking, it should not be harder because of lacking firmware and software as well as poor support. With raceflight everything went as smooth as possible, just as it should have been right from the start. I am super mad but I agree it does not justify an angry language, hence I edited a bit my previous post. I guess the best thing to learn from my terrible experience is that unless you want to spend your time in settings, you should avoid kiss.
 

PsyBorg

Wake up! Time to fly!
Mentor
#30
I guess everyone has their opinions based on personal experience. It seems like you now know more then people like Final Glide and Mr. Steele about KISS gear as well as the guys on winning race teams using KISS.

All I can say is as long as you are happy with your current set up Rock on.

Just please stop passing false information about something you have little knowledge of because of a bad experience.
 
#31
It seems like you now know more then people like Final Glide and Mr. Steele about KISS gear as well as the guys on winning race teams using KISS.
Just please stop passing false information about something you have little knowledge of because of a bad experience.
My experience is going from not being able to fly at all, to terrible flying, to the FC going crazy and attacking me. 11 stitches and a ripped finger nail. Call it what you want. I do not pretend to have more experience than these guys, I say that I have built and configured RC planes for over 15 years and I never had such a terrible experience, and that my betaflight and raceflight experiences went smooth with no issue at all.

Facts, not opinions. If you want to call it false then I cannot stop you, but I have a responsibility to share my experience so others may avoid going through the same one and learn from my mistakes.

Time to close this thread
 

PsyBorg

Wake up! Time to fly!
Mentor
#32
Well still gonna say... This looks more to be inexperience unfamiliarity then gear so far. Show us some pictures of the build and some screen shots of the settings for both the quad and the radio and I am sure we could diagnose the issue. Until then I have to side with the gear being fine. If it were not fine a lot more people then you would be having issues.

I am sorry you injured yourself as that is never fun and I am sure if you were more familiar with proper set up practices you would not have been in the position. We have all been there mate including myself. I nearly took my own head off and buzz sawed a friend doing something in a not so safe manor. We just got very lucky.

I wish you a full and speedy recovery.
 
#33
Hi All,

I came across this post many times as I've been researching the exact same issue I had and haven't had an answer to give or gotten enough info back out yet to figure my prob out. I built up a brand new 6s 5" freestyle quad and from day 1 just had non stop trouble with hot motors and the same motor spaz outs shown in his videos. Basically it boiled down to a badly design frame with terrible resonance. Kiss fw is alot more sensitive to vibration hence the faster the reaction when compared to the bf fw, the noise would get into the gyro and cause the motors to spaz out via the D term. This is proven by having extreme hot motors with all defaults and to get cooler motors you had to drop D gains down to almost nothing. I then put the identical kit on a solid base style frame and with defaults the quad came down much cooler. Switching to betaflight FW or a betaflight fc made no difference at all either.

I see he is using a chameleon frame so can only imagine there was excessive noise in the motors as those frames are designed to work within the kiss fc limits and defaults.

A blackbox log would then be the last step to assess the noise spektrum and custom set the filters and notches based on the noise bands that got through the adaptive filtering to cut out all the noise. But not all frames and motors can be remedied in my opinion on kiss FW for the simple reason in my opinion that kiss is not going to reward you well if you do not give it atleast a decent solid frame and good quality motors. If you manage to filter out the noise, you will have extra delay and additional prop wash but also risk damage as a simple bent prop will throw out the tune again.

If this gets any attention, I will gladly comment back my final results amd progress as well as share the spektrums from the failed quad build. I have ordered a impulserc reverb frame and am going to give this another try when it arrives tomorrow. What an exceptional learning curve this has been. It gave me greater respect for this sport, it's equipment and all of its players.

Have a great new year everyone!
 

PsyBorg

Wake up! Time to fly!
Mentor
#34
Being a KISS user for 3 years now and still flying an original RR5 Aliein I can say that some of your thoughts are not totally correct.

KISS FC's are designed to handle more noise. They even reccomend no soft mounting of anything other then the O rings that come with the FC.

As far as your statement about adding lag that is either your radio gear or someones not understanding tuning well enough. Even the default settings are good enough to fly with out rotor wash if you know how to avoid it to begin with.

KISS FC v1 is still to the day better made and more well behaved then a lot if todays fancy bells n whistles laiden FC's
 
#35
Yeah man, I'm not disputing your experience. I find most of the time, conveying what it is we are trying to say is more difficult... But what you are saying or misunderstood me on, is yes, you can make kiss fly with anything on defaults, the point I am conveying is that, the more filtering you need, the more latency is introduced in the flight controller, this is known fact. But lets be clear here, we aren't talking about a flying quad, we are talking about a performing quad, you can put any quad together, hit arm, and on defaults it will fly to some degree. Not great, possibly you won't have motors left, but it will fly. Hence the race by the developers to get the adaptive / dynamic filters working so well because it allows for additional filtering to be turned off or made narrower, there is tone or video out there on this, including from JB himself. It takes care more and more over time of bad parts as the code advances. So yes, we can tune out the vibration / jello / shakes with PIDs and Filters but sometimes at a cost of flight performance the worse the parts get. Fact is, the more delay, the less the ability for the flight controller to mange the prop wash. Yes you can fly around it, but not always... For some it doesn't matter, for some it does. Can't beat everyone with the same stick. That's my whole point I am making and the small 20% odd of frames / motors / variables that influence flight outside of defaults.

You see the general just I get from forums and responses are the same, "It should just work", "we don't understand your issue", "defaults worked for me". Not every case is the same, we know this yes, there will be cases where the defaults don't work. What I am seeing, no one yet has enough experience to be able to say, yes it is this. So we are all grasping and learning together.

If we dumb this down completely, and say, ignore all settings, changes, leave perfect defaults, then attempt to fly, if it doesn't fly within parameters i.e bad oscillation, super hot motors, we would usually then be left with 1 option, change parts until the issue goes away because we must have a bad part right, because as you say, defaults cant be to blame, Kiss is better than betaflight at ignoring noise? Fact is, as you guys all say, it has to work, mine works, has always worked, and I've never had an issue, you are doing something wrong obviously. So by ignoring software and FC, we can isolate the bad part, which in cases appearing now seems to be happening more often with new equipment. Sadly though it is not economical and affordable to keep buying different parts till you find a combination that works. It may just be the motors that were selected, based on how they vibrate coupled with the choice of frame, you get a resonance you can't isolate. So you would change the frame, or the motors, and now suddenly all works. What our tech now allows, is to take the stuff that didn't work, and adjust the system to make it work when the default scope or range doesn't cater for the setup. Getting those adjustment parameters is where all this arguing and fun comes in.

Give you some more examples of,

Kiss - did you know that if you are not running 1k loop on either V1 or V2, the adaptive filter doesn't work, so come Mr newby, fresh off betaflight and wants to upgrade to kiss, he sets up his rig, activates 4k pid loop on V1, or 8k on V2, he arms, his motors smoke... oh dear, what happened? Well, you had absolutely no filtering except a little on the yaw Gyro, the D term noise amplified and cooked his motors.
Lets say he was sharp enough to spot the issue on arm, now he spends how much time trying to work out all that was the issue was the fact that the loop was not set to 1k... I see a great opportunity here for a system that detects your parts, reports and stores this data. As you get it working or others do, you can share in this data to assist each other with profiles and settings. e.g. oh i see you are running those cheap motors from banggood, you would need a special notch filter at approx 245hz to help with their noise, now there in under an hour, we can have the issue sorted.

Notches are off by default now. A noisy (new) setup will oscillate at mid throttle. you need the notches to be turned on or you have to back off P and especially D, which hurts flight performance. Have you ever installed blackbox and used it to run tests across different frames and motors and watched how the spectrum moves across the frequencies but mostly stays in the default range of the adaptive / dynamic filters, but not always to fully understand what the filters are doing? Or were you just lucky enough, that in 3 years you always managed to have your stuff operate well enough inside default? Let's not tell people they don't know what they are doing and rather tell them, hey we don't know either but lets put our heads together and figure this out or, tell the man, you bought nonsense, rather buy better parts because we don't know how to help you tune and filter the bad stuff, yet.

What is the major change in betaflight 4, dynamic filter improvements... And why people ask, because as the hobby grows, more and more parts and setups are becoming available that can't fly on older FW because of poor quality, the parts are just too noisy and that fw didn't know how to deal, so they are changing the code to adapt for more noise to help us all get in the air easier, quicker, more issue free...

And that is my rant done. Peace and happy flying to all! Defaults or not!
 

PsyBorg

Wake up! Time to fly!
Mentor
#36
Preachin to the choir mate...

I am one of the people that try n tell people that jumping on the bandwagon and cookie cuttering their builds will usually cause more problems.

You can build two exactly same quads and thete will always be slight differences and quirks.

Beta flight as good as it is has fried more gear than I can keep track of because (for example.. followed Bardwells or Steele or whom ever)
And the gear they have did not like those settings.

I am doing and documenting a start to finish quad build series and will be pointing this all out.

Videos are a tool not an instruction manual to be copied or set as law. More if an example how to get from point a to point b
 
#37
I have to agree with you on those tutorial videos, a number of times, Bardwell has caused me more problems than actual help. That and half of everything that is shown I feel is biased... But I must say, having the opportunity to build a bad quad is one I would recommend to anyone wanting to learn the tech side quickly. Even if you can afford the good parts, I would recommend at least one cheap build. The learning curve is just phenomenal and sets you up to deal with most common probs quickly.

I have managed to resolve my issue that I had when I found this forum. My fix was, changing the frame. I tried out the iFlight battleaxe on a 6s build, first with the new 2207 1800kv XING motors, then with the MARK II 1750kv Motors. Results were, on default, regardless of flight controller type used, the quad after 30's would have steaming hot motors. An enormous amount of people were oblivious to the fact that this issue was an issue and a serious one to begin with and were unable to help me or point me in the right direction. They all just assumed I did something wrong. I was lucky enough to have spare components lying around that I could test with to try and isolate the issue but didn't have much luck, which told me that it could be the frame. At this point, my own ignorance turned into wisdom. I picked up all my quads and started to tap on them with the screw driver, weird yes, but in doing so, you could clearly hear the left over ringing resonance of the frame, all other frames was a very short, high pitch ring that ended quickly, the battleaxe however, rang with a much lower frequency and for much longer, you could feel the vibration stay in the frame and this just from a tap.

Long story short, I tested by moving the components to a solid base plate frame and on default everything came down cool. I ordered myself an ImpulseRC reverb frame, maiden flight was yesterday and on default she is happy. Motors are running smooth & cool, the HD feed is crystal clear. I even managed to do a little tuning. On the battleaxe frame, I had to bring the D's down to 2 before the motors cooled down. On the Reverb, i have it over 16 already.... Bit of a pointless fact, but to the experienced, this shows the noise amplification by the D term making it past the filters on the bad frame, so much so, it was not worth trying to filter it out.

Some more things that I found that could cause this issue on a new build:
- Loose arms, not tightened down properly. Frame has too much slop. Frame is poorly designed with arm mountings that don't hold the start of the arm. e.g. the battleaxe arm screws are near 4cm away from the start point, they mount the arm basically outside of the centre of the frame. There is nothing holding the inside of the arms except for the top and bottom plate, Any vibration through the arm is amplified as it's acting like a small lever to the centre of the frame. All other great frames, the arms are screwed in underneath the body in the centre.
- Bad motor bearing - I've seen this time and time again. Just because the motor is new, don't trust that the bearings are fine. Always run them on the bench via the motor tab and feel for excessive vibration. High quality motors are not even safe from this. I've had a steel motor come in with a bad bearing and caused the whole quad to fly terrible no matter what you tried. Swapped the bearings and all was perfect again.
- Bad Esc or ESC Settings - This does happen, although not that often, an esc in a set could have a small enough electrical inaccuracy that it makes a motor / arm behave faulty. You can spot this in blackbox with an over active / under active motor or in your FPV feed as a dipping arm on punches.

Thanks for taking the time to respond my posts. All the best for your write up. I look forward to checking it out actually. Would it be all on this forum?