330-gram Mini Quad Upgrades

3689567334?profile=original

I recently rebuilt my mini quad with a new camera, modified ESCs, custom wiring harness and a few other minor tweaks designed to make it neater and less messy. Initially I had just knocked this frame together a selection parts to make it fly.  Even on the stock KK settings (latest update) it flew reasonably well, however there were a few problems and obvious areas where improvements needed to be made. As can be seen below.

3689567406?profile=original

Here is an account of those improvements, along with a list of relevant parts, and some images and flight video.

FASTENERS

The AQ50D frame was supplied with all metal fasteners and spacers which seemed to be overkill in terms of strength, and an obvious area from which to shed some weight.  18 M3x12mm aluminium frame spacers were replaced with 10mm nylon, 2 M3x15mm aluminium spacers were replaced with 15mm nylon, and the majority of the 32 Stainless Steel fasteners were replaced with nylon.  This exercise, while relatively expensive, reduced the weight of the structure by a significant margin.

RC RECEIVER

The V8R4-II is a tiny little receiver with the very smallest of antennas (it is barely longer than an inch) which obviously limits distance between the element and whatever the receiver is mounted to.  I have had varying success with this receiver depending on the type of airframe I place it on.  For example, it is positioned directly beside the battery (just 12mm away from the battery itself) on my balsawood octocopter and yet the integrity of the link is faultlessly reliable, whereas I have experienced dropouts on this little Carbon Fibre frame (with the flight battery and BEC combination consistent on both airframes) especially then flying in build up (noisy) areas.

Considering the very tiny weight of the receiver module, I decided to address the issue by constructing a custom lead to run between the V8R4-II and the KK board which would also serve as a support to raise it well clear of potential sources of interference and to allow for improved signal propagation.  While it works perfectly, I did perhaps make this link longer than it should be.  I actually have another V8R4-II (which my cat had the audacity to use as a teething aid, but which still functions fine) - so I might stripped it down, remove the pin headers and dump the cardboard case in favour of a heat shrink sleeve in order to create a leaner version.

3689567459?profile=original

KK2.0 BOARD

I had purchased a specific lightweight GPS receiver in preparation to upgrade this frame with an APM however, concerned that it might be a bit too much to try and fit on this tiny little frame, I decided against it.

The top of the KK board was already a cluttered mess, with the previous setup and I wanted to add to the complexity by utilising the onboard voltage monitoring circuit and associated buzzer and also use the KK Board for 5V power distribution.  To solve the problem I used tiny pin connectors, soldering pin sockets onto the bottom to allow stealth connection of battery voltage, buzzer and the 4 signal wires for the ESCs.

3689567474?profile=original

ESCs & BEC

When setting the quad up originally, I chose to isolate the BECs from on another.  I did so by eliminating each BECs ability to bridge at the KK board (although I am aware that this is unnecessary).  I then powered the KK board from the output of ESC#1, my 200mW 5.8 VTX from ESC#2 and the 2.4Ghz FrSky receiver from ESC#3 with ESC#4 pin removed and insulated.

The BECs in the little 6A Turnigy ESCs I am using are tiny SOT-223 linear regulators rated at 800mA, however I found that they struggled to deliver anything like that reliably (regardless of whether powered from a 2S or 3S battery).  Without going into too much detail, I had intermittent problems with video link (horrible blizzard), RC link (drop outs) and with the KK board (mostly the screen would drop out or the board would simply not wake up) - all of which caused by voltage sage, with the IC getting extremely hot.

The solution was to add a separate 3A Hobbywing BEC to the mix, which solved all my problems.

With an alternative 5V solution sorted, the 4 linear regulators were essentially worse than dead weight, as they seemed to also be generating much of the waste heat emanating from the ESCs.  I decided to remove them entirely.  While doing this, I also opted to remove the 100uF caps in order to further simplify the form factor of the ESCs, as well as add a set of tiny female pin sockets for faster more convenient installation and servicing (previously, I just soldered everything in - making it hard to repair/swap out motors).  A pair of 470uF 25V electros located at the power distribution point replace the 4 x 100uF caps, simplifying installation and more than doubling capacitance.

3689567431?profile=original

BATTERY

I tested a few different batteries with this frame, ranging from a 40-gram 2S 900mAh to a 192-gram 3S 2200mAh.  The small 40g 2S gave great sporty performance, but the COG was better with the 3S 2200mAh (even if it was a tad heavy).  I decided to create a T-shaped unit with two 2S 900mAh packs, which provides double the capacity of the single 2S 900mAh, allows a better COG, but which weighs less than half the 3S 2200mAh.

PROPS

I started with 4025 props which worked great and were very quiet and smooth, but once equipped with extra FPV gear it was a bit under powered.  It is now wearing 3-blade 5030s, which are just as nice (even without balancing) and offer plenty more lift.  I would have likely to have found a good quality 2-blade 5030 for better efficiency over the 3-bladers, but I failed to do so.

WIRING

The entire wiring harness was remade from 22AWG Superflex Silicon wire in order to neaten things up and minimise chassis wiring.

CAMERA

I have been using a #26 808 HD keychain camera which I modified with an add-on lens designed for the iphone (providing very wide field of view) for video recording and FPV duties for a while.  This worked well and gives great quality video, however my lens mod is not very durable, and the recording video format of the #26 808 is difficult to work with.

Recently I purchased a Mobius HD cam which generates mov files and can be imported into iMovie for easy editing.

The Mobius Camera is simply attached to the quad by a velcro strap, and is held down against the 3A BEC.  A very tiny about of vibration reduction is (potentially) provided by a thin 1mm piece of double sided tape used to attached the BEC to the frame.

3689567489?profile=original

BUILD COMPONENTS:

FRAME

1 x Alien Spider AQ50D PRO 250mm Mini Quadcopter Carbon Fiber - $64.27USD

8 x ACP-AC-0024 M3 Rubber Washer - $1.36AUD

10x Nylon Standoff Spacer M3x6mm Male x M3 Female  - $2.80AUD

10x Nylon Standoff Spacer 20mm M3 Female x M3 Female - $3.58AUD

25 x Black Nylon M3 x 6mm Philips Head Screw - $3.81AUD

10x Nylon Standoff Spacer 10mm M3 Female x M3 Female - $6.05AUD

4 x M3x10mm Nylon Screws - $0.80AUD

18 x M2.5 x 16mm Stainless Button Head Socket Screw G304 Allen - $3.20AUD

DRIVELINE

2 x 470uF 16V 105C Radial Electrolytic Capacitor 8x11mm  - $0.40AUD

1 x Micro size wire Shrinkage Wrap Φ1.0mm>Φ0.6mm (150mm B/R/W 3pcs)  HC-039 - $0.99USD

22AWG High Quality Silicon Wire (60 Core, O/R/B, 1m)  HC-JR22AWG - $4.99USD

4 x Turnigy Plush 6A/.8bec/6g Speed Controller (HK HK Warehouse) - $30.60USD

4 x Turnigy 1811 brushless Outrunner 2900kv (HK AUS Warehouse) - $43.84USD

5030 (black) Three Blade Propeller (ABS) (3pc) (HK AUS Warehouse) - $2.50USD

5030R (black) Three Blade Propeller (ABS) (3pc) (HK AUS Warehouse) - $2.43USD

ELECTRONICS                                          

New Hobbywing 3A UBEC 5V 6V Max 5A Lowest RF Noise RC BEC - $4.89AUD

Hobbyking KK2.0 Multi-rotor LCD Flight Control Board (HK AUS Warehouse) - $29.99AUD

FrSky V8R4-II 2.4Ghz 4CH Receiver (HK HK Warehouse) - $13.62AUD

2 x E-Flite Blade SR Hook and Loop Battery Strap B400 - $1.50AUD

2 x 7.4V 20C 900mAh LiPo 2S 7.4 Volt RC Akku Battery - $15.98AUD

3 x Dean Connectors (Pair) - $1.50AUD

CAMERA                                          

Mobius ActionCam Full HD Sports Camera 1080P 30FPS 720P 60FPS - $77.26AUD

SanDisk 8GB 8G Class 10 MOBILE ULTRA Micro SD Micro SDHC - $12.70AUD 

TOTAL COST - $351.61AUD ($314.09USD) + some $ for shipping

TOTAL WEIGHT (Including both 900mAh 2S lipos and Mobius Camera) 328-grams

Uploaded test footage can be viewed here (looks terrible), Stabilised version can be seen here (also looks terrible). Or the original mov file can be downloaded here (looks great).

FURTHER PLANNED MODIFICATIONS

1. Substitute the unnecessarily large and heavy dean connectors for something more suitably sized for the task.

2. Relocate BEC from top of frame to allow integration of Moon Gel for jellow free video.

3. Upgrade to new KK2.1 board.

4. Test 6030 props.

5. Reinstate 200mW 5.8Ghz VTX to allow FPV flight (at back on top of battery, maximising distance from RC receiver)

6. Find better way of attaching motors to frame, ideally one which integrates some form of mechanical noise isolation. I am thinking of sandwiching/compressing a piece of bicycle inner tube between the motor mount and frame.

7. Gain a better understanding of the tuning process, so that I can eliminate the terrible toilet bowling effect which occurs whenever the craft is descending.  I can eliminate it on my other airframes, but not this one.

8. Commence constructing <50g micro Brushless Gimbal and 2-Axis Micro BLGC.  Once completed, tested, tuned and ready for installation, I will rotate the AQ50D frame 180-degrees (effectively running it back to front) in order to accommodate a front mounted gimbal mounting location.

I am really very happy with the baby quad.  It has been a joy to build and I am also really inspired by the performance of the amazing little Mobius Camera - having simply strapping it to the airframe with no effort towards improvement (ie - no prop balancing done, no vibration reduction solution, etc).  Pausing on any frame in the video shows a picture perfect still, and I cant wait to test out the high res still image capture function.

I would invite any feedback on further improvements that could be made.  Also, the front motors are 6mm closer together than the rear.  I assume this is an issue that needs to be addressed and can be tuned out (so to avoid changing the existing dimensions), but I am currently unaware of how to do this.

E-mail me when people leave their comments –

You need to be a member of diydrones to add comments!

Join diydrones

Comments

  • Arthur - There is a bit going there but I think it all fits nicely, and the chassis wiring is as minimal as can be. Keeping everything neat, compact and isolating all the noisy stuff from all the sensitive stuff is my goal.

    Crash - Yeah, I love the simplicity, ease of use and predictability of the KK. I obviously have the receiver failsafe set to drop the craft out of the sky in the event the RC link is broken. Being less than ideal, I am careful of what I fly above. I am still very much a greenhorn limited to flying line of site, and it's tiny weight makes it less of an issue.
  • Nice litte flyer, I guess it will outperform your GTI on the first meters ......

    A basic flightcontrol is always a warranty for fun and no sudden fishy behaviour.

  • Developer

    Quadzimodo: There is a log going on there. But at the principle is a L-C filter. (with L being your wiring)

    Those capacitors are there to provide the high-demand current needed by the switching mosfets on the ESC, for a good design there will also be caps. of smaller values on the PCB (the big electrolytic caps. also have a high inductance). Moving those too far away from where they are needed makes them lose their porpoise, or not. I wrote that "or not" because I'm not sure of the involved values here, so the effect will not be noticeable.

    Here is one of the best books on the subject: http://www.amazon.com/Electromagnetic-Compatibility-Engineering-Hen...

  • Ruwan - 12-15 minutes when discharging to 6.6V

    Arthur - Thank you so much for the tip. The idea that inductance could be a factor hadn't even crossed my mind.  It is an issue that I wish to understand though, as I have another build underway which uses the same modified ESCs and relocated filtering Capacitors under development that I really want to get right. An image of the layout of this other project is pictured below for your reference.  I expect that inductance will be much less of an issue with this design due to the shorter path lengths involved (the base plate is just 110mm wide). but I will investigate how to calculate this so to be sure of the potential impact.

    3692918232?profile=original

  • Developer

    The quad looks nice. But here is one quirk:

    A pair of 470uF 25V electros located at the power distribution point replace the 4 x 100uF caps, simplifying installation and more than doubling capacitance.

    That may not be the best thing for your system, the position of those caps can matter (I don't want to cal anything now, but you can search for EMI and wire inductance power supply).

  • Moderator

    Nice work! What is the flight time you get (w/ camera)?

This reply was deleted.