BLD Module Final Fit + Testing

[JohnC]

We are about to fit and test fly the first production standard of hardware and software of the BLD brushless driver modules. I will update this thread as it goes along so those who are interested in the upgrade know what to expect. The plan is to follow the upgrade through to some flight test videos using a Panasonic GH2.

The mount is a regular cinestar 3-axis gimbal.

The first stage is going to be brushless Tilt+Roll with Pan driven by the original servo. As people know, Pan needs more power for the motor than Tilt or Roll and the idea is to see what kind of results can be expected with this configuration. Once this is done the pan will upgraded to brushless as well.

The G-Lock can drive a mix of brushless and servos on a gimbal so the one unit will be driving all 3 axis.

This is the gimbal with the hardware kit of the brushless upgrade installed. It has iFlight 5108 (24 pole) motors on each axis and will be running on 4S. You can see the mounting of the upgraded G-Lock radian module and the connection through the side arm tube which will mate with a signal master module.

If you have any questions, feel free to post them here.

[JohnC]

Ran into the first problem pretty quick : the adjustable tilt bars did not go high enough to allow the tilt axis to be balanced.

The solution was to drill an extra hole in each tilt bar to allow it to be mounted higher up with adjustment only possible on the lower part (as shown on the picture).

This is the gimbal with the camera fitted and balanced in tilt and roll.

_

[Shooting High]

Great to hear about progress John. I have exactly the same setup. GH2 on a 3 axis Cinestar using the same brushless conversion and motors on tilt and roll (Alexmos) with Radian controlled servo & pulley on pan. I must say that when the tilt and roll are super smooth with brushless motors it really shows the shortcomings in the servo based pan. Hopefully your firmware will improve that, but I believe that I will eventually end up going brushless on all three axis.

I had the same problem with raising the GH2 enough to balance it. I ended up using a QMount mini which raises it up just enough to balance and has the advantage that it makes it way easier to mount/dismount the camera and you can easily make small adjustments to the forward/back position of the camera. The mount I used is here.

Keep the updates coming please.

Andrew

[JohnC]

The G-Lock has much better slew acceleration control than the radian and I think it might make a difference when using the cinestar mount with those 'o' rings. I have had comments back which says the pan is a lot better on the G-Lock - so we'll see how well it works. Part of the problem with the radian is that the initial slew movement could kick off some oscillation if the 'o' rings are soft.

Of course it will never stabilise as well as brushless but it's a reliable solution which might be acceptable in trade against the technical problems and annoying lack of holding torque for the brushless.

We will see ...

[Shooting High]

I suspect that part of the reason that people have struggled with finding enough torque to hold the pan axis is that the Cinestar landing legs act as weathercocks. I am going to try mounting the CS3 without the legs using a retractable gear system. I think that there is a good chance that this will reduce the requirement for torque on this axis. It might also help reduce unwanted movement using the servo/pulley system.

[JohnC]

I think the camera-ball might be due a comeback for the brushless gimbals. I never saw much value for servo gimbals - but definitely can now.

[JohnC]

The installation is now complete. The G-Lock has more options for gain control than the Radian did and I tweaked the pan settings a little bit.

It all looks good - so time for a test flight. Unfortunately the workshop here in Thailand is not ready yet so I have to wait until I can get out somewhere with a bit of space (maybe tomorrow or Friday).

[tornvig]

Hi All,

WOW super cool project. I´m  also having some of the "old" useless radians in the draw. I was just about to sell them, but now Ill keep them and wait for your fantastic upgrade 

I made mu own pan update and it works really good with 5208 motor and my 5D mIII. Have a look.

Regards
Bo

Attachments:

[JohnC]

Looks good Bo, it could be the best solution for pan without going to a big (+ heavy) motor for direct drive.

The first batch of production BLD modules arrived yesterday so I've been replacing the development BLD board with the real thing today and hoping to get out for some test video tomorrow ...

[JohnC]

All ready for a test flight but the downlink monitor failed so could only fly blind. Hence it flew with pan turned off for most parts so I could pan using the airframe.

Also this machine has some jello vibration because I have the O-rings packed with some silicon gel to firm it up against oscillation and it picks up some airframe resonance vibrations which is quite annoying.

But apart from that, it's not too bad for a first flight. Looking on the bright side, the vibration is a good test for the G-Lock in less than ideal conditions.

This is from the GoPro on a quad:



And this is from the GH2 on a cinestar mount with brushless tilt and roll (Pan disengaged mostly)

[Shooting High]

This looks very good John. Congratulations. As you say, there is some jello, but the stabilisation looks very good. It would be nice to see some footage in windy and/or fast flying conditions. Do you have any feeling at this stage for price and release date? Will the Signal Master be available at the same time. I'd love to see a special deal for all three components together ;-)
Andrew

[jchiogna]

Will the G lock allow the placement of any camera without further fine tuning or will there have to be  different gain settings for each camera?

[JohnC]

Thanks Andrew. I must say the Cinestar mount with GH2 tuned up much better than the Go-Pro mount. It looks very smooth on the tuning graphs. The jello does spoil it a bit, as does the vimeo video conversion because the original file is better.

We are just working out some details on pricing, dealers etc - but stand by for an announcement very soon. We will probably distribute the first batch directly from Thailand with a special offer upgrade package which includes all modules + G-Lock upgrade, then after that the hardware will just be available through dealers.

jciogna: Brushless gimbals are very sensitive to all PID parameters and it's not easy to make a general setting which works for all cameras. Normally it needs to be retuned each time you change something. I do have some ideas to make this easier to live with in the field which will be released in future firmware upgrades once time allows.

 

[JohnC]

For those who haven't seen it from the CARVEC Systems facebook link, here is a clip from a handheld 3-axis mount with a GH2 camera.

The handheld version is still in the development stage - but works pretty good as it is. The movements of the camera are all done with the smooth-shoot mode (like follow mode).

[JohnC]

Been busy this last week or fitting the brushless motor to the pan axis - and the results are so much better than the servos, I would consider it an essential upgrade. It seems to have eliminated the oscillation on the rubber 'o' rings - presumably because the drive is much smoother and has less torque 'bite' than the servos. This test was primarily for a flight test of the new pan 'follow' mode and I was very pleased with the results considering this ship is not the smoothest and the servo-gimbal video needed a bit of work in post-stab.

I do seem to have a bit of a problem getting vimeo HD video to come anywhere near the smoothness of the original using the Premier Pro presets.