Photohigher skyline rsgs

chris hornby

chris hornby

Member
DennyR said:
I think that the addition of a small spring to take out that movement would make it even better. The dampener could be causing some reduced resolution at small rates of change. The spring would improve that
Click to expand...
Can you let me know where and how to add the spring?
 


T

thijmen5

Member
Can you all install 1.1.6 without problems? When I install this FW my servo's start spinning like crazy! Both servo's are reversed. What did I forget?
 


S

schnied

Member
Here's a testshot of my rsgs gimbal with 1.1.6
With 1.1.3 I had no rolling motion when I yaw the Copter, now I have, unfortunately.
But this one looks more stable than 1.1.6.
What do you think.
The yaw bug looks like a precession error of a normal gyro. Can this also happens to MEMS-Gyros?

pw: skyline

Schnied
 
Last edited by a moderator:


chris hornby

chris hornby

Member
This is my first flight after calming the jitters. There are all sorts of other issues, mostly down to the extreme conditions. Normally I would not think of filming in a high wind but if anybody is watching the Olympics, the sailing events are here on Portland so you can check the conditions on the TV! Anyway you can make up your own mind but I don't think skyline jitter is evident here. I can only confirm this when I fly in calm conditions, and we dont often get those here on the island!

Wind speed 24KpH Flying Octo with DJI in Manual mode. (No auto stabilisation on the airframe) Normally I would film in calm conditions in Atti (stabilised) mode. Waiting for quiet weather to give it a fairer test. Filmed with Canon 550D and gopro.

 
Last edited by a moderator:

MultiHexa

MultiHexa

Member
I'm extra-time flying back and forth and back and forth. If it were now the last rucker be removed in the new software (Let us call you once 1.1.7), you could almost compare the skyline with the Zenmuse.

I hope this gets through PhotoHigher yet. Would be a shame if not.


 
Last edited by a moderator:


T

thijmen5

Member
Looking good Christian, wish I had something near that. Can you provide us the settings? I am having a nex5n and AV130 so it should be pretty much the same. (but first I need to get that firmware running, servo's spinning like mad when powering up in 1.1.6)
 

D

darkovujasinovic

Member
I do not get it how you managed to get it so smooth as my 1.1.6 is very rough even when just sitting still with the lowest gains
What did you do ?
How did you set it , which servos ?
 

MultiHexa

MultiHexa

Member
Thank you for the flowers

I've just set everything back to 1.1.3. Then again the 1.1.6 set up and plays everything by hand.
With the AutoSet it does not work for me. After the Autoset trembles with me everything.
I have everything set without AutoSet. Servos Nick Savöx and Roll Savöx.
 


D

DennyR

Active Member
I just do not know what to do with this, I now have a really bad cross axis error and the tilt servo just started to rotate one way only maybe I can get more sense out of 1.1.3

I think the gyros are saturating. There is not one single function that works correctly.

There was some talk about a faulty batch of boards. I think some of these may work ok whilst others are crapp.
 
Last edited by a moderator:

D

darkovujasinovic

Member
Well, I did the same as you Multihexa (1.6.1 manual procedure ) and I got the perfect result on roll while tilt is still a bit jittery .
With autotune it is horrible
 

DJIFlyer

DJIFlyer

Member
darkovujasinovic said:
Well, I did the same as you Multihexa (1.6.1 manual procedure ) and I got the perfect result on roll while tilt is still a bit jittery .
With autotune it is horrible
Click to expand...

Here is a tip that I believe will help your pitch jitters. I just did this and it improved things considerably: Your camera tray has a series of mounting holes that run vertically. These are the holes that have the little screws which mount the camera tray to the gimbal. If you are using a 550D/T2i like I am you will find that the camera is sitting too tall on the tray as the gimbal comes from the factory. Look at where those little screws are relative to the CENTER of your lens. I'll bet that they are slightly beneath the center of your lens. In a perfect world the pivot point of the tray should be lined up dead center with the lens. But since this is not a perfect world and PH has sold us all a complete piece of s**t so... drop the tray ONE ADDITIONAL HOLE downward from the center (as I just explained). This will induce a TINY bit of pendulum effect and will minimize the chattering that takes place. This worked for me and I have a few other trick but I am waiting for parts to arrive before I say more. If you want to see a real chatter box: try removing that thick, black, shiny spacer that is placed vertically between the left side of the camera tray and the left side of the gimbal (the opposite side from Photohigher's "S**tline module". That thing is in there solely for the purpose of creating friction (there's some first class engineering for ya). If you remove that you will remove the friction and this thing with jitter meth-head on crack. Give it a try, it will take you 5 minutes to figure out if this will help. I did it and saw a significant improvement.
 

nicwilke

nicwilke

Active Member
DJIFlyer said:
Here is a tip that I believe will help your pitch jitters. I just did this and it improved things considerably: Your camera tray has a series of mounting holes that run vertically. These are the holes that have the little screws which mount the camera tray to the gimbal. If you are using a 550D/T2i like I am you will find that the camera is sitting too tall on the tray as the gimbal comes from the factory. Look at where those little screws are relative to the CENTER of your lens. I'll bet that they are slightly beneath the center of your lens. In a perfect world the pivot point of the tray should be lined up dead center with the lens. But since this is not a perfect world and PH has sold us all a complete piece of s**t so... drop the tray ONE ADDITIONAL HOLE downward from the center (as I just explained). This will induce a TINY bit of pendulum effect and will minimize the chattering that takes place. This worked for me and I have a few other trick but I am waiting for parts to arrive before I say more. If you want to see a real chatter box: try removing that thick, black, shiny spacer that is placed vertically between the left side of the camera tray and the left side of the gimbal (the opposite side from Photohigher's "S**tline module". That thing is in there solely for the purpose of creating friction (there's some first class engineering for ya). If you remove that you will remove the friction and this thing with jitter meth-head on crack. Give it a try, it will take you 5 minutes to figure out if this will help. I did it and saw a significant improvement.
Click to expand...
I used to have an Xaircraft hexa with their camera gimbal... I had 2 rubber bands from the sides of the gimbal cradle to the landing skids and that worked to remove the deadband. It wasn't much pressure, but enough to reduce chatter.
 

DJIFlyer

DJIFlyer

Member
nicwilke said:
I used to have an Xaircraft hexa with their camera gimbal... I had 2 rubber bands from the sides of the gimbal cradle to the landing skids and that worked to remove the deadband. It wasn't much pressure, but enough to reduce chatter.
Click to expand...

I can see where that would help. I am absolutely certain that I have a solution to the pitch chatter but I will have to wait a few more days for parts to arrive. I'll keep you posted.
 

T

thijmen5

Member

This is what I got. My jitter get's increases everytime there is little wind or at descend. Sometimes it's allmost pointed towards the sky. My horizon got better, but still isn't what it should be.

The video is at FW 1.1.3, but my software says 0.0.2?
 
Last edited by a moderator:

J

jes1111

Active Member
All this talk of deadband - correct me if I'm wrong, but "deadband" refers to the servo's centre position only and how much "margin" there is either side of (say) 1520us before the servo tries to move. So if this was a deadband issue, the jitter should only show up at the centre position. If, on the other hand, it's related to mechanical backlash in the drive system (i.e. the servo's backlash plus the belt drive's backlash) then the jitter will occur at any angle. The revelation that varying/removing any restraining friction on an axis alters or removes the jitter issue further confirms that it's likely to be backlash at the root of it. This suggests that PH tuned the algorithm to cope with an "ideal" setup and didn't test across a sufficiently wide range of variables to ensure that the system would cope.

Has anyone tried this "on the bench", i.e. just hooking up loose servos without anything attached to them?
 

You must log in or register to reply here.
Top