Mustang7302's iQuad Build Log

Cyberdactyl

Misfit Multirotor Monkey
Oh yes. And there's an added feature to partial interior reinforcement of the boom. With care, you can add a small but significant load shedding nuance by slightly (the tapered tip ~95% the dia. of the tube's I.D. for commom CF tubing) tapering the interior reinforcement at the predicted failure point. It allows the point of failure to shed load over a wider area before catastrophic failure occurs.

This can also be mimicked by using softer reinforcing materials such as soft wood or hard foam.
 
Last edited:

Mustang7302

Senior Member
Naming/painting/cleaning up your flying toy implies a dedication. Such an implication may be useful someday. ;)

Does your Q-Brain get warm? My ESCs do and more than I would like. I will rebuild my Knuckle I copter to get better airflow to them.

I should think the standoffs might make the copter a bit more efficient having the sensors closer to the same plane as the rotors. They may also increase air flow to the ESCs. I expect they also make the copter more flexible offsetting the new efficiencies in a 'push' of sorts.

You might consider CyberD's 3M foam pad/tape solution to allow you to eliminate the standoffs.

I like the hybrid. Zip tie landing gear and wood for cheap build/fix and vibration reduction on the booms. CF for rigidity, performance and weight reduction. You took the knuckles out to reduce the air dam effect on a small copter and you have a craft that weighs 220 grams less than mine, has more power with similar motors and the same size rotors, gets similar if not better flight time and uses a smaller, cheaper battery.

Damn fine build Mustang.
:applause: :applause: :applause: :applause: :applause: :applause: :applause: :applause: :applause:

It isn't easy to get in to feel the Q-Brain, but from what I can tell is that it does get warmer than ambient, but no where near being concerned about. The SK3 motors also get a bit warm after flying around for a battery, but again no where near being concerned about.

The stand offs may indeed be a push, for the exact reasons you listed. I have noticed that the frame has a bit of torsional flex imparted by unexpected collisions with the ground. The flight controller never seems to care however if the frame is flat or not, it just shifts how it sits slightly in a hover. Regardless, this is something I want to address.

I really like the simplicity of the zip tie tie downs for the flight controller, but I just dont have the space for it. I have some double sided carpet tape (think duct tape on both sides) that I am going to use to afix the foam to the top CF plate and the FC to the foam pads. Then I will remove the hook side of the velcro from the ESC to make for enough clearance to sandwich the CF plates down to the frame's support booms.

One of the biggest goals with this build was repairability and getting in the air without the need to keep pre-manufactured spares on hand or waiting a week or two for parts to come in. After that was just making it as simple as possible. The knuckle plates were eliminated to reduce the need for 12 more holes to drill in the frame, and then making sure it all lined up right; although there is a small weight savings there too.

10 to 11 minutes seems to be the sweet spot with the 1800mAh 3S battery. That didn't however keep me from running one too long, just enjoying flying this setup so much, and puffing one of the cells in the pack. I have three 2200mAh 3S Zippy Compact, which should weight very close to what the Turnigy Nano 1800's weighted, on order from the HK US warehouse now. As a part of remounting the KK2.1 I mentioned above, will be checking the polarity direction for the voltage monitor and getting it hooked up.

The light weight helps a ton with making flight times more efficient, but I've already had a number of crashes where I would expect to have broken a prop blade at the hub or busted delrin landing struts; so far the only real damages has been a crack in on CF prop about 3/4" from the end when it made contact with a pole sticking up (which still flew just fine afterwards!).

This setup has been months in the planning with lessons learned from my own experience, taking in lessons you've learned in your builds, and many from the great community on the FT forum.

Very nice project. Looks good. I also like the idea of using ties as a "ratchet screw". :cool:

I have been working on a full carbon fiber H-frame that uses no holes whatsoever on the booms or main spars.

I found 1.5mm CF will deflect a bit if semi-cantilevering plates by clamping one side and having no other support for several inches. Also, the 10mm square tube HK sells will oval slightly under hard clamping pressure unless you get the screws snug against the tube. The really like the square tubes and they are a decent price, but it would be nice if the inside round diameter was maybe 1mm less.

HQuad.jpg

That is an interesting idea for sure. I see a torque wrench being a good idea to make sure you get snug enough to keep everything clamped together without uneven pressure and warping. In your next post with the internal support to help spread to point of critical failure is also a great idea. If I ever get around to doing a mini-quad, I might use some of these ideas.
 

Mustang7302

Senior Member
Got around to getting some video from this past weekend with the little 808 #16-D keychain camera attached to my hat or to the belly of the quad.

In this video the camera is on my hat, but is pretty boring. I found that the quad in the wind has a lot of oscillation on the stock settings, so I did some tuning lowering the P and I gains by a bit each time I put it on the bench. Other than that I was working on trying to incorporate more swooping into my flight and spent a lot of time learning the muscle memory for trying roll with yaw and yaw with roll (more roll with coordinated yaw works well). There were a couple of soft impacts and one I thought for sure would have broken something on previous builds.


The second video has the camera strapped to the belly, up against the battery and a piece of foam between it and the battery strap. It is interesting to see the footage from this view and was happy that there was no jello in the video. This video shows more practice swooping around and has a pretty significant crash, which it got back into the air from too.


There is still more tuning needed, especially with the wind, and I will work on it little by little. As long as it flies well enough to enjoy, right now that is the mode I am in.
 

cranialrectosis

Faster than a speeding face plant!
Mentor
Isn't it nuts how the one obstacle in the field always manages to reach right out and make its presence known?

:mad::mad::mad: The shrub to the right hates you. :mad::mad::mad:


Beautiful copter, Mustang.
 

xuzme720

Dedicated foam bender
Mentor
Isn't it nuts how the one obstacle in the field always manages to reach right out and make its presence known?

:mad::mad::mad: The shrub to the right hates you. :mad::mad::mad:


Beautiful copter, Mustang.
I keep warning people about how sneaky and fast trees are. Shrubs are just their evil minions! They have somehow gotten the PVC involved in their machinations, now!
 

xuzme720

Dedicated foam bender
Mentor
Could be...I almost had some payback here... Things begin to go south at almost 2 minutes in (I think they had a jammer or scrambler) and I manage to get a cow in my sights around 3:30 but the plane got scared just before retribution!

 

Mustang7302

Senior Member
I've put in a dozen or so flights over Christmas, but after that it has been pretty bleak for flight time between weather, other projects in spare time, computer games, and the wife's prerogatives. It has flown very well and quite resilient to crashes, which is most of what I had hoped for.

However, there has been oscillation with mild to moderate wind gusts which causes the flight controller to work too hard and takes away from efforts of more aggressive flying and swooping about. Despite an brief attempt a tuning the PI gains, I could not make much of a difference in the oscillation when faced with gusts of wind; to be clear though I have very little experience in gains/limits tuning.

Over on the RCGroups forum Rolf Bake, the original designer of the KK2, has chimed in that he is no longer a part of the KK project and detailed a number of issues that Hobby King did. Most of the flaws are not fatal, but I think observable with my experience to date. There is a V1.6 firmware available which may help this oscillation issue, but other people smarter than I am have taken to the code and started running away with making updates.

One of these really smart people is Steveis, which has released a hand full of new firmwares making changes along the way. His latest firmware tackles the calculations with the accelerator/gyro to help dial in auto level mode, among other changes. His updates along the way can be kept up with his blog: http://www.rcgroups.com/forums/member.php?u=367321

Even though he has released firmware V1.12S1Beginner (which he a "KK2 look alike" for the sake of easy setup and tuning), I just flashed the prior version for the advanced tuning options (not that I know what to do much with them though). As of this evening that version is V1.11S2. To flash the KK2.1 I used the Arduino microcontroller, like I did for the KK2 in my Bat Bone build, by using the guide with the following link for the pin layout to the ISP header: http://arduinodev.woofex.net/2013/02/11/how-to-flash-kk-2-0-using-an-arduino-uno/. From there I pulled down the latest beta version of LazyZero's AVR Programming application and used the AVRDude utility via command line. LazyZero has also been kind enough to include the hex files from Steveis' work, which made it pretty straight forward to flash with.

With the command line in windows I change the directory to the path that AVRDude is nested in, after having decompressed the runtime package. From there I use the application to acquire the hex file, then manually moved it to the same folder the AVRDude utility is in. Lastly the flash process is as simple as this one line of text in the command line: "avrdude.exe -P COM4 -b 19200 -c arduino -p m644p -v -e -U flash:w:KK2V1_1V11S2.hex:i"

Even though the last two paragraphs may seem like Greek to some people, there will be some who have a little experience on the matter and can sort through the bumps if they go the less convenient route of using an Arduino as the flashing interface. If someone with an Arduino is really trying to go through the effort of flashing, I will see about setting up a more detailed tutorial at a later date.

Since the only 3S battery I had in arms reach when flashing the KK2.1 was a partially charged 1300mAh, I still took it out for a brief flight. The wind was pretty calm with the sun being set, and I was flying by the glow of the lamp by the back door in my small back yard. But from the couple of minutes I noticed that the iQuad felt much more locked in. There almost no sway/drift when changing pitch/roll position and was a different kind of smooth feeling. I hope to get it back out tomorrow after work with a few 2200mAh batteries and really give it a run down; more feedback to follow.
 

Craftydan

Hostage Taker of Quads
Staff member
Moderator
Mentor
Sounded more like Geek than Greek . . . but then again, I'm fluent in geek ;)

Stevis's builds are awesome, but I'm waiting for RC911's port of 1.6++. His mixed acro/SL mode alone makes me feel the 2.1 board is inferior to the 2.0 board . . .
 

Craftydan

Hostage Taker of Quads
Staff member
Moderator
Mentor
Last Tuesday he posted he was testing it, so "soon", but he's a pretty thurogh tester. I'd say any day now, but more likely near the end of a weekend.
 

Mustang7302

Senior Member
Went out on this beautiful calm morning to do some tuning on the KK2.1. I had previously lowed the P and I gains and limits to get rid of the minor bobbling oscillation when centering on auto level mode; but found it made it loose the locked in feel and was mushier. So I figured I'd default the board, set it up again, and start looking at bumping PI gains and lower self level gains.

To be clear, there has been damage done in the past. I've cracked the rear motor boom from the end towards the center for about 3 inches long; I jut put some wood glue in the crack and clamped it shut for about 10 minutes, all has been well since.

But on the second test flight feeling out the gains in acro mode, I had the copter get away from me and I chopped the throttle to keep from damaging me or anything else. The resulting impact landed on the front motor boom breaking off one side and sheering the prop adapter off the motor on the other side.

uyyf.jpg


So I will be ordering a new SK3 1275KV motor and putting my spare motor boom into action over the next week or so. I also have some GWS props from Altitude Hobbies on order to use when I know there will an increased risk of crashing; oddly no damage to the CF props in this last crash.
 

Mustang7302

Senior Member
It was designed to be easy to repair from crashes. If the prop adapter hub hadn't sheered the crews into the motor's bell, I could be back in the air in about 10 to 15 minutes of that crash. Its the ordering of new parts which gets to me; waiting for them to come in.
 

cranialrectosis

Faster than a speeding face plant!
Mentor
Ouch!

Sorry Mustang. Order two motors and have a spare. Eventually you will lose a bearing or get jumped by that post again and want it.
 

Mustang7302

Senior Member
The new motors came in today! Pretty impressed with HK and the USPS for getting an order placed on Saturday in my box on Wednesday with a federal holiday in between.

Took about 25 minutes to disassemble the hardware and motor boom, swap motors on the ESC harness, drill six holes in the new boom, and re-assemble everything. Didn't get lucky on the motor direction on the first stab this time :black_eyed:.

Took off nice a smooth within a few minutes after I was done and I spent about 10 minutes playing with the P & I gains. I went as high as P:180 / I:90 and had just a little quick oscillation in acro mode. Come down to P:160/I:80 and it was pretty smooth in acro mode. Putting it into self level then brings quick oscillations at a hover. Tried lowering the SL gain as low as 30 to see no difference, so I put it back at 80. Bringing the main gains down to P:120/I:60 did smooth out the oscillations some, but as soon as it was in a mild wind it back to moderate oscillations. At this point I figured there would only be a few minutes of flight time left on the 2200mAh battery, so I took it out for some moderate maneuvering. It flew well, other than the oscillation at level, and enjoyed flinging it around for a bit. Going to work on the gains some more soon. If anyone has recommendations for tuning, I am all ears.

On a side note, there was a rattle that I am fairly sure wasn't there before. It was either wiring getting blown about or perhaps the other motor that didn't sheer a prop hub has a damaged bearing. Next time I take it up I will see about isolating one of the options.