Success with minor issues!

[FiberToast]

So I'm new to the forum so first off i want to say hello to everyone and thank´s for many hours of reading and many solved issues along the way.

And the greatest of thank you to Flitetest for making all this happen!


I have been tinkering away up here in Sweden during the long and dark winter building a DW Tricopter 2.6HV Deluxe
But finally summer have come and it´s time for FPV.

This is my first FPV platform and i have to say it feels like i had the fortune of instant success. Im a lousy LOS pilot but flying FPV just feels much more natural.

My build is basically the same as Davids from his Rcexplorer.se site with the exemption that I'm running multiwii
Flip2.5 from RTFQuads.

To say the least starting with Multiwii and having no experience in FC´s was quite bite of info to swallow.
Thankfully there is a ton of info around if you just take your time to find it.

There was one big obstacle along the way.
Jittering tail servo.
I bought what i thought was a better quality Turnigy servo Hispeed MG digital and I'm shure it´s great but mabe not for tricopters or multiwii. I played with gyro filters and servo filters in the MW sketch but with little result, moving average gyros seams to help but not totally. In the end i changed to a slower MG digital servo the Bluebird servo David recommends in the FT tricopter build video combined with moving average gyro in the sketch and behold the jittering went away.
I would still want to use a real good servo even if it cost a´lot just to be shure it will hold.

If anyone have a recommendation for a good quality servo please tell.

Then there is a few things i would recommend if starting out with multiwii
#1 Get the bluetooth, makes it easy to adjust PID and other settings.
#2 Run MW V2.3 and be sure to run the latest MVconf or you might have some bugs.
#3 Calibrate you're speedcontrollers and always run Simonk
#4 Don't be afraid to experiment with the sketch you can just upload a new if you mess up.
#5 Get a CNC cutter and CAD and a lasercutter and a watercutter and a 100 other fun stuff to cut your frame and other bits and pieces in stead of doing like me and cutting G10 and Carbonfiber by hand
#6 Don't use the barometer until you know how it works ( i smacked in to the ground on deecel l and gave my self 2 more days work)

Now there is a few things i can't figure out and therefore i turn to you guys and gils here.

I have some kind of glitch in my tail servo it is not consistent and only comes now and then.
There is no jitter but from time to time while flying or on the ground the servo goes full left and full right in a micro second and then everything is normal until it happens again. This is really annoying and i can't figure it out.
You can see it in the video at 0,34 and 6,34

I have a vibration issue in my camera tray in high speed that i think is either the 1mm carbon plate or to soft metal in the vibration absorbers.

Would you mind looking at the linked video and telling me what you think about the long distance focus on the Gopro hero3+ Black. Some people are saying you have to refocus the lens. Is this true for all go pro´s of that model or just the early ones. I don´t know how much better it could be but i may be blind.

Here is some pictures of the Tricopter
The tail mechanism is from Armattan and is really strong and solid
I include my PID´s, they are by no means perfect but they work, I'm still trying to get them tweak´d out.




Again Thank´s to all the great people here on this site.

 

[FiberToast]

Oh and i have a bussing noise in the video when my FPV gear is connected, I'm using Davids power regulator and filter and I'm flying from the go pro as i think it´s nicer than my ccd
The video is still uploading to youtube.

 

[cranialrectosis]

Buzzing sounds and quirky intermittent flying issues sound like vibration from an unbalanced motor or rotor.

Are you sure everything is tight and the motors and rotors are in balance?

 

[FiberToast]

Yes the rotors and motors are balanced. Everything is running smooth until this glitch with the tail happens. It can be 5 min between them.
The noise in the video is defenitly something from the electrical system as it goes away if i disconnect the live out cable.

If i could get this stupid video to upload to youtube. its given me errors 2 times and takes 90min for a 1Gb file is that normal
it´s my first time uploading

 

[cranialrectosis]

The hum in the video is electrical. I hear a rattle though.

Throughout the entire video, I hear low frequency and mid frequency vibrations (rattling). The pitch changes as you increase or decrease the forward speed of the copter. When you are pushing forward there shouldn't be any cavitation caused by the rotors being in their own wash. This to me is the best indicator that you have an unbalanced rotor or motor or a loose motor or rotor or boom. These issues can be electrical (loose bullet connector maybe), but I doubt it. The hum is electrical, but the rattle is most likely not.

Do you use a Hobby King power distribution board?

 

[FiberToast]

Hi Cranial
No i don't use a distroboard at all and all the cables are soldered to the esc´s the only bullet connectors is on the motor leads and the female ones is soldered directly to the esc´s and held tight with epoxy.
No parts from Hobbyking at all exept LiPos and the servo.

I found today that i have a small crack in one of my arms lengthwise wish gives a lot less torsional strength in that arm then the others. Also my tail assembly is beginning to show a little slop wish i will correct. The rattle was not there from the beginning but i noticed it first after i changed to the servo I'm using now.

The props is seriously balanced so then it have to be the motors but T-motors should be good don't you think?
there is no jello or vibrations in the video from what i can see at least not in hover or slower speeds.

I strengthened the camera tray today with a small piece of carbon in the front and it took out a lot off the shaking in higher speeds.
Throttle PID attenuation also took away some more of the motors irregular rpm in fff.

The hum seams strange to me because i have the power filter installed and i don't power the go pro from the flight battery, only take
video out. I had no such hum in my ccd camera on the audio channel with the same power supply.

Suppose i have to get a BO hex and a dual camera setup as i feel that hi speed close to the ground is really addictive.
Made about 10 flights today.

So any opinions about the go pro focus. Is it worth messing with refocusing the lens and voiding the warranty?

 

[cranialrectosis]

Your video is isolated from the vibration by your camera mount. It is the FCB that may cause the oddities due to vibe hitting the sensors on the FCB. Jello is a good indication of vibration but with your setup, is not imperical evidence of vibration or the lack thereof.

How is your flight controller isolated from vibration?

IMO your first step in isolating and correcting the issue is insulating your flight controller from vibrations and that may be as simple as correcting that rattle.

The electrical hum and the camera focus are outside my purview.

 

[FiberToast]

I totally agree with you that removing wibes is quie.
But that does not help the random glitch in the tailservo as it occours even with the motors un armed on the ground. It can be completly still for a minute then it twitches left right and then goes still again.
The FCB is isolated with gyrotape.
Thanks for your help

 

[cranialrectosis]

I totally agree with you that removing wibes is quie.
But that does not help the random glitch in the tailservo as it occours even with the motors un armed on the ground. It can be completly still for a minute then it twitches left right and then goes still again.
The FCB is isolated with gyrotape.
Thanks for your help

Sorry, I did not understand that the twitch happens on the ground and during flight.

 

[FiberToast]

No worrys.
I dont think its a brownout but rather a glitch in the fcb or some kind of bug.
Will update with a new video later and fix the cracked arm.

 

[FiberToast]

My feeling is that its some kind of error building up in the pid loop that releases and then starts building again. But i dont understand MW well enough to tell if thats possible

 

[FiberToast]

Guess my FCB have Torrets syndrome.

 

[jhitesma]

Have you watched in the configurator while it's twitching and seen if there's a signal going out on the servo line?

If the configurator is still showing the servo at 1500 when it twitches then I'm thinking it's something in the servo itself and not the FC. But if the configurator is showing the servo signal changing...then it probably is something in FC causing it.

It could be something in the PID loop (most likely the I I'm thinking since that usually handles errors over time) - you may want to try eliminating the I term from your Yaw PID...but I'm not convinced that's the issue.

It could be the long wire from the FC to the servo acting as an antenna and picking something up that's causing the twitching...you could try a ferrite on the servo lead.

But I'm still thinking it may be something in the servo itself. Or it could be that moving average you have enabled as it could be building up error and then clearing out. Generally the moving average causes more problems than it solves and the problems it does solve can often be corrected with better vibration isolation.

 

[FiberToast]

Hi jhitesma.
The servo is connected directly on the ESC with only the signal wire running inside the carbon boom. I had the same problem with the other servo i have been using. Maybe i should try changing the ESC. I will have a look in the GUI but i think i checked that before.
The moving average did for me about the same as setting the gyro and servo filters down to 10 and 42hz.
I was thinking that running the filters would impact the PID´s in a negative way so i tried the moving average instead.
Before setting that up i had a terrible tail jitter even with ESC unarmed, it would go away as soon as i got in to hover but the copter would shake itself to bits on the ground. The sporadic twitch have been there all along.

Also i have been checking for vibrations using the GUI after balancing props and can´t see anything much.

 

[Cyberdactyl]

I found today that i have a small crack in one of my arms lengthwise wish gives a lot less torsional strength in that arm then the others. Also my tail assembly is beginning to show a little slop wish i will correct. The rattle was not there from the beginning but i noticed it first after i changed to the servo I'm using now.

Hi FT,

I had the same problem with one of the 10x10mm CF tubes on my tricopter several months ago. Something happened, not quite sure, I think it was from a fairly minor crash, no crack but the torsional strength got really bad on one boom. I could twist the boom 30 degrees or more with little effort, about the strength of PVC. The other booms, and new ones, resist even 5 degrees of twist with great effort.

I fixed it pretty much back to its original strength by taking a strip of carbon fiber plate 9.5mm x 1mm (just a hair under the tube width so there's no sharp edge) x approximately 70% of the boom length and epoxying it to the boom on two opposing surfaces. I used the high gloss CF plate, so I lightly sanded the surfaces that would be epoxied. Again, there's no need to do the entire length of the boom, just 70% or so.

It was an easy repair because I didn't have to disassemble anything.

 

[Cyberdactyl]

Here's a closeup of the repair.

 

[FiberToast]

Hi Cyberdactyl.

Thats smart, i have been repairing these kind of split booms on micro helis before by putting CA and covering in shrink-wrap. Your method adds more rigidity i guess.
Another thing i did on my booms was to cut a short about 2cm piece of thin brass tubing with the same outer diameter as the carbon booms inside. I then epoxy this piece inside the boom before drilling my hole in the boom. This helps prevent splitting and does not add a lot of weight while i can still run the wires inside. It stops the frame from compressing the CF boom.

I think some of these booms are not the best as the thickness of the wall is about 0.1mm and they compress easily and then split.

I suppose they don't make wowen square booms or how its called. Same like round ones.

 

[Cyberdactyl]

Yep, pultrusion carbon fiber tube, whether it be square or round, is considerably cheaper than woven wrapped CF tube. Of course it is a healthy percentage weaker for the same cross section. Still, the pultrusion CF 10x10mm tube at HK is an ok deal. And if you order it in 300mm lengths, instead of 750mm, you don't need to pay for the large box and the premium shipping cost. If you get the longer ones, get more than one, and maybe grab some long round tube, since you're going to get smacked with the long box fee and shipping cost.

And you're right, I learned quickly the HK 10x10mm tube can not be clamped overly hard unless there was a reinforcement column against the tube wall being compressed. Meaning, the screws clamping needed to be touching the tube. You STILL get inward deflection or bowing of the tube on all four surfaces, but not quite as bad. If you don't run the wiring inside, it's an easy fix to allow you to clamp pretty much as tight as the fastener allows. All you need to do is epoxy a short piece of wood dowel snug & tight, maybe 2cm long in the area of the clamping point. I designed a break-apart H quad I never built that purposely has two tubes on either side of the clamping screw for this exact reason.

 

[FiberToast]

Cyberdactyl, you would´t possible have a good set of PID`s for tricopter for MW 2.3 was just thinking something to compare my own with and see what difference it would make. I realize that all builds are different but anyway.

Im realy liking having a conversation in this forums instead of just reading as i have done for years.

 

[Cyberdactyl]

Sorry, no. One of my multirotor 'gaps' is that I only use the KK2 series.