Mini Tri Test Build
- Thread starter Twitchity
- Start date
Mustang7302
Senior Member
Got some renderings from Twitchity for the motor pad dimensions on his booms, toss it into SketchUp and came up with this last night:
I printed the top and bottom half to make sure clearances were sufficient, and so far so good. Still need to model the center pin and test it. Then figure out the alignment for a 9 gram servo.
I printed the top and bottom half to make sure clearances were sufficient, and so far so good. Still need to model the center pin and test it. Then figure out the alignment for a 9 gram servo.
Twitchity
Senior Member
That tilt looks nice Mustang. Is this going to be 100% printed, or will it use a metal pin for the pivot? What kind of 3D printer do you have? That's another item on my list of toys I want to add to my collection, but I haven't really researched the printers that much.
I'll try to get a video of the tri today after work if it's not raining like it's been doing for the past week. Hopefully the video will help isolate the problem I'm having with the tri. So far I've adjusted the P and I values on Yaw, and reduced the looptime to 2000 from 3500 (didn't notice any difference doing that).
I i remember correctly, the values were somewhere around P=4 I=2.2 for Pitch and Roll, P=8 I=4.5 for Yaw. I've reduced the Yaw values to 1.5 (I believe) and .2-.5 on the I and it seemed to have removed the constant wobble, but the tri is still under constant Yaw. I've moved the servo over a tooth to help eliminate the problem, but it just spins the other way. I adjusted the yaw trim on my transmitter and but still no luck.
It appears from the one indoor flight I had last night it gets more uncontrollable the longer it stays in the air. If weather permits, I'll try to get outside and get some video of it flying when I get home from work.
I'll try to get a video of the tri today after work if it's not raining like it's been doing for the past week. Hopefully the video will help isolate the problem I'm having with the tri. So far I've adjusted the P and I values on Yaw, and reduced the looptime to 2000 from 3500 (didn't notice any difference doing that).
I i remember correctly, the values were somewhere around P=4 I=2.2 for Pitch and Roll, P=8 I=4.5 for Yaw. I've reduced the Yaw values to 1.5 (I believe) and .2-.5 on the I and it seemed to have removed the constant wobble, but the tri is still under constant Yaw. I've moved the servo over a tooth to help eliminate the problem, but it just spins the other way. I adjusted the yaw trim on my transmitter and but still no luck.
It appears from the one indoor flight I had last night it gets more uncontrollable the longer it stays in the air. If weather permits, I'll try to get outside and get some video of it flying when I get home from work.
Mustang7302
Senior Member
I was pondering the idea of being 100% printed. It is going to largely depend on how smooth the action is with a 3D printed plastic pin.
I picked up a Solidoodle 4 a few months back when they dropped the price from a grand to around $700. It took minimal fiddling to start getting successful prints with tolerances I have been pretty happy with. I am currently printing in white ABS.
Twitchity
Senior Member
I was able to get the tri out for a test flight. I can't for the life of me get the tri to stay in one place. It's always in constant yaw and drifts to the left. The drifting I can probably fix with a gyro calibration (again), but I'm having problems with the yaw. I've tried adjusting the trim while its in the air, but the tri either stays in a left or right yaw.
Here's some screenshots from BF
Yaw and pitch are off because I was adjusting the settings on my transmitter
And a video... It does have an extremely fast yaw if nothing else.
Here's some screenshots from BF
Yaw and pitch are off because I was adjusting the settings on my transmitter
And a video... It does have an extremely fast yaw if nothing else.
Hey Twitchity what a great lookin tri. It looks like it flies really nicely too despite the issues you've described.
When the servo is centered, does the rear motor have any tilt to it? Are you running all three props in the same direction or are you running one of the front ones counter-rotating? Sorry if I missed this info if it was posted earlier in the thread.
I had a similar issue with my large tricopter due to the servo offset being too high, causing too much of a tilt on the rear motor and the tri hovered with a bit of a yaw angle to it. This was running all three with the same prop. I switched one motor to counter-rotate with the reversed propeller and leveled out the rear motor at center and that seemed to make it go away.
When the servo is centered, does the rear motor have any tilt to it? Are you running all three props in the same direction or are you running one of the front ones counter-rotating? Sorry if I missed this info if it was posted earlier in the thread.
I had a similar issue with my large tricopter due to the servo offset being too high, causing too much of a tilt on the rear motor and the tri hovered with a bit of a yaw angle to it. This was running all three with the same prop. I switched one motor to counter-rotate with the reversed propeller and leveled out the rear motor at center and that seemed to make it go away.
Twitchity
Senior Member
I'm running counter rotating props on the front.
Specs:
DYS 1806 Motors
EMAX 12a ESC
Acro Naze32
1300mah 45-90C 3S battery
5030 propellers (for now until I get it flying right)
This is the exact same parts off of my FPV quad that I disassembled to build this tri, so I know all of the parts are in perfect working order.
I'll play around with the tail servo some more, but I had the rear motor as vertical as I could get it when I attached the servo. I had a larger tri with the KK2.1.5 board and it was extremely easy to setup properly (once I got my motor issue sorted out). I'm sure there's one little thing I'm overlooking for this tri that will solve all of my problems, I just can't think of it yet.
Specs:
DYS 1806 Motors
EMAX 12a ESC
Acro Naze32
1300mah 45-90C 3S battery
5030 propellers (for now until I get it flying right)
This is the exact same parts off of my FPV quad that I disassembled to build this tri, so I know all of the parts are in perfect working order.
I'll play around with the tail servo some more, but I had the rear motor as vertical as I could get it when I attached the servo. I had a larger tri with the KK2.1.5 board and it was extremely easy to setup properly (once I got my motor issue sorted out). I'm sure there's one little thing I'm overlooking for this tri that will solve all of my problems, I just can't think of it yet.
C0d3M0nk3y
Posted a thousand or more times
I think you want the rear motor tilted a little to the left if your rear prop is clockwise, or a little to the right if your rear prop is counter clockwise. I do this by adjusting the servo midpoint.
How are you mounting the NAZE board? Is it mounted on some kind of isolation foam? Try mounting the Naze on a vibration isolation foam. I use the door seal foam from HW store, and mount it with the sticky side to the board and use 3M double sticky tape to mount the other side to the frame.
I had exactly the same issue on my H-Quad with the board rigidly mounted on a frame. The quad would constantly, and annoyingly yaw to the left. After mounting it on some foam the yaw was perfect without any other changes. Give it a try and see if it helps. I would recommend 1/8th foam, or any other way to isolate the board from vibrations.
- Bogdan
Twitchity
Senior Member
The naze is mounted the exact same way as my FPV quad that flew without any problems (nylon standoffs with o-rings under the FC).
I got another video today of the tri. I tried to keep my thumb off of the sticks as much as possible (mainly the yaw). You can see the tri slowly swings back and forth. Is there a feature similar to the servo filtering on the KK boards? My PIDs are the same as above.
I got another video today of the tri. I tried to keep my thumb off of the sticks as much as possible (mainly the yaw). You can see the tri slowly swings back and forth. Is there a feature similar to the servo filtering on the KK boards? My PIDs are the same as above.
Ohhh hmmm that is a different issue than what I was thinking. My bad, ignore everything in my last post regarding the servo tilt angle.
I'm not very good with PID tuning but I would say try adjusting the yaw PIDs.
Maybe one of the following might help:
Upping P by 0.2
Upping I by 0.005
Upping D by 10
I'd suggest testing each adjustment individually. Seems like a very low yaw oscillation which I think means PIDs too low
I'm not very good with PID tuning but I would say try adjusting the yaw PIDs.
Maybe one of the following might help:
Upping P by 0.2
Upping I by 0.005
Upping D by 10
I'd suggest testing each adjustment individually. Seems like a very low yaw oscillation which I think means PIDs too low
Twitchity
Senior Member
Well I finally got back around to messing with the tricopter. I had removed the ESCs to use on my latest quad build, but ended up getting KISS ESCs for that build since they were back in stock a few days after I removed the ones from the tri. After my quad build I started back with the tri, but my soldering station decided to stop working on me. Well my new soldering station arrived last night so I got back to work on the tri.
As far as tuning, I changed the servo pwm rate from 50 to 100 because I'm using a digital tail servo. Outside of that, I reset the board to factory settings to start from scratch. I was still getting the really fast oscillations on the yaw servo so I began adjusting the PIDs. I believe I ended up around P=2 I=.10 and D=40 for my final settings. The tricopter now flies a lot better, but there is still a very minor oscillation and the tri over-corrects ever so slightly coming out of a yaw. As you can see from the video it was rather dark out, but I think I'm finally at a good starting point to do fine tuning.
As far as tuning, I changed the servo pwm rate from 50 to 100 because I'm using a digital tail servo. Outside of that, I reset the board to factory settings to start from scratch. I was still getting the really fast oscillations on the yaw servo so I began adjusting the PIDs. I believe I ended up around P=2 I=.10 and D=40 for my final settings. The tricopter now flies a lot better, but there is still a very minor oscillation and the tri over-corrects ever so slightly coming out of a yaw. As you can see from the video it was rather dark out, but I think I'm finally at a good starting point to do fine tuning.
What tail tilt mechanism are you using?
Twitchity
Senior Member
Not much of an update, but I got some more flight time in with the tri this afternoon. I'm happy with the way it flies so far, but I still need to mess around with the PIDs just a little bit more. I remember seeing something along the lines of in-flight leveling in the latest version of Baseflight; is this their version of auto-tune or something different? If it's auto-tune I would like to give it a try so any information about it would be great. I did a quick search but couldn't find anything about it.
I decided to be a little reckless with the tri this afternoon and bumped up the roll/pitch rate, and RC rate at the same time to what my quad is running (that was a bad idea as you'll see at the beginning of the video. I turned left the roll/pitch rate the same, but turned the RC rate back down to .90 from 1.25 and flew around like that for the remainder of the video. I had another crash (user error) but nothing broke and was back up in the air. I should have kept the camera rolling for the first crash as it didn't break anything either.
I decided to be a little reckless with the tri this afternoon and bumped up the roll/pitch rate, and RC rate at the same time to what my quad is running (that was a bad idea as you'll see at the beginning of the video. I turned left the roll/pitch rate the same, but turned the RC rate back down to .90 from 1.25 and flew around like that for the remainder of the video. I had another crash (user error) but nothing broke and was back up in the air. I should have kept the camera rolling for the first crash as it didn't break anything either.
I'm thinking those are DYS 1806s. I'll bet the lipo lasts a while.
Will those booms handle 22mm motors?
Will those booms handle 22mm motors?
C0d3M0nk3y
Posted a thousand or more times
It's more like an in flight accelerometer calibration for auto level.
Twitchity
Senior Member
Bogdan, I was surprised that nothing broke in that first crash, especially the tilt mechanism. I didn't get the weight for the tri flying in the above footage, but I slapped on my FPV gear just now and took it out for a quick spin. AUW including a 1300mah nanotech batters is 445g. This thing is sluggish (or I'm used to my quad with 2204's and 6045 props), but it's to be expected with the DYS 1806's running 5030 props. I'll be switching to 6045's tomorrow when I head out to this field I've been looking at to do some FPV flying practice.
Cranial, the booms on the tri in the video are drilled for 2204's, but I'm just using the inner holes to mount the 1806's to. This tri is hacked together from various parts from other builds
The arms came off of my Armattan clone quad, stole the tilt mechanism from my large tri (very first multirotor I built), and cut the frame from scrap G10. Take a look below at the CF version Since I don't have anymore 2204's (and my hobby budget is tapped since I had to buy a new soldering station this month), I cut the booms to have two 2mm holes to mount the DYS motors, and two 3mm holes for the 2204's. Once I upgrade the motors I can use a round file to enlarge the 2mm holes to 3mm so I won't need to cut a new set of booms.
Code, thank you for the information on the in-flight leveling. I was pretty sure that's what it was, but I had hopes it was for an auto-tune Do I just select the box and it's activated for auto level tuning, or is there more to it?
And now more pictures! This tri is loaded down... I don't think I could fit another piece on there if I needed to. Also... Don't pay attention to my battery connector... I ran out of XT-60 connectors and this was the next best thing. I must check the connection 5 times before I plug in the second wire each time I go to fly this thing. Like I said, this tri was slapped together with whatever I could find And yes... my work space is a mess.
Mini CF Tricopter. I was trying to go for an Apache Helicopter type of theme with this little thing.
Sitting on top of my quads FPV pod
Top shot: You can see the 2mm and 3mm holes for the 1806 and 2204 motors really well in this shot
And with 6" props for size reference
Cranial, the booms on the tri in the video are drilled for 2204's, but I'm just using the inner holes to mount the 1806's to. This tri is hacked together from various parts from other builds
The arms came off of my Armattan clone quad, stole the tilt mechanism from my large tri (very first multirotor I built), and cut the frame from scrap G10. Take a look below at the CF version Since I don't have anymore 2204's (and my hobby budget is tapped since I had to buy a new soldering station this month), I cut the booms to have two 2mm holes to mount the DYS motors, and two 3mm holes for the 2204's. Once I upgrade the motors I can use a round file to enlarge the 2mm holes to 3mm so I won't need to cut a new set of booms.Code, thank you for the information on the in-flight leveling. I was pretty sure that's what it was, but I had hopes it was for an auto-tune
Do I just select the box and it's activated for auto level tuning, or is there more to it?And now more pictures! This tri is loaded down... I don't think I could fit another piece on there if I needed to. Also... Don't pay attention to my battery connector... I ran out of XT-60 connectors and this was the next best thing. I must check the connection 5 times before I plug in the second wire each time I go to fly this thing. Like I said, this tri was slapped together with whatever I could find
And yes... my work space is a mess.
Mini CF Tricopter. I was trying to go for an Apache Helicopter type of theme with this little thing.
Sitting on top of my quads FPV pod
Top shot: You can see the 2mm and 3mm holes for the 1806 and 2204 motors really well in this shot
And with 6" props for size reference
Looks like it has a limit of 5" rotors with the cage attached. Will it handle 6" rotors as a convertible?
Edit: wait are those 6030s????
Edit: wait are those 6030s????
