How I balanced the Polar Pro ND Filter [video]

Jaybee said:
The Editor said:
That won't work - If you move the gimbal around without it being attached to the Phantom it will try and fight itself due to the fact that there is no feedback loop information coming from the accelerometers inside the Naza.

The Zenmuse works on a closed loop basis and takes into account not only its own data from internal gyros and IMU/GCU but also utilises a secondary 'feedback loop' from the Naza itself - This is why it's correction and stability is so good.

You need to do any testing with the gimbal properly attached to the Phantom.

Right, fair enough. However, the other axis don't vibrate violently during this testing (they are quiet and smooth unless you go crazy and tilt to extremes) but the yaw axis has crazy vibration. Why would only the yaw servo fight itself?

The Editor said:
You should also do a full calibration of the Phantom and an X1 calibration as well.

Then try, and see what happens

Sure, I had done all that.

I would imagine because the yaw axis is in whats known as 'follow mode'. This means there is a deadband of around +/- 5 degrees where the gimbal is getting aircraft information on the yaw of the craft itself (via Naza and Compass). It then counteracts the yaw with opposing torque in the opposite direction. Outside of the deadband the gimbal will 'think' the craft is attempting to actually turn/yaw and the gimbal will try and follow (with an exponential curve).
You maybe getting away with hand holding the gimbal for Pitch and Roll (to a point) but the yaw axis will be much less forgiving and more locked into the aircraft movement as it needs to 'sense' not only yaw drift but also true yaw/turns. This is why it's having a spazzy fit !!
 
The Editor said:
I would imagine because the yaw axis is in whats known as 'follow mode'. This means there is a deadband of around +/- 5 degrees where the gimbal is getting aircraft information on the yaw of the craft itself (via Naza and Compass). It then counteracts the yaw with opposing torque in the opposite direction. Outside of the deadband the gimbal will 'think' the craft is attempting to actually turn/yaw and the gimbal will try and follow (with an exponential curve).
You maybe getting away with hand holding the gimbal for Pitch and Roll (to a point) but the yaw axis will be much less forgiving and more locked into the aircraft movement as it needs to 'sense' not only yaw drift but also true yaw/turns. This is why it's having a spazzy fit !!
Of course, yeah that makes perfect sense. Cheers.

I think it's still safe to say that this gimbal is more 'noisy' than the 2D, and that the extra vibrations can cause, or contribute to, the jello effect from being physically connected to the GoPro. It's pretty clear it's down to the new yaw axis and the difference in function to the other axis', as you've pointed out there, makes be wonder if this is all software related which will get nailed in an update.

I've deduced that the audio noise (on the footage) from the unit I am sending back is solely to do with the USB connection, it isn't present when I wire it to a back plug from a 2D - but this in turn causes yet another issue of the repeating 'tilt spasm'.

The whole situation is puzzling, confusing, and frustrating :x
 
Will adding a ND filter get rid of the jello in the video below. This was shot at 1080@30fps in the middle of the day to really pick up any jello. I have tried all the dampener balls and the best ones are the white. I was hoping I would not have this issue, but I do. :x I had the H3-2D on a few days ago and flew in about the same conditions and there was not any jello. I have the v1 with the gopro connector in the back. Please give me your input please. I want to get rid of this ASAP!!!

http://youtu.be/qWMcUnAZdc8
 
Jaybee said:
Here's how I balanced the Polar Pro ND Filter today.

Hi Jaybee and everyone else. I have the same Polar Pro ND filter on my H3-3D.

I counterbalanced it solely with fresh, new, BluTak.

One thing you need to be careful of, is that when you mount your weights on the back of the GoPro, (these are the counterbalance weights that stop the GoPro flopping forward due to the weight of the Polar Pro), keep those weights as far to the left as possible (as viewed from the front of the Phantom, looking directly at the GoPro). The reason is that the Polar Pro is already heavily off balance, due to the lens being offset to the right. If you place your counterbalance weights directly behind the Polar Pro filter, you will then need additional weight to bring the roll axis into balance. The Polar Pro weighs aprox 11grams, so if you take 11grams of BluTak and stick it on the back of the Gimbal, directly behind the GoPro, but as far left as possible, then you will have your forward tilt axis balanced. Now you'll find you need to add less weight on the left motor (actually, I added the BluTak to the small flat black area just behind the motor). I found I only needed about 8grams of BluTak to balance the second axis.

Here's a short film I produced here in NZ using this setup. https://vimeo.com/94486996

This guy has done a good job of creating a video showing the benefits of the Polar Pro. https://vimeo.com/91578759
 
shartlza said:
Will adding a ND filter get rid of the jello in the video below. This was shot at 1080@30fps in the middle of the day to really pick up any jello. I have tried all the dampener balls and the best ones are the white. I was hoping I would not have this issue, but I do. :x I had the H3-2D on a few days ago and flew in about the same conditions and there was not any jello. I have the v1 with the gopro connector in the back. Please give me your input please. I want to get rid of this ASAP!!!

http://youtu.be/qWMcUnAZdc8
Well as you can see from al my posts above, it didn't help my current setup and I am sure the jello I am getting is caused by vibration from the gimbal itself + possibly interference for whatever reason from the USB setup, as opposed to the dampening balls not doing they're job.

You're not alone, we all want this **** sorted!
 
Jaybee said:
Well as you can see from al my posts above, it didn't help my current setup and I am sure the jello I am getting is caused by vibration from the gimbal itself + possibly interference for whatever reason from the USB setup, as opposed to the dampening balls not doing they're job.

You're not alone, we all want this **** sorted!

Don't get me wrong, I know a lot of it has to do with the gimbal itself. I just did a little DIY project to see if a ND filter could help somewhat. So I took my Oakley sunglasses, popped out a lens and taped it to the gopro :roll: to somewhat darken the light coming into the lens. It actually got rid of a bit and I could only see it in one spot. I will probably just get the ND filter and put it on and counter balance. I will do this till maybe DJI puts out a firmware update that could fix this crap!!! Also I have the v1, so the USB is not to blame for all of it, even though that noise on your one video was absurd. Lets hope all of this can be fixed by a firmware update and we are not just screwed with a defective design!! (where is the cross your fingers smiley??)
 
andysydney said:
... I have a Phantom 1.2 with the H3-2D.
It seemed to take longer to warm up and balance but it seems to work okay…
http://team-blacksheep.com/products/prod:jelloguard_h3
Hi Andy
Good to know that the H3-3D works with your PH 1.2 and H3-2D (P330CB upgrade kit I suppose?)
I have in mind to buy the H3-3D to replace my H3-2D and would make sure it will work with my configuration, before taking any decision.
I own a Phantom 1.1.1 or 1.2(?) bought last october, but I am not sure of its version (seller couldn't tell me)
It has self tightening props, spring to center throttle, two antennas, but Naza is V1, not V2.
As to the transmitter: it had no tilt lever mounted (I did the upgrade lateron) but PCB inside is marked Phantom II, 2.4 ghz (strange :?: )

I installed the P330CB-H3-2D upgrade kit and connected my Zenmuse H3-2D directly on this PCB without can bus.
It works well except in strong winds.
There is only one particularity: When I give a tilt command with the lever, it's responding with an unpleasant delay. Don't know if its normal?
My questions:
Did you plug the H3-3D directly to your H3-2D upgrade board of did you have to use a can bus?
When you write that it seems to take longer to warm up and balance, how long does it take exactly?
 
The Editor said:
Outside of the deadband the gimbal will 'think' the craft is attempting to actually turn/yaw and the gimbal will try and follow (with an exponential curve).
You maybe getting away with hand holding the gimbal for Pitch and Roll (to a point) but the yaw axis will be much less forgiving and more locked into the aircraft movement as it needs to 'sense' not only yaw drift but also true yaw/turns. This is why it's having a spazzy fit !!

I believe one of the recent H3-3D firmware updates changes this behavior markedly. The gimbal is now fed rudder input from the tx, and will preemptively rotate with the direction of intended travel so it has more "deadband" in the opposite direction to stabilize the turn while keeping the landing gear out of view.

In short, the gimbal now knows unequivocally when pilot is intending a yaw and when the craft is yawing due to attitude correction. No idea though how that is mixed with the standard follow-mode +/- 5 degree trigger.
 
ElGuano said:
The Editor said:
Outside of the deadband the gimbal will 'think' the craft is attempting to actually turn/yaw and the gimbal will try and follow (with an exponential curve).
You maybe getting away with hand holding the gimbal for Pitch and Roll (to a point) but the yaw axis will be much less forgiving and more locked into the aircraft movement as it needs to 'sense' not only yaw drift but also true yaw/turns. This is why it's having a spazzy fit !!

I believe one of the recent H3-3D firmware updates changes this behavior markedly. The gimbal is now fed rudder input from the tx, and will preemptively rotate with the direction of intended travel so it has more "deadband" in the opposite direction to stabilize the turn while keeping the landing gear out of view.

In short, the gimbal now knows unequivocally when pilot is intending a yaw and when the craft is yawing due to attitude correction. No idea though how that is mixed with the standard follow-mode +/- 5 degree trigger.

Hi ElGuano,

I forgot about 1.0.0.6 !! You're right, it did change the characteristics of the yaw stabilisation (I'm not sure for the better!).
I believe what now happens is the deadband floats against virtual 'endstops' that are moved by the rudder input.
In other words - no rudder input = Deadband normal at +/- 5 degrees from centre then exponential damping back to centre.
Rudder input = a defined 'push' (I'm not sure if this is proportional) of the deadband outwards to help prevent the legs getting in shot for the P2.
 
I sent my gimbal back to the supplier. They tested it and found nothing wrong with it, no jello, no audio tone.

However they tested it with a newer Phantom 2 (with larger hole) and they also tested it with a Phantom 1.1 (?)

Not sure what the hell is going on as I've had loads of issues with it, why would it work perfectly with a newer P2? If you check my footage in the video below, look at the correlation with the audio tones and jello (at the 00:05 mark, grass in shot).

Really fkin sick of this now. Gonna get a refund and buy another 2D, at least I could get good looking footage with it, albeit with the yaw wobble.

https://vimeo.com/95396632 (password: jaybee123)
 
Really fkin sick of this now. Gonna get a refund and buy another 2D, at least I could get good looking footage with it, albeit with the yaw wobble.

https://vimeo.com/95396632 (password: jaybee123)[/quote]

Jesus that sounds bad! One thing, take off from the top of your case, not from in the grass. Esp when grass is so long, sometimes you'll get dew etc on your lens.

Cheers
 
nzvideoguys said:
Jesus that sounds bad! One thing, take off from the top of your case, not from in the grass. Esp when grass is so long, sometimes you'll get dew etc on your lens.

Cheers
I wouldn't put the Phantom on the grass if it was wet buddy.
 
Hi JayBee, you're not the only one getting this buzz on the video. Generally when the motors are running it gets drowned out but it's very noticeable when they aren't.
 
PhantomSam said:
Hi JayBee, you're not the only one getting this buzz on the video. Generally when the motors are running it gets drowned out but it's very noticeable when they aren't.
Good to know, cheers.

What I just can't understand is; my supplier tested the 3D I sent back, and they had none of the issues... they even sent me the GoPro footage which had no jello and no audio tone before take-off. The only difference was; they tested it on a newer Phantom 2 with the pre-cut hole.

I'm thinking of waiting a month or so before I buy another one, hopefully these issues will get resolved in time (as they did with the 2D last year) either in more reliable hardware builds, or firmware.
 
I got the same polarizer, but have not used it yet for the same concern of weight. Good idea. I'll need a scale now to measure washers.

Happy Flying
 
evonbart2 said:
I got the same polarizer, but have not used it yet for the same concern of weight. Good idea. I'll need a scale now to measure washers.

Happy Flying
If it's any help, I don't bother counter balancing any more. Not had any issues yet, all working fine.
 
gr8pics said:
Guys, are there two different models of the Polar pro ND filter?
Mine is designed for attachement on the case, and not directly on the lens :?
Yes, you need the 'Naked' version.

filtre-polar-pro-nd-naked-2.jpg
 
Not saying what I am doing is right but I have now had AT LEAST 10 hours of flying with an unbalanced polar pro and have experienced zero problems..... Even had a very hard... well lets be honest... crash landing with it and everything instill fine!!
Also many of that 10 hours was in Aruba which has constant trade winds of 15+ Mph with gust and sometimes constant up to 30MPH.
[vimeo]https://vimeo.com/100108054[/vimeo]

[vimeo]https://vimeo.com/99885831[/vimeo]

[vimeo]https://vimeo.com/99884359[/vimeo]


I have also gone from 1 flight with the polar pro to the next without and everything was perfect!
 
Hey guys,
Very sad that DJI know of the buzzing problem in the audio of the gopro videos and have not to date yet given us any indication that they are looking into it. Pathetic really. We spend all this money and this is how they treat us. Anyway back to my point. For those of you running FPV on a screen with volume control, if you up the volume on the screen before the phantom 2 is switched on, you will hear the annoying audio sound. I only found this out at home after my first video uploads when i was checking out the black pearl and getting used to its functions.

Some say its caused by the usb and some say its a firmware problem. I have no idea, all i know is that its frustrating as hell. The boys i bought mine from are having the same issues with their P2's AFTER they did the latest update ( before the update they worked fine) I had the audio problem before the update. So i cant wait till someone finds the problem.
As for the Polar Pro ND filter, well that has solved all my problems I run it at 2.7K and sort it all out in my editing programs. Just remember guys, if you use the ND filter on your gopro make sure you use some fishing line to tie the filter to the arm of your phantom as they can come off mid flight and once that happens the gimbal works really hard to counter balance the weight loss on the lens and can damage the motors.
As far as I'm concerned DJI you need some PR work guys. You need someone to let us know where your at with this problem. I have written to them 3 times and not 1 reply. (4 weeks now) disgraceful service really.
 

Members online

Forum statistics

Threads
143,066
Messages
1,467,358
Members
104,935
Latest member
Pauos31