Post your videos as file attachments here. One video per topic please!

Moderators: sg, mods

User avatar
By DMarley
#398569
red wrote:
dbotos wrote:
red wrote:How high is the voltage that you feed to the servo?
Red,
Just 5 VDC from the Arduino so far. The servo can take 4.4 to 6.6 VDC. Guess I could see if a higher voltage supply for the servo would help.
David
David,

I would suggest using a straight 5Vdc from the power source, with only the signal voltage coming from the gadget. An oscilloscope might show that servo signals are dragging down the power to the servo, as is now. This power issue may or may not be causing some jitter at the "level" spot.

In RC aircraft, constant "corrections" by the servo can drain the batteries. I used a three-pin voltage regulator chip, 11Vdc in and 5Vdc 2A output, to stop that jitter. It was similar to this part, but feel free to innovate. US$.68 each. 8)
https://www.digikey.com/product-detail/ ... -ND/585989

For the chip I used, the regulated voltage would "float" upwards, no load, but a small-load (high resistance) resistor would hold the 5Vdc steady, when there was no other load.

:mrgreen:
Yes!

What I would try first is to provide 2s voltage (~8.2 v) to the processor and imu (the arduino uno can accept 7-12 vdc) and another battery supplying the servo with whatever it requires, but connect both battery grounds together. If that helps solve most of the problems, then provide a stepper or linear voltage reg to one battery to power the servo.... or just get a high-voltage servo. As is now the servo is most likely drawing down the voltage to the processors and creating a sort of ripple.
User avatar
By dbotos
#398571
red wrote:Everybody in your game needs an Oscilloscope. Here is a DIY oscilloscope that uses your PC sound card, and a few extra pieces. The guy mentions Hantek, if you want an off-the-shelf commercial solution, still much cheaper than a fer-real commercial oscilloscope. Have fun . . . 8)

http://homediyelectronics.com/projects/ ... illoscope/
-
Red,

Thanks. At some point I'm sure I'll need one. Free to sub-$100 is pretty good considering what the Tektronix DPOs like what we have a work cost. :shock:

David
User avatar
By dbotos
#398572
DMarley wrote:Dave,
Using arduino ide delays to attempt to correct problems just creates more problems and prevents the processor from doing things efficiently. Instead of delays, try to do a moving average calc to smooth out the outputs from the imu. Graph your inputs vs averaged inputs vs time and see what works that way. The raw inputs are probably scattered a bit. You'll have to find a running average set length that will provide reasonably smooth curves yet yield fast enough response rate. That is if putting the servo on it's own power supply doesn't correct the problem.
Try using the Atmel Studio IDE rather than the arduino ide. You'll be glad that you did later on if you make the switch. It's not the easiest to print things, but it has a much better debugger, and the whole of the ide is professional and lightyears beyond arduino's poor attempt.
You may find that wrapping the servo wires around a torus ring a few times may help reduce signal interference.

All I want for Christmas is my two front teeth .... oscilloscope and a logic analyzer!
Doug,

Doing the moving average is a good idea. Bouncy in gonna be bouncy out unless you do something to smooth in between. :thumbsup:

I'll have to check out that Atmel Studio and see if I can dig up a torus.

Thanks,
David
User avatar
By dbotos
#398574
Another thing I'd like to figure out (that Doug's mention of rolling average reminded me of) is how fast does the system need to be? If glider roll rate is X deg/sec (for typical recreational flight), how much faster should the gimbal system be? Is the response dominated by the actuator speed (i.e. the time to take input from the IMU, calculate a correction, and send that to the actuator is so much smaller than the actuator movement time that it's negligible)?

I'm kind of thinking of it like a constant game of tag between the glider and the gimbal. The two are standing there and the glider runs off without warning. The gimbal has to be watching the glider's position to say "hey, glider's not standing right next to me anymore, but is now x units away". Gimbal's brain has to process that and send a message to its legs to go catch glider as quickly as possible. But gimbal has to factor in that during the chase, glider's position could be still be changing from that initial position of x units from start. And gimbal doesn't want to run past glider and start some sort of oscillatory pattern.

In reality it's more like gimbal trying to stay in one place and glider's pulling on the rug beneath its feet, pulling it along to some other position. But I think the tag analogy is more straightfoward and could be used to show the potential benefits of PID control:

proportional: glider is x units away; therefore gimbal needs to run certain distance or speed

integral: glider has been just out of arm's reach for a while - gimbal needs to move closer or pick up speed

derivative - glider is running at certain speed with head start; gimbal must extrapolate position to intercept glider based on gimbal's faster running speed. If glider slows down, gimbal should slow down too so as not to run past.
User avatar
By DMarley
#398585
Placing the imu on the camera's mount is going to complicate your pid calcs quite a bit I believe, since now the system is a positive feedback loop, rather than 'neutral' feedback. This method will either make you get up close and personal with Harry Nyquist and PID theory, or you will just end up with no hair.
Placing the imu on the glider-side will be much less math overhead for the processor, and less for you as well. I would start there, then if you decide that you want to continue further, then place the imu on the camera side. There is plenty of PID related programs for the 328p out there that you probably don't have to completely generate your own code. I'd first search Adafruit's site, then Github, Brokking, Andreas Spiess, etc.
User avatar
By dbotos
#398616
DMarley wrote:Placing the imu on the camera's mount is going to complicate your pid calcs quite a bit I believe, since now the system is a positive feedback loop, rather than 'neutral' feedback. This method will either make you get up close and personal with Harry Nyquist and PID theory, or you will just end up with no hair.
Placing the imu on the glider-side will be much less math overhead for the processor, and less for you as well. I would start there, then if you decide that you want to continue further, then place the imu on the camera side. There is plenty of PID related programs for the 328p out there that you probably don't have to completely generate your own code. I'd first search Adafruit's site, then Github, Brokking, Andreas Spiess, etc.
Doug,

Keeping the IMU on the glider side definitely keeps thing simple from both a physical and a calculation standpoint. No wires going across the rotating interface to the IMU and the calculation is essentially "if the glider is rolled X amount this way, roll the camera X amount the other way". One potential downside, if your actuator has no position sensor (encoder, potentiometer, etc.) of its own, is that you're actuating open loop and you don't really know if the camera ended up where you told it to go or where it is at any particular moment. But maybe that isn't critical if the nature of the glider dynamics and the actuation system get along okay. I think you'd still want some input smoothing (like the running average you were talking about doing on the IMU roll angle) and/or an output positioning tolerance (e.g. ±1 degree or whatever chosen tolerance you want the camera to be level within), so that it's not doing micro-motions responding to "noise" from the IMU input or trying to get to some perfect output angle.

I got my NEMA 17 stepper set up the other day and turning with one of the example sketches. Didn't do much other than that since I actually got to fly this weekend. :thumbsup:

Maybe I'll throw the IMU back on the stator side of the servo test box and see how well I can get it to balance that level vial.

David
User avatar
By dbotos
#398746
No gimbal tinkering lately, but I did get to test out my new lower-profile static camera mount on the keel of my Falcon this weekend. On Memorial Day weekend, I stuck it on the keel for the first time, but with that mini ball mount still on there. It was fine, but I realized the ball mount was just adding weight, complexity, and size, so I got one of these little two-hole aluminum mounts (lots on eBay / Amazon) and put it right on the aluminum channel:

http://www.dx.com/p/aluminum-alloy-skat ... TbcSk2weHs

I swapped in one of the knobs that came with my SJCAM (same M5 x 0.8 thread), which have the finger-tightenable knob, but also a Phillips screwdriver recess to get it even snugger if needed (I employed the latter).
Attachments
2017_06_Jun_04_F5_2_02_resize.jpg
2017_06_Jun_04_F5_2_02_resize.jpg (104.43 KiB) Viewed 3444 times
low_pro_cam_mount_left_6-4-17.jpg
low_pro_cam_mount_left_6-4-17.jpg (209.57 KiB) Viewed 3444 times
low_pro_cam_mount_right_6-4-17.jpg
low_pro_cam_mount_right_6-4-17.jpg (174.62 KiB) Viewed 3444 times
User avatar
By dbotos
#399546
I may be ready for that gimbal soon. Giving myself a harness wedgie yesterday in about a 40-degree banked turn:
Add description
Add description
7-9-17_F4_banking_over_launch_resize.jpg (157.62 KiB) Viewed 3271 times
I later worked on making more gradual turns. :thumbsup:
User avatar
By dbotos
#399591
Thanks. Do you mean side to side-to-side or rotating? I try to keep my eyes on where I want to go despite body shifting. So if I'm shifting my body/head to the right, I'm probably turning my head slightly to the left to keep my eyes on target. Kind of like an azimuth tracking gimbal. Lateral head motion may seem amplified by the fact that I'm pretty straight up in that training harness so my head, being higher up, is closer to the downtubes.
User avatar
By DMarley
#399607
Hey Dave. Nice launch and landing.

I'd probably be a bit tight as well in your shoes and skill level, attempting to land between those corn rows!
Look up PIO (you probably already have). It's an exciting resonant frequency response. And since you're a quick-minded, technical kind of guy, here's a good mechanical description of the problem:

Resonance occurs when a structural system is forced at a frequency corresponding to one of the natural frequencies of the system. These are frequencies at which the system naturally likes to oscillate. If the forcing frequency corresponds to one of these natural frequencies, the force and the response of the structure are in phase with one another, so positive work is done by the force, and the energy in the oscillation increases. Eventually, damping and nonlinearities in the system limit the forced response.

An anti-resonance occurs at a frequency between two resonances where the response of the system due to oscillation from the lower natural frequency is out of phase with the oscillation due to the pilot's higher frequency control inputs (jabs or bumps). The resulting oscillations then tend to cancel each other. ...But then, you probably already knew that. :)

So.... all ya gotta do is input your responses at a much higher frequency than that of the pendulum and the glider system. Anticipate the roll, and merely bump it quickly, or just ignore the oscillations if not of large amplitude.
Spending a whole, good-weather week at Steve's would probably highly accelerate your learning curve, too. Then you could come play at Big Walker Mountain! Wooo Hoooo!

Read the comment by the author of the vid....
https://www.youtube.com/watch?v=JmJM7Uv ... 3F&index=1
User avatar
By dbotos
#399608
Doug,

Thanks. It certainly could be PIO. I'm still getting the hang of how much input is enough (and when), especially off tow.

I'd love to do a week's worth of training - I just don't trust the weather to be good for that long. Especially now that I'm back to being a wind direction snob, since I've outgrown the shorter E/W runway. 8) Sunday was my 9th lesson at Blue Sky (also did two on the dunes at Outer Banks), so hopefully I'm getting close. I'd like to start working on trying to land in a defined section of runway or quasi spot-landing on a given target. Hopefully I can get out to the mountains this Fall.

David
By harris
#403540
Hi,guys.you can try this new gimbal which born for mobile filmmakers-zhiyun smooth 4 [www.zhiyun-tech.com/smooth4].]Really good features and nice price.only $139 ,but full with amazing functions such as vertigo shot, roll-angle time lapse, focus zoom in and zoom out.