View Full Version : Discussion ATTOpilot camera trigger set-up
professor
Jul 14, 2009, 06:33 AM
Good morning, I am now reliably and consistently flying waypoints thanks to all of you! I installed the camera in the data bay of my uas and plugged the actuating servo into port #7. I modified the WP and SET files to the best of my knowledge and yet, no pictures. Can this installation be tested on the ground somehow? Are my code files correct for camera trigger using a micro servo mounted to the top of a camera? Thank you!
WP Latitude Longitude Alt Triggers Airspeed checksum
$00000,40.58248,-86.84979,0200,00010032,50*CS ' WP entrance Prairie Acres
$00001,40.58231,-86.84134,0200,00010032,50*CS ' WP back intersection Prairie Acres
$######## END OF SETTINGS ############
Tom Harper
Jul 14, 2009, 06:52 AM
Professor,
The trigger is the fourth position from the right. You are using the fifth:
$00000,40.58248,-86.84979,0200,00001032,50*CS ' WP entrance Prairie Acres
$96 and $97 are the pulse periods that will drive the servo. 2100 is probably over driving the servo, and the two have to define two positions. On my set up the rest is 1100 and activate is 1900.
On the ground, when you switch to AUTO the trigger output will be in the REST position.
A '1' in the fourth position will take a single image at the WP.
Tom
Mecha
Jul 15, 2009, 12:05 AM
I believe the fourth digit from the right is where The professor has it (the fifth one from the beginning). I do not know what the last two digits are for, where you have the 3 and 2.
$87-$90 are different time intervals you could set for different waypoints
and $91-$94 are for different distance intervals again for different waypoints.
In mine I take a picture every three seconds so I have
WP Latitude Longitude Alt Triggers
$00000,38.14508,-76.42837,0075,00002000*16
and on the setfile I have
$87, Trigger time interval (2) = 3
If you want to take a single picture at the waypoint then you will need
WP Latitude Longitude Alt Triggers
$00000,38.14508,-76.42837,0075,00001000*16
and set file does not matter
Your main problem as I see it, is the pulse width on line $96 and $97
The servo pulse should be more or less from 1000ms to 2000ms, where 1500ms is the servo center. Having both limits on 2100ms does not allows the servo to travel in ether way. On mine I have 1600ms for the active and 1500ms for rest. I use an HS56MG and only move about 3mm to depress the button.
I also see these are wrong in your set file, make sure that your servo limits are set correctly:
$24 set it to 2000 and $31 to 1000
Testing the setup
If you set the trigger for a time interval like in my case, you can test it on the ground by just turning the system on. It should start taking pictures right a way. Another way would be to get the coordinate of a point on your street and walk the plane from your home to that point at which it should take a picture. Make sure your altitude is set to 1 or 2 meters.
I hope these helps
Tom Harper
Jul 15, 2009, 08:46 AM
Mecha,
He is uaing the forth digit from the left.
Tom
professor
Jul 16, 2009, 01:38 PM
I really appreciate your help guys. Still no pictures, but I am using a waypoint out in my driveway so I don't have to fly the whole mile circuit each time. That is saving wear and tear on those Lipo's. Navigation is so smooth and accurate with this ATTO that I am really excited to get the camera trigger figured out. Here is my WP coding and I modified the SET text as noted above: WP Latitude Longitude Alt Triggers Airspeed checksum
$00000,40.58377,-86.84723,0001,00001000,55*CS ' WP Driveway for camera test ONLY! Do not fly!
$######## END OF SETTINGS ############
$00000,40.58248,-86.84979,0200,00001000,55*CS ' WP entrance Prairie Acres
$00001,40.58231,-86.84134,0200,00001000,55*CS ' WP back intersection Prairie Acres
Mecha
Jul 16, 2009, 02:05 PM
Post your SET file again to look at it. And make sure you do not have any more $ sentences after the $######## END OF SETTINGS ############ in the WP file.
Is the servo moving at all? Did you try to have it on a time sequence rather than at the waypoint? try the following just for testing purpose
$00000,40.58377,-86.84723,0001,00002000,55*CS ' time test for camera ONLY! Do not fly!
$00001,40.58248,-86.84979,0200,00002000,55*CS ' WP entrance Prairie Acres, Do not Fly
$######## END OF SETTINGS ############
On the set file change lines
$87, Trigger time interval (2) = 3
$95,Trigger Type = 1
$96,Trigger Active = 1300
$97,Trigger Rest = 1700
$98,Trigger Duration = 2
This should move the servo every 3 seconds and hold for .4 seconds
If you already changed lines $24 and $31 and the servo does not move then I would assume a bad servo, or it is plugged incorrectly to ether Atto or the receiver.
If it works then change the 2 for the 1 on the WP file and it should do it.
professor
Jul 17, 2009, 11:28 AM
I changed all the parameters as suggested above. Attached are my WP and SET files. I used 2 different servos and have replaced the lead with a new one. This is a new set-up and has never taken photos. Any help appreciated.
Tom Harper
Jul 17, 2009, 12:22 PM
1300 - 1700 may not be enough to activate the camera servo.
One way to test this is to plug the camera servo directly into the throttle socket on the receiver (not Atto). Power up the transmitter and rcvr and move the throttle control, and watch the servo response. If you have an osscilloscope you can measure the activation and rest points. If not you can estimate them. The throttle lever is linear, if you have not programmed it in the transmitter. With the idle lever down, the throttle will move from 1000 to 2000 over it's range. You can estimate the position of the lever when the servo activates the camera or you can count clicks to get a pretty accurate measurement.
The above exercise will also tell you which value to use for 'rest'.
Keep at it you're almost there!
Tom
Mecha
Jul 17, 2009, 02:48 PM
Tom's suggestion will definitely tell whether the servo works or not and the range of motion.
Professor- Please let me know if i understand the problem. The servo does not move at all, Regardless of the WP file having a 00002000 for time sequence, or a 00001000 for waypoint trigger.
with the 00002000 the servo should begging moving after Atto has had a home position locked. Make sure you are getting a GPS lock and have the mode in Autonomous, (disconnect the motor).
the SET and WP files look good. You can try the servos maximun range 900ms-2100ms
professor
Jul 17, 2009, 02:49 PM
Thank you Tom for the encouragement. I forgot to change the 1 to a 2 in the WP text for the last test. I did that and the camera takes pictures every few seconds. That rules out camera, servo, wire, or ATTO. I am now waiting to fly with the code returned to 1 to see if it triggers at the waypoint. I am very hopeful. Thank you.
Mecha
Jul 17, 2009, 02:55 PM
Excellent!!! Look forward to see some aerial pictures.
Tom Harper
Jul 17, 2009, 06:41 PM
Let's see the pics!
professor
Jul 18, 2009, 08:54 PM
Thank you guys. I took the plane out for a flight and it performed perfectly. The pics weren't that great due to the light conditions but the picture shows the exact GPS point I selected from Google earth. I have no idea why news crews don't use UAS instead of helicopters. I am flying what I like to call "enhanced visual" meaning I keep the plane in sight but am at the limits of my vision. I was doing this with a 10 foot span plane packing two large cameras but that is much easier to see at 1000 feet and 1/2 mile away than my little 6 foot development plane. I have attached the pics.
I am considering offering a course here at Purdue University on UAS Hardware and Applications. I have 6 years of struggles to share with the new batch of kids. We have deployed the 10 foot span model for roughly 50 missions with hundreds of photos taken for a myriad of agricultural purposes. It is not autonomous but uses a gps downlink in real time to a laptop for manual location and taking of the pictures. It has worked well and is very reliable with the onboard gyros taking care of the attitude at extreme heights and distances. I am basically a voice activated autopilot when we are flying that plane. Thank you again guys. I am off to experiment with some ideas I have for applications of this technology in agriculture.
he last picture is a joke that Homeland Security may or may not find humorous.
Tom Harper
Jul 18, 2009, 10:05 PM
Congratulations professor!
Let me know if I can contribute. I'm using Atto for Archaeology.
Tom
tekrunner
Jul 19, 2009, 01:36 PM
You guys know if there's any special settings required if you're using an IR prism switch instead of a servo?
Tom Harper
Jul 19, 2009, 02:52 PM
What's required to actuate the device?
tekrunner
Jul 19, 2009, 06:50 PM
What's required to actuate the device?
When I hook the IR switch into channel one of my radio receiver and move the stick to the left the switch activates and a picture is successfully taken. I'm going to test it with Atto here in a minute and give an update.
tekrunner
Jul 19, 2009, 07:51 PM
Well I set atto up with a servo just like the professor and it moved every 3 seconds in the same direction as if the servo was plugged into channel one and the stick moved to the left. However when I hooked up the IR switch it doesn't activate when it's plugged into atto instead of the servo. So the IR switch isn't getting exactly the right signal from atto as when it's plugged into channel on of my receiver.....
tekrunner
Jul 19, 2009, 08:19 PM
OK, I just changed $96 and $97 from the professors settings to 1000 and 2000 respectively and everything works great now. Quite a sight to watch.
Tom Harper
Jul 19, 2009, 10:10 PM
Fantastic - Looks like you've got it!
tekrunner
Jul 19, 2009, 10:42 PM
Fantastic - Looks like you've got it!
Thanks, it's quite a sight to see the camera autonomously taking pictures every five seconds. Can't wait to fabricate a camera box and get it flying!
vBulletin® Copyright ©2000-2009, Jelsoft Enterprises Ltd.