Hi, I've uploaded a clip shot at 2.7K to youtube. Please check around 20-25 seconds in, you will see a wavey wobble in the imagery. Can you please advise what is going on since I also have another problem:  "Gimbal disconnect" errors and lost camera control in the app.   I'm using a M600 + RTK + X5. I've tried contacting support via mail but keeps getting "default answers back".

It seems like it could be vibration jello, does it seem like the aircraft vibrates a lot? Do you happen to have an Inspire or a M100 that can rule out an issue with the camera itself? Are all components up to date? Did you happen to notice this out of the box or after a firmware update or a hard landing?

Hi. thanks for the reply. Video quality has been like this out of the box and there aren't any significant abnormal vibration from the M600 during flight (On the contrary I'm surprised at how stable the craft is but I did notice some of the props tighter than others). Don't have any other aircraft to test on. All components still brand new (Only 6 hours of flight time total). Also the gimbal disconnect issue is quite annoying for mapping application.
I see there is another thread also discussing this at http://forum.dji.com/thread-64186-1-1.htmlAre the remedies mentioned there worth any try?

I've checked the difference in tension between the blade sets and found 3 arms that has 1 blade that is stiffer than the others. Also on one motor, the blades are much looser than on the others. As stated, I've flown around 6 hours with 60 landings (none of which hard). Still have to look into the blade tensions.

Moving the camera closer to the center of gravity of the craft does NOT resolve the effect. Also flying into the wind seems to increase the jello effect which has me thinking it can also be some sort of aerodynamic buffeting that is contributing.

Only success I've had so far and it "seems" to reduce the gimbal disconnect instances as well is the following: I've measured the vibration coming through to the gimbal plate (sprung plate onto which the camera mounts) and did an Fourier transformation on the frequency data. One of the harmonics is at exactly 30 Hz in the Z-Axis (up/down). This would explain the jello at shooting at frame rates close to this number in fps since the camera is vibrating at the same frequency as the frame rate.  I then added 85g of weight onto the centerline of this gimbal plate and did another shot. It seems that this reduce 80-90% of the problem since the 30Hz peak has come down to 13Hz. My reason for going for the 85g is to match the weight difference between the X5 and X5R. Are there stiffer rubber dampers available? It is worrying that these natural frequencies and harmonics are coming through from a design perspective.

This is just and improvement and not a resolve of the issue! Still looking for answers and solution

Jello almost gone. Retorqued the propellers to loosen them a bit since some of the blades were tighter than their opposite on the same prop. Found a lot of white flakes from the washers upon removal causing the tension issue. I will test wether this eliminate X5 Jello as soon as weather permits.

Now on the disconnect issue: iSkyUnmannedSys mentions the same connection error I'm having over at http://forum.dji.com/forum.php?m ... =1&extra=#pid637697 (#7)

This is getting really frustrating and I'm starting to lose customers. Please help me getting this sorted. I purchased the setup in good faith and after careful research and I really don't want to regret my decision.

Propellers re-torqued a little looser - craft flies a bit quieter but jello still prominent in video footage.
Seems a lot more when flying into the wind that downwind.

Had to restart the M600 2 times due to camera settings showing N/A in the app (Whenever this happens I can still control the gimbal from the RC as well as touching and dragging in the Go App). When I select info, it shows gimbal disconnected. One of these restarts gave the error immediately after turning the power on.

I've been in discussion with support since 27 December 2016 but they keep asking me default questions which I am now beyond and the person I'm talked with has now gone quiet since 3 January 2017.

What must I do to get my system functioning properly. I cannot use it to do what I purchased it for and video recording is also not usable. Please help!

Seems like I'm having a bit of a monologue here.

Any idea where I can get stiffer dampers for the gimbal mounting plate. I added the foam earplugs (cut in half along the axis) as suggested elsewhere on the forum into the gimbal vibration dampers which seems to transform the jello into pure camera vibration/shake. This means that I will (hoping to) find the sweet spot with stiffer dampers on the mounting bracket. With these plugs installed, you can clearly see the vibration coming from the drone on the footage but the jello effect is gone. I did take video of the motors in flight and there are NO noticeable vibration from them. I can upload the video if you want to have a look. I will also add a 2 kg payload to the M600 to see if that makes a difference when flying tomorrow as this should also reduce vibrations from the M600 in flight due to the higher inertia.

Doing research and development on behalf of DJI here it seems.

Jello still persists with no resolve.
Gimbal/Camera disconnect issue still persist with no resolve.

I am now running out of options!
What can I do to get this sorted out?

Any help out there would be appreciated.

4 January = last feedback, can someone from DJI please take interest in the problem and help a customer resolve a problem with an unusable system worth $12,000.

Here are some  footage shot today illustrating the jello problem I have: https://www.dropbox.com/sh/zm2qq ... 72rK0fPb8CXmJa?dl=0Tested various camera shutter settings - 60 - blurred footage, higher settings jello prominent.

DJI PLEASE! It is frustrating to turn customers away because my investment does not work as advertised!

lt Posted at 2017-1-14 12:09
DJI PLEASE! I'm sick and tired of turning customers down because my investment does not work as advertised! The silence will not magically make the problem disappear! Own up!

I can't believe your not getting support, hard to believe with the money we spend on these tools..

Another day with camera disconnect and more jello in the shots. I've uploaded another clip with jello as well as footage of the motors whilst in flight as well as on the ground to the link provided in #11

DJI please advise!

DJI please advise!

DJI please advise!!!

DJI, can you please advise!!! I feel ripped off. Seriously cannot believe that this is how you treat customers that spent good money on integrated products only to have the problems I'm experiencing.
It cannot be that difficult to help me sort out the problems: 1 - Camera disconnecting during missions/sessions, 2 - Jello whilst taking video (hovering, flying slow speed and high speed).
The evidence provided is surely proof enough that the problems do exist.

Hence I cannot use the system reliably even if I wanted to and believe me I seriously like the way the M600 flies but at the moment it is completely worthless to me.

If this is an isolated case, at least admit it and work with me to assist in resolving the issues. If it is more widespread which I think it is, please also help us sort it out.

https://forum11.djicdn.com/data/attachment/forum/As suspected and mentioned previously in #4 in which I used an external accelerometer with the M600 stationary there was a peak in vibration at 30Hz coinciding with the 30fps. I've attached an FFT of the dataviewer of a random flight and this seems to confirm the vibration I got. This seems to be the cause of the Jello in the videos. What do you suggest to get rid of this. I know adding mass will bring down the frequency but surely there must be some other means of solving this because the vibration should not have a peak at this frequency.
Same data prevalent in FFT done on Gyro Data

I do not fly a M600 but I do have a M600 Pro with X mount and X5 using the 15mm lens.  I have not seen the issue but normally I film at 60fps.  Have you tried to vary your recording frame rate?  What is the result?  To answer some of your questions, there are no different vibration isolation mounts.  I would like to see a picture of how you have the X mount mounted on the M600 rails.  Depending on whether it is an early version or one of a later versions.  I have noticed the difference in the tightness of my props as well.  However it has not produced any jello.  I am an old single rotor helicopter pilot and set the props by feel.  We have posted the torque settings.  However it requites a micro torque wrench and they are not cheap.  

Regarding the gimbal disconnect issue.  What device are you using, iOS or Android?  What model?  Typically the disconnect is device related.  You can also click on the HD icon and use custom channels and reduce the transmission rate to the lowest available.  That will also help.  I apologize that we have not assisted you.  However, lets see if we can help going forward.  

Hi Tahoe Ed. Thanks for the feedback.
With regards to the Jello/Vibration:  I've tried all the tricks in the book/forum/internet without success. Higher framerates produce the jello as seen in the videos uploaded. Lower framerates seem to reduce the jello but introduce a vibration blur (normally in the z-axis). I was hoping that this would be a cure but there are no sweetspot I could find varying the framerate or any camera settings. From an engineering standpoint if vibration exist 30hz you will get harmonics at 60 and 120hz as well. With regards to the dampers - I tried adding those foam earplugs (as suggested elsewhere in the forum) as well as the cable tie route but this only increase the vibration/jello in my case. I also tried increased the sprung mass (adding 85grams to the gimbal plate) with no reduction in the vibration/jello.  The only thing I havent tried was to add weight to the M600 frame and this should reduce the natural frequency at the cost of flight time hence not a practical solution.
With regards to mounting: I've tried moving the whole mount backwards (more towards the CoG) but in my opinion this only increased the issue of seeing the props in the image when flying forward and had no significant effect on the vibration/jello.
If possible, can you redirect me to the torque settings required for the prop tensions. I did try them with a looser setting (by hand) but reverted to the default torque setting as per the markers on the bolts.

Camera/Gimbal disconnect:  What happens (at completely random intervals) is that the camera works fine one moment and the next all camera settings change to N/A. If I then exit the app and reopen, The whole camera panel is missing. Most of the times I can still see the image and can move the gimbal with the R/C and in the Go App but no possibility of taking a photo or video when this happens (3rd Party apps - Map Pilot also issues a Camera/Gimbal disconnected error when this happens and I then have to abort the mission). Only way to rectify this currently is to do an RTH and reboot which is not ideal when doing survey missions because it's time and battery life gone. I did try different cable routing options but no difference in the effect (Initially thought it could be bad cable connection on either the CAN or Video cable).
Device in use: I'm using an iPhone 7+ but have used an iPhone 5S and iPad as well as different cables between R/C and the device, also with no noticeable difference in the frequency of disconnects. I also have tried manually selecting channels and reduce the bandwidth but this is not ideal as it will have a random impact on range. The issue has also happened with the M600 stationary.

As you can see I'm at a point where I'm left a bit frustrated (didn't expect to have these issues) but still hopefull to resolve the problems and once we get this sorted, we'll be in the market for 2-3 more units or one more M600 and an I2 and Mavic.
I appreciate your input in this matter and really hope that we can resolve the issues. Thanks in advance for the assistance

Just had another failed video shoot at a game farm with the words "JELLO" written all over the footage (less than 1 minute of usable footage out of 20 minutes of video and also picking up a blur in the photos now with the rest just plain garbage and I refuse to accept the fact that I have to live with this "issue" when paying $1700 for a camera and around $12,000 for the complete kit). Left completely disgusted not being able to resolve the issue and having no input on this from DJI apart from the standard apology and some random unrelated questions is adding fuel to the fire. If this is too much of a technical issue which I can understand (I'm a mechanical engineer) please refer me to the right people that will be able to resolve the issue. Maybe voodoo or some other dark magic chant will work better than asking for advice on the forum.... (I know sarcasm doesn't really fit the forum but frustration is driving me through the walls with the issue at the moment).

Back to the facts and more proposals (I think I'll send a bill to DJI for trying to get their issue with the M600/RTK/X5 working):
The only other suspicion I have causing the Jello is unbalanced props. I'm trying to find my milligram scale to check variance in weight between the blade sets. Please let me know if I'm on the right track or else I might set fire to the M600!

Please feel free to edit or delete this post or ban me but in doing this you will only be admitting that the problem is not being addressed by the appropriate person from DJI. I sincerely and honestly want the product to work as per specification and enjoy having made the purchase and generate a good income stream but at the moment I'm losing my mind, losing customers and losing credibility.

hello It,
When I first filmed with the M600 original I too had much vibration and as a designer of F 1 vehicles I have plenty of test equipment to play with. I first moved the camera platform under the CoG as you have but this was done to reduce the leverage effect in any transmitted X,Y movement then I balanced the camera gimbal, I added 3g to the horizontal arm and 5g to the rear face of the camera body because I suspected I had motor stutter, vibration between poles on the gimbal motor. this procedure reduced 80% of the "jello" I then looked on hobby king and bought the stiffest vibration dampers I could buy and then put tie-raps through the dampers and pulled tight to reduce most of the movement, 90% vibration gone. I then removed the props and put a iPhone vibration app running to measure the relative frequencies agains each other, as this was a direct comparison between motors it did not matter about the accuracy of the device, it just had to have repeatability. one of the motors had a higher reading than the rest so I added aluminium tape to dynamically balance the motor than ran the motor tests again at four speed steps to see if it was more balanced. 95% fixed. Now only when I have the lens pointed to the ground and do a high G lateral movement will I see "jello" I don't see how any of the cameras including my high-end cinematic cameras can ever be vibration free given they are hanging under a huge source of vibration with air density  and wind velocities all doing their best to bugger things up. I will say the bigger the gimbal motor is and the more poles it has the better the footage is. We are at the cutting edge of this technology at present and it must be excepted despite the desires of DJI to make a perfect film platform it will never happen so we will always compromise something. Should DJI make false claims  no, but they are a business and the game is to make money how ever they are able and a few bent truths seem to be the norm nowadays.
I will say in defence of the company they do try to help when possible, however it is in their time span not ours.
You seem switched on technically so keep testing and reading the forums and you will get to a place where you can use the tech to meet your needs, with or without DJI.


ricci

Hi Ricci, thanks for the suggestions. Will evaluate and revert. I fully understand that a drone is basically a "vibration machine" and natural frequencies and harmonics can cause serious havoc if even one part/component is out of spec if the issue is marginal and in best efforts most people DO get usable footage. Personally I feel that the dampers provided are a compromise on cost/performance but then again who would pay upwards of $1000 for a professionally designed damping solution that is perfectly tuned to the circumstances and equipment and I do not blame DJI for this. It is just the way things work in business.
My problem being that not even post production can fix the erroneous problems on the footage do confirm that I do have a problem (Jello as well as what I suspect the same vibrations causing random disconnect of the camera). I have also been thinking about adding mass to the M600 to reduce the frequency of vibration. As a fellow engineer and R&D man doing analysis and development I really appreciate your feedback. I can also understand that this might be an isolated case but merely acknowledging this from DJI's side alone does not cure the problems.
I can fully relate to your detailed analysis and I will definitely try it out (just didn't expect it to go this far in terms of technical details because it is marketed in a different manner).

I do believe that DJI is doing themselves harm by not properly educating the people who are supposed to help customers with the product queries and neither do they have a referral service where we can directly make contact with someone in their technical/engineering department to try and sort out these problems. Most new drone users have what I call "teething" problems if any mostly related to not fully understanding the workflow and operation and expect it to work straight from the word go and this is where the forum can serve most people perfectly but I do feel that some of the problems encountered by some of the people buying the equipment is way more in the technical realm and understanding and do need engineering input from the designer or engineer should a problem like mine arise. Sure you can relate to this.

I am merely venting (maybe unnecessary) but justified due to the fact that I did not foresee the level of technical inputs required from myself and also the fact that I am being completely ignored on the support site. My support request has been unresolved for 25 days and last communication from DJI was 11 January suggesting I do the firmware upgrade already done and mentioned in the discussions prior to the comms and then marking the issue as "RESOLVED". If the person I am dealing with do not have the skills or knowhow, the least I expect is an acknowledgement and referral to someone who does have the knowledge.

I just want to reiterate as well that I am really impressed with the Matrice 600 + RTK as a package and it gives the operator unmatched confidence in operations. I just ended up with a system that is brought to its knees by 2 "stupid" problems that are not being resolved by DJI.

Hi,
just mounted new X5R on my M600 and have same problems.... but i have a very-very strong feeling that these wobbles are some kind of electro magnetic interfering. i am cameraman (25 years in business)  and i had similar thing when brining Betacam camera near to strong magnetic field.
i have a damper plate from Ronin MX attached to M600, so i have dumpers from Ronin plate and dumpers on X5R itself.
+ i have a micro vibrations on the video when using 25mm and 45mm especially. will try to upload video later. 100% sure that it is because of the mechanical issue.

Hi Dmitry_R. I also went down the same EM interference avenue and you can also test it out. Take video at high shutter speed and you will get the effect as per video uploaded (jello). It seems that the effect is most prevalent in fully automatic shooting mode. The moment you reduce the shutter speed to 90 or 60, the jello effect changes into periodic vibration giving the video a stable look but you will then notice a blur in the vertical axis when looking at the footage (You will also get the vertical blur even though it is just 1-10 pixels vertically when the vibration occurs and you take a photograph). I've also read elsewhere that people are using an ND filter to reduce the effect but (and I stand to be corrected in this regards) it makes no sense since using this, people will only mask the problem and not remove the vibration.

This leads me to rather believe that it is vibration related. I don't want to discount your experience as a cameraman so you can try it out for yourself and see what happens - suggest you use Autopilot or Litchi and program repeatable runs to evaluate the different shutter and F-stop settings to see what works best as a compromise but you will definitely see the jello/blur with the different settings.

I will also keep fooling around until I have this resolved (as per ricci2's suggestions which makes completely sense from a mechanical perspective).

Best of luck and lets keep pushing until we get this resolved because it is very frustrating spending time and money to go out on a shoot or survey and then returning egg-faced because you end up looking more like an amateur and a fool than a pro that knows what you are doing.

lt Posted at 2017-1-21 08:53
Hi Dmitry_R. I also went down the same EM interference avenue and you can also test it out. Take video at high shutter speed and you will get the effect as per video uploaded (jello). It seems that the effect is most prevalent in fully automatic shooting mode. The moment you reduce the shutter speed to 90 or 60, the jello effect changes into periodic vibration giving the video a stable look but you will then notice a blur in the vertical axis when looking at the footage (You will also get the vertical blur even though it is just 1-10 pixels vertically when the vibration occurs and you take a photograph). I've also read elsewhere that people are using an ND filter to reduce the effect but (and I stand to be corrected in this regards) it makes no sense since using this, people will only mask the problem and not remove the vibration.

This leads me to rather believe that it is vibration related. I don't want to discount your experience as a cameraman so you can try it out for yourself and see what happens - suggest you use Autopilot or Litchi and program repeatable runs to evaluate the different shutter and F-stop settings to see what works best as a compromise but you will definitely see the jello/blur with the different settings.

Hey lt,
i guess my aim is the same as yours are: to find the solution.  and very fast.
why i think it is not mechanical issue: after i noticed the problem we took M600 up in the air, get the Jello, land it, stop the motors - and the Jello was still there! on stoped motors on the ground.
dont have any video but i saw it and my camera assistant did. more then that. in the office i switch on M600 and there were no jello at all ! i took it up in the air today and there was no jello at all for 20 min flight at the same place.

this jello with motors off makes me mad....

+ micro vibrations is there!!!

will move the gimbal to the center and fly tomorrow.

Good luck with your efforts and keep them posted. Its driving me mad as well.

I will also continue trying to find a solution. From a mechanical viewpoint it might also be related to the way the props are set from one flight to the next (I do doubt that this could have a huge effect but it is a possibility) as well as one of the props being a bit looser than the other on the same arm, because I also have more and less jello sometimes from one flight to the next. I'm still waiting on Tahoe Ed to reply with the correct torque settings to set the props.

There might also be an aerodynamic effect (buffeting) at play at higher speeds when flying into the wind. I have noticed this a few times but can never repeat the problem to get the effect.
If the actual root of the problem cannot be identified everything else we try is based on calculated speculation so lets keep trying to identify the actual cause of the issue.

ps. You will notice that when moving the gimbal towards CoG that the propellers will be in the footage when filming straight forward, but in your case, I think it will be worth a try (This should change, as Ricci2 stated, the X-Y swing motion/moment if you are doing aerial surveys).

Thanks lt,
here is a sample of the video
https://www.youtube.com/edit?o=U&video_id=eu4W_lXSdVM
will post more if there will be any success.

Hi chaps,
Here is something to try, I have just pointed a high speed camera at the gimbal mount for M600 and where it locates at the inner cut-out section where the tongue is located I have noted a z, axis movement extreme enough the tongue touches the top and the bottom of the location tab. I have inserted low modulus silicone foam to damp the movement now but have not flown yet, I have now put plastic bolts and washers through the vibration dampers to compress the rubber to the point of binding. Inspired by your problems I have looked at my setup and noted the mounting plate is very stiff in the forward and aft directions but has noticeable lateral deflection. I would recommend trying to make the mounts as solid as possible then carry out these tests to identify the force vectors effecting the camera. take off and face the lens forward and focus on a horizontal line and also a vertical line like a rugby goal, acquire footage to see if you have rolling shutter, then rotate the m600 90º and rotate the camera 90º to the vehicles forward axis and point the lens again at the same horizontal, vertical axis, record and analyse footage to see which has the greatest shutter roll. which ever is the greatest damp that direction on the camera mounting base. we have huge vibrations on the cars and tests like this help guide us in suppressing them. regarding EMF, use some MU-metal found in old crt monitors as shielding around the electronics to the base of the m600 above the camera or failing that add a 0.5mm thick sheet of coper as a shield and ground it to 0v if possible, if there is an issue there it will soon vanish.Love to know the results as I am grounded for a week or so due to location.

happy hunting

ricci

i

hello Dmitry_r
Just read your observations on the rolling shutter at standstill which reminded me I had some very high frequency movement of the gimbal motors which could be felt on your fingertips when the camera was put in certain positions. I had come across this on an earlier drone and gimbal setup so I spent time dynamically balancing the gimbal with the lens cap off, as mentioned above. I put 3g on the horizontal arm and 5g on the rear of the camera so now in any position the camera is balanced and holds its location when powered off. if you do find the camera is not holding a position at its extreme throws add flattened lead fishing shot cut to fit and glued in place. one note, turn on the m600 for 5 mins to warm the gimbal motors up, I found cold motors damped out of balance situation.
This might sound like a small issue, however all the little changes are leading in the correct direction hopefully and as every M600 is diferent we all face our own challenges unless there is a fundimental engineering design fault then watch out DJI the people will bite back.....wont we????

Ricci

lt Posted at 2017-1-9 11:29
Seems like I'm having a bit of a monologue here.

Any idea where I can get stiffer dampers for the gimbal mounting plate. I added the foam earplugs (cut in half along the axis) as suggested elsewhere on the forum into the gimbal vibration dampers which seems to transform the jello into pure camera vibration/shake. This means that I will (hoping to) find the sweet spot with stiffer dampers on the mounting bracket. With these plugs installed, you can clearly see the vibration coming from the drone on the footage but the jello effect is gone. I did take video of the motors in flight and there are NO noticeable vibration from them. I can upload the video if you want to have a look. I will also add a 2 kg payload to the M600 to see if that makes a difference when flying tomorrow as this should also reduce vibrations from the M600 in flight due to the higher inertia.

To make you feel worse, the rolling shutter effect you have is far worse than one gets from a USD 1K Mavic even in 8m/s winds.

If all you say about vibration is correct, it does logically boil down to having resonance node that given wind speed, aircraft speed and that payload (X5) is at the X5 giving vibration. You have addressed this already in a previous post.

As you suggested, try a different cam. Also given that the main harmonic is at 30Hz try a different frame rate.

The damper balls seem to be of only one type. This is not the main issue. I would not put screws etc.
Try shifting weight around the craft. I'm sure you'll hit the sweet spot for your config.

If this was a systematic error, you bet your life others would complain.
As for the disconnects: try wobbling the cable connecting the camera to the flight control system. If your signal breaks up there, you have an answer. It drove me nuts on an Inspire and I thought this was a firmware update but it turned out to be a cable issue. Fixed and no problems since


Hope you solve your problem!

Hello Mike the cat,
I think others could be having such issues with rolling Sutter if you look at the read count of this thread, maybe folk are just excepting of such issues and try to do post editing to remove some of the shake.
Would be interesting to know if this is isolated or more widespread?

ricci

Here are some semi-successful feedback based on the suggestions above with notes.
1: Added 3g to the horizontal motor - This balances the camera nicely.
2: Added 5g to the back of the camera - I removed it later because with the standard lens the 5g to the back makes the camera point upward immediately. Mileage may vary but I think this is lens dependent and would be required if mounting one of the other lenses.
3: Did test flight with just the 3g to the gimbal. Footage still had Jello present.
4: I then had a good look at how the X5 gimbal mounting bracket was designed and something hit me (I think Ricci also mentioned something to this effect): The way the fixed plate is mounted directly onto the drone create high stiffness in a horizontal plane but being a single sheet of carbon fiber with foam inside there is flexibility in the vertical direction due to buckling/bending and if vibration coming through from the M600 is at the right frequency, the plate will resonate in this motion. This in combination with the arms going down to the dampers cause these arms to swing and vibrate in the horisontal plane whenever the plate flexes and then translates into a vertical vibration through the dampers to the sprung plate onto which the camera mounts. I hope this makes sense and can draw a sketch to explain this if someone wants a better understanding.  This motion made me think and I then added an 85 g lead weight strip (those they use to balance mags) towards the front side of the fixed plate. This immediately took away about 60% of the jello I have been experiencing. Taking the concept and reasoning further, I added an aluminum strip to create stiffness in the front of this plate and moved the weight towards the back of the plate resulting in 90-95% reduction in jello). I will get a new weight strip tomorrow because I want to try a 100 grams to see how far the envelope goes.

After this I did a few trial runs taking footage at varying shutter speeds and found jello present (albeit 90-95% less) with the automatic settings on the camera but bringing down the shutter to anything longer than 500 results in almost no jello. Didn't test shutter speeds shorter than this but will do so tomorrow. Even though this is  a much better scenario and I find the footage "acceptable", I personally feel that mileage may vary when going out and doing filming in real world scenarios and I will report on the issue if it occurs again. I would classify the issue as improved but not resolved.

I also feel that DJI should seriously look into this and improve the design of the bracket because I feel that if it is this sensitive to vibrations, there must be some sort of revised design released to customers. A DJI Customer is not supposed to be in the business of finishing purchased products that are marketed the way they are currently if the design is this sensitive to manufacturing variances and vibration from the M600.
Video link for youtube

        

hey lt, thank you for sharing and digging in!

thank you ricci2 for your participation!

i did some tweaks and it went well.

1. i moved the gimbal to the centre of the aircraft

2. i put back the front bracket

3. i put 2 washers and 2 ND filters for balance the gimbal with 45mm olympus lenses just for the test

and i have cage for ronin mx installed with extra tough ruber bumpers.

all this dramatically improved the picture. no wobble for the first 10 min, but then it came back. i had to stop because of the snow and after landing found lots of ice on the arms and body, maybe it was the reason of the wobble after 10 min?

in the whole i would say that the picture is more o less usable now. the wobble is practically not noticeable, especially on movement. i think on 25 mm and 12-15 mm the picture will be good (i hope so).

will do more test and post here the results.

Great news on your fix Dmitry. It seems that we both had a busy drone day today. Just hope we have a long-term fix here! I'll post the moment something gets worse on my side. Happy flying!

Hi chaps,
This is why I love this subject, the people who fly are design engineers and best of all they are honest and share. My type of people.
Dmitry I use an Olympus mz 12-42 with a uv filter and a weight adaptor Dji supply, they have adaptors for all lens types,  saves using ND filter which might compromise the lens settings, I think it was 10 euros.
I can weigh it if that helps.

ricci

Hi Chaps,
After looking at your pictures and then my craft I have found another place of weakness, the diagonal arms connecting the bottom plate, the ones facing forward, they offer very little lateral stiffness and if it were possible a diagonal brace between the arms would help tie the beast together. due to space and its odd shapes this is not an easy mod but I have some 4mm ally tube with brass threaded inserts from an old frame that can be cut to fit inside the front and rear arms and screwing in place adding some rigidity to the weak design. I will put the horizontal tube at mid distance on the stays to firstly divide the flex by 50% while transfering loads horizontaly and also half (aprox) any hamonics transmitted through them. will post pictures upon completion.

Dmitry you said you had ice buildup. When we do low temperature thermal testing in a test cell we coat the camera and equipment that would be effected by the ice in a water repellent coating that slows down the ability of the ice Cristal to grip on the metal parts. There are a few products out there that can do this, one is called "rainx" used on car windscreens or if you are in a fix and need something quick you can use shaving gel, teflon spray or teflon dry lube, silicone spray or even spray floor polish as it contains silicone.  Prolonged exposure to dense water saturated air will win eventually but a coating might give you more time in the air. how bad are your winter temperatures?
just thinking out loud...:-))

ricci

ricci2 Posted at 2017-1-22 19:17
Hi chaps,
This is why I love this subject, the people who fly are design engineers and best of all they are honest and share. My type of people.
Dmitry I use an Olympus mz 12-42 with a uv filter and a weight adaptor Dji supply, they have adaptors for all lens types,  saves using ND filter which might compromise the lens settings, I think it was 10 euros.

Hi Ricci,
i have 4 different fix lenses and i need to use ND filters to keep shutter speed not more then  100 (normally 50, PAL Area) and iris 5.6 - 8 for the best picture look.  my plan is to get more cheap lens hoods, drill a hole and use tiny bolt with washers for balancing the gimbal. or something like that. tests-tests-tests.. so dont bother to weigh your adaptor, but thanks for the offer.      

Hi Ricci

Thanks for your comments. This is exactly the movement that got reduced by adding stiffness to the top plate of the bracket. I know that I'm not eliminating all the movement coming through the down and forward facing arms and had the same thought about putting a brace between them. I think and might also try just putting a lasercut plate between the 2 forward arms that can tie them together at the dampers. Same with the set at the rear. This will have the effect of them vibrating together instead of opening and closing that translates into vertical movement of the sprung plate. I think this is one of those areas we discussed previously where DJI had to make a compromise in terms of cost vs efficiency and this is kind of biting them back on the grounds of the jello I experienced. The 2 sets of 2 arms should basically form a "U" shape each and if not rigidly connected at least be forced to vibrate synchronously in the same plane.
Once again thanks for your input. I am a much happier customer now that the jello seems to be gone or should i say reduced because inherently the design would still introduce vibration.
Also like the idea of coating the electronics with some repellent and will try it out sometime.
PS. The fixes provided also gives a reasonable explanation why I got the gimbal/camera disconnects. Haven't had one since this leading me to believe that the disconnect originated at the connection/contacts caused by the same vibration that create the jello. Will also report back if this is not permanently resolved.

Rgds
LT