Project SK1 Begins

Today (9/24/16) marks the beginning of the next project. It’s starting a little prematurely due to breaking both arms on the B1. Due to that and the fact that I was going to do an overhaul of the electronics, I decided it was best to start on the next project. Although based on the B1, it will have these changes:

1. The use of a 4-in-1 Racerstar 30a V2 ESC
2. Further reducing the cross section
3. Capability to run 6s (the Brotherhobby T2’s are rated for 5s, but I have confidence in these motors).
4. Finally trying 4khz/4khz on the fc.

Update 9/27/16:

The frame is complete and the electronic configuration has been laid out:

Compared to the B1:

• The cross section in the XY plane (top) has been reduced by 955mm²
• The cross section in the XZ plane has been reduced by 400mm²
• Motor distance has been reduced from 250mm to 230mm
• Frame weight is 60g (15g less)
• The front motor arm has a Z offset of -10mm (might need more) to accomodate the shift in the CG of the 6s battery. This can be adjusted using spacers.

The only thing I am waiting on now is the speed control which is in transit.

Update 10/7/16:

The Racerstar RS30Ax4 V2 30A ESC finally arrived and is installed on the SK1. I still need to finish up the Vtx, the OSD, and the flight controller set up (SP Racing F3 Evo):

Update 10/9/16:

FC has been flashed to Betaflight 3.0 and everything appears to be working. Tonight I will finish the Betaflight setup and make sure the ESC is to the latest BLheli S version. I hope to get in at least 2 flying days this week. First will be flights using a 4s to get some blackbox data to see how well everything is aligned and balanced.

Weight as shown is 300g. However, in regards to speed only, weight has an EXTREMELY small effect.

 Update 10/16/16:

Initial flights went poorly due to my inexperience with PID tuning. The main issue is the motor jitter/stutter which only happens around 50% throttle and up. After playing with motor timing, notch filters, lpf, etc, I finally set my lpf to 15hz, cut the P term on yaw, and also cut my D terms in half. Since I am unable to fly around where I live, I am left to testing the results by holding the quad in my hand without props and quickly changing throttle, roll, etc. Before, the motors would stutter with quick throttle changes. Now they appear to be running much smoother.

This delay allowed me time to order a 5s battery now that they are in stock. This makes me feel a little better since now I can go from 4s to 5s and then to 6s. I was worried going from a 4s directly to 6s.

I hope to FINALLY have real results by Tuesday 10/18…

Update 10/20/16:

Delays… Stutter continued and ended up crashing on Tuesday. Apparently, the stutter wasn’t due to PID settings/filtering, but build errors (not surprised!). Upon rebuilding, I found a motor wire that was damaged that might have been grounding on the frame. I also hard wired the esc to the fc. Stutter is now completely gone but the weather has been horrible. Great flying weather if you’re a duck!

Next attempt will probably be on Tuesday, possibly Monday.

Update 11/02/16:

This project has been slowed way down due to flight controller problems (and not to mention a crash due to the OSD dying in mid flight).

Gyro issues

I thought the issue were the PID settings, but apparently the MPU6500 and MPU9250 have issues with high powered motors. Not exactly what from, but seems to be pointing to the gyro sensor being sensitive to the stronger magnets and/or high voltage spikes which some claim to have fixed with a 1000μF 35V capacitor. I purchased one, but have not installed it.

Bad bearings

After the crash (using a 5s lipo), I had noticed nearly all the bearings in the motors were bad. I am assuming this was due to the gyro noise which caused excessive heat in the motors.

Revisions

Opting to slow down the pace, I decided to make the following changes:

• Use a flight controller with an MPU6050
• Switch to a minim OSD
• Use higher quality voltage regulators capable of 36V input
• Replace all motor bearings with VXB bearings
• Possibly try a different set of motors with a lower kv value to lower the amp draw/voltage drop

Update 11/12/16:

SK2

With the wind, cold, short daylight, busy school schedule (my kids), and no easy access to a large field, time is even more limited to do some test flights. However, this leaves me with more time for building and thinking up of other concepts. I ended up tearing up the SK1 and rebuilding it in an effort to strengthen the arms.

After some VERY rough crashes, I was impressed how tough these frames have been except for the attachment point on the arms (and the motor mounts which I will get to later). Instead of drilling through and puting a screw through the arm, they are now held on with clamps and reinforced by connecting the front and back arms with a carbon fiber plate. Although a slot has to be put into the arms for the wires, this will not be an issue since any significant bending forces from a crash will not  reach this area due to the mounts.

This design also creates a compartment that the battery can fit snugly into which makes it easy to integrate the battery into the overall shape of the quad and reduces turbulent air flow. Electrical tape is used to secure in the battery instead of velcro.

The Frame

The pictures above basically show enough information on how the frame was built. Not including the fasteners, there are 5 pieces to the frame:

• Main plate – a 38mm x 110mm x 1/16 inch (1.6mm) carbon fiber plate. Next time, I will probably use a 2mm thick plate like this one here
• 2 side plates – 18mm x 136mm x1/16 (1.6mm) carbon fiber plates
• 2 arms – 10mm outer diameter x 8mm inner diameter x 176mm carbon fiber tube which can be found here

The Body

The bodies on my quads are not exactly pretty, but they do the job. Since I typically change designs as I build and don’t know what the final design of the quad will look like. I have found that a functional body can be very quickly made by using a combination of electrical tape and .010″ (0.25mm) thick delrin (acetal resin). Both are extremely cheap, and if I am not happy with the results, I can start over without having invested tons of time and money. To help with making sharp bends in the delrin, a lighter is used. One bad thing about delrin though: it only likes to bend in one direction, otherwise it cracks! This still hasn’t been a big issue.

• Delrin for the body can be found here

Pylons Finally Explained (click here to see the updated pylon design)

Here is a closer look at the pylons (I haven’t finished sanding down the rough edges, etc. yet). Below are pictures of the front ones which have added spacers to help lower the center of thrust closer to the center of gravity.

The pylons have 3 main functions:

1. Increases the distance between the motors. Because of this, the central part of the frame can remain relatively short.
2. Provide a way to put the motors at any angle.
3. Increases the radius of rotation of the motors. As the motors are rotated on a more aggressive angle, the increased rotation radius prevents the propellers from “overlapping” (as viewed in the direction of movement). This in turn prevents the back propellers from getting dirty (or turbulent) air.

The holes/taps on the clamps are for 2.5mm screws, but I ended up drilling and tapping them for 3mm screws (only the clamps used on the motor mounts).

• The 10mm clamps can be found here (5mm wide)
• The 10mm clamps used to attach the arms to the main frame can be found here (6mm wide)
• The motor mounts can be found here

Update 11/29/16:

It Finally Lost the Jitters!

I’ll try to make my rant short, but this was the absolute worst string of electrical bad luck I have ever had. Apparently when my 4 in 1 ESC was damaged, my 12 volt regulator was also damaged, but it was not at all apparent. These were the issues caused by the bad regulator:

• When doing a fast roll with air mode on and throttle at 0, motors would stutter and the FPV video would cut out for a few seconds (this only happened when the camera was hooked up).
• I could not get an external Black Box to work.
• The radio receiver would cut in and out.

After going through 3 flight controllers, 2 receivers, 7 straight jackets, 2 video transmitters, and even replacing my motors, I thought everything was good until I hooked the OSD up and noticed the vtx cutting out again. I VERY hastily replaced the regulator last night so I could fly today after work.

After all the craziness, I ended up with a Seriously Dodo FC and Spedix ES 30 HV ESC’s.

Maiden Flight

Fairly uneventful since the sun sets quickly this time of year, but I was ecstatic that it finally worked. Things went great, but it is still an unfinished product. These few things still need to be done:

• Put the T2’s back on.
• Flash the FC to Betaflight
• Align the motors
• Replace the aluminum motor mounts with carbon fiber ones
• Rotate the motor angle to 45º (only for speed runs)
• Add an external Black Box

UPDATE ON THE BLACK BOX: I noticed this past weekend that motor 1 was a 2600kv(!?!?). Apparently it snuck its way into the bakers dozen of red bottoms that I have… now the black box makes PERFECT sense!

I’m out of time to get into flight details, but it felt extremely smooth and responsive (compared to what I am used to). At least I got the video up:

Pylon design update:

I got one step closer to making a stronger pylon design. I ended up getting rid of the aluminum motor mount and just cutting out a 5mm x 33mm x 1.6mm thick piece of carbon fiber and drilling screw holes through them:

The End?

Like the F1 project, I think this project is finished other than a few minor design changes. I am going to keep the red bottoms on (they are my favorite all around motor) and keep the T2’s for the next project in which the goal will be 110% speed. With all the technical bad luck I had, there was plenty of time to think about the next project.

Although I never attempted a speed record with it (but who knows?), I am very happy with the way it feels in flight; especially the way I can throw the momentum, invert, and stay there for what feels like forever.

Connector Update:

I did a quick update on the connector so the SK2 can use the same modified batteries as the C1/C2 projects. Also, it cleans up the design by not having the connector/battery cable hanging out the back.