Just thought I'd share my first experience into building multirotors, which happens to be the Flite Test bat bone. It might help someone new avoid some of the sillier decisions I made while on this endeavor, and it might help me learn some new things as well. I've been flying the tricopter for about two weeks now, and I think there's enough to go on for a post. I had no flying experience with anything prior to this besides a Nano QX, and just started the hobby.
Initially, I went with the recommended 3s 2200mAh Park300 motors with 8045 props setup on the store, and this is what it looked like built:
It took me two days just to get the tricopter not to spin like a beyblade, and a week of tuning and practice to get crash-free flights. I hadn't done enough research on what PID settings do, and I didn't realize how reliant acro mode was on them. My friend who is getting into this hobby with me had no experience with multirotors and started his build with a Naze32, which flies really well on stock settings. Meanwhile, I was still trying to get the KK2 not to yaw left while flying forward...
I installed the FPV equipment sometime in the middle of process, and broke the tail on the FPV maiden. I learned during this crash that PID settings matter a LOT, especially for the rudder on a tricopter. Also, I realized that my motors were underpowered, and a linear throttle curve is not good. I had my max throttle set to 100%, which meant there was no room for the gyros to compensate for the yaw at full speed forward.
Also, please don't try flying full speed forward on your FPV maiden like me. Unless you like destroying motors and tilt assemblies. The aftermath, after plummeting tail-first into the grass:
I decided at this point that I needed more lift, for that extra wiggle room for oh sh*t moments, and to carry my gopro and FPV equipment around. meehan on the Flite Test forum also wanted to load a tricopter with full FPV gear, and I used his/her thread as a starting point.
After that, I replaced the motors, esc's, and props with SunnySky X2212 KV980 IIs, Multirotor Superstore 20amps, and 1045 Gemfan props. I also cranked up some of the P and I settings for the yaw, and set the max throttle to 80% on my Tx. That resolved most of the yaw issues, but there was still some drift that I couldn't get rid of, which I eventually learned was due to the KK2's gyro not reporting perfectly or something due to changing conditions. The added lift also gave me the power I needed, and flight times were still around 9minutes before the 10.5v alarm starts going off. It also sounded amazing!
Then on my second set of FPV flights at the football fields I stripped the tailservo in a crash. (pilot error) Goddamn it.
At this point, I was already set on upgrading to a Fortis Titan frame due to the better crash resistance and the stronger, less sloppy yaw mechanism, so I didn't feel like getting another expensive digital servo just for the Batbone. I biked to the nearby Radioshack, picked up a $12 analog servo, and epoxied the servo horn to my stripped Turnigy's horn. Even with half the torque, it seems to fly well! Just had to crank up the servo filter in the KK2 a little bit and mess with the PIs a little. I'm expecting it to explode one day though.
So here's the current state of my Bat Bone. After flying and crashing a lot for two weeks, I'm really happy that the booms and front motor mounts hold up pretty well to crashing. I'm a bit disappointed in the strength of the tilt mechanism though— the landing gear is integrated into the mount for the servo, so over time even with light crashes it eventually breaks the tilt mechanism's servo mount.
Above is the remainder of the broken landing gear from the front, CA'ed back together many times. I zip tied rubber weatherseal on them in the hopes that they would break less easily. Even though the delrin plates are supposed to pop out in the event of a crash, what usually ends up happening is tiny beams in the delrin plate points will break. I feel like there's a few points in the plates that could have been reinforced way better to make them more crash resistant, but that would have added slightly more weight as well.
I got tired of waiting for parts that break frequently with my flying inexperience, so I decided to design and print print my own RCExplorer style landing gear for the tri. After some crashing I've confirmed that they break properly at the 4inch zip ties that keep them on the booms, and take seconds to minutes to put back together on the copter. The tilt mechanism's landing gear cannot be removed as it's reinforcing the servo mount, so I made sure all the printed legs were 2cm longer so that they are more likely to take the impact.
I'm still tuning the PIDs, but I'm going to switch the flight board to a Naze32. Getting the KK2 to this point felt like more work than it was worth, and I feel like I'm starting to see the limits of how well it can fly (not that I'm anywhere close to them yet). Even on stock settings, the Naze32 can already be more stable than my KK2. Even now, I still haven't sorted out a tiny bit of yaw drift that I think is the result of the gyro's 'no movement' reading constantly flickering between 511 and 512. I'm running the SteveIs 1.6(?) firmware.
That's it, and I hope someone finds this valuable or entertaining to read. Note though, that I am a terrible pilot and everything I fly will break four times faster. Well, at least now have a bunch of Park300 1360kV motors sitting around for.. something. I'm not sure I'd go for a KK2 on another build again though, and I'm pretty sure my next build is going to be a mini hex or quad. I'll have more stable videos once the guts on this gets moved to a Titan and I get the tuning right, but for now, here's some LOS testing footage. Good night!
Make sure servo wires are well connected:
Initially, I went with the recommended 3s 2200mAh Park300 motors with 8045 props setup on the store, and this is what it looked like built:
It took me two days just to get the tricopter not to spin like a beyblade, and a week of tuning and practice to get crash-free flights. I hadn't done enough research on what PID settings do, and I didn't realize how reliant acro mode was on them. My friend who is getting into this hobby with me had no experience with multirotors and started his build with a Naze32, which flies really well on stock settings. Meanwhile, I was still trying to get the KK2 not to yaw left while flying forward...
I installed the FPV equipment sometime in the middle of process, and broke the tail on the FPV maiden. I learned during this crash that PID settings matter a LOT, especially for the rudder on a tricopter. Also, I realized that my motors were underpowered, and a linear throttle curve is not good. I had my max throttle set to 100%, which meant there was no room for the gyros to compensate for the yaw at full speed forward.
Also, please don't try flying full speed forward on your FPV maiden like me. Unless you like destroying motors and tilt assemblies. The aftermath, after plummeting tail-first into the grass:
I decided at this point that I needed more lift, for that extra wiggle room for oh sh*t moments, and to carry my gopro and FPV equipment around. meehan on the Flite Test forum also wanted to load a tricopter with full FPV gear, and I used his/her thread as a starting point.
After that, I replaced the motors, esc's, and props with SunnySky X2212 KV980 IIs, Multirotor Superstore 20amps, and 1045 Gemfan props. I also cranked up some of the P and I settings for the yaw, and set the max throttle to 80% on my Tx. That resolved most of the yaw issues, but there was still some drift that I couldn't get rid of, which I eventually learned was due to the KK2's gyro not reporting perfectly or something due to changing conditions. The added lift also gave me the power I needed, and flight times were still around 9minutes before the 10.5v alarm starts going off. It also sounded amazing!
Then on my second set of FPV flights at the football fields I stripped the tailservo in a crash. (pilot error) Goddamn it.
At this point, I was already set on upgrading to a Fortis Titan frame due to the better crash resistance and the stronger, less sloppy yaw mechanism, so I didn't feel like getting another expensive digital servo just for the Batbone. I biked to the nearby Radioshack, picked up a $12 analog servo, and epoxied the servo horn to my stripped Turnigy's horn. Even with half the torque, it seems to fly well! Just had to crank up the servo filter in the KK2 a little bit and mess with the PIs a little. I'm expecting it to explode one day though.
So here's the current state of my Bat Bone. After flying and crashing a lot for two weeks, I'm really happy that the booms and front motor mounts hold up pretty well to crashing. I'm a bit disappointed in the strength of the tilt mechanism though— the landing gear is integrated into the mount for the servo, so over time even with light crashes it eventually breaks the tilt mechanism's servo mount.
Above is the remainder of the broken landing gear from the front, CA'ed back together many times. I zip tied rubber weatherseal on them in the hopes that they would break less easily. Even though the delrin plates are supposed to pop out in the event of a crash, what usually ends up happening is tiny beams in the delrin plate points will break. I feel like there's a few points in the plates that could have been reinforced way better to make them more crash resistant, but that would have added slightly more weight as well.
I got tired of waiting for parts that break frequently with my flying inexperience, so I decided to design and print print my own RCExplorer style landing gear for the tri. After some crashing I've confirmed that they break properly at the 4inch zip ties that keep them on the booms, and take seconds to minutes to put back together on the copter. The tilt mechanism's landing gear cannot be removed as it's reinforcing the servo mount, so I made sure all the printed legs were 2cm longer so that they are more likely to take the impact.
I'm still tuning the PIDs, but I'm going to switch the flight board to a Naze32. Getting the KK2 to this point felt like more work than it was worth, and I feel like I'm starting to see the limits of how well it can fly (not that I'm anywhere close to them yet). Even on stock settings, the Naze32 can already be more stable than my KK2. Even now, I still haven't sorted out a tiny bit of yaw drift that I think is the result of the gyro's 'no movement' reading constantly flickering between 511 and 512. I'm running the SteveIs 1.6(?) firmware.
That's it, and I hope someone finds this valuable or entertaining to read. Note though, that I am a terrible pilot and everything I fly will break four times faster. Well, at least now have a bunch of Park300 1360kV motors sitting around for.. something. I'm not sure I'd go for a KK2 on another build again though, and I'm pretty sure my next build is going to be a mini hex or quad. I'll have more stable videos once the guts on this gets moved to a Titan and I get the tuning right, but for now, here's some LOS testing footage. Good night!
Make sure servo wires are well connected:
