The challenge
https://intofpv.com/t-challenge-sub-250-...%80%93-10k

My entry into the 30minute / 10km Sub250g Copter Challenge. I'm planning a few extra self-imposed extras as several people have finished the challenge and I wanted to add some stretch goals..

Since there are at least 3 viable digital systems avialable on the market I wanted to add at minimum a digital FPV system to the build. 

My Caddx/Vista Nano setup is 26g with nano camera and polar antenna.

I've built a few OpenHD VTX systems well under 20g, and when getting closer to the weight of the DJI system with an ASUS AC56 stripped adapter ranges of over 20km+ have been flow. So the limitation, unlike the DJI system of 13.1Km will be batteries. Hopefully the Frostbyte comes out with some high power and long range VTX options to make it into the tests.

For flight controller software I feel that any quad going on long range missions needs to have a reliable way to get back home, and for extra fun navigate autonomously like a true drone. Multiple software packages exist, iNav and Ardupilot I've used with success.  

My stretch goals:

- Digital FPV System
- Navigation/True RTH
- Fully autonomous flight to 10km and back (manual takeoff permitted as iNav must be launched)
- Greater than 45minutes of flight / Greater than 60min of flight
- Greater than 15km distance / or max range of DJI 13.1km
- One way flight with alternate landing zone 20km+ away from takeoff point

If anyone has any ideas to push the 250g limits please comment.

My current test platform hardware:

Motors:
- 1404 / 2500kv
- 1404 / 2750kv 
- 1404 / 3000kv
- 1404 / 3650kv

Props:
- Gemfam 4" Bi-blade
- HQ 4" Tri-Blade
- Gemfan 5" Bi-Blade
- HQ 5" Tri-Blade
- HQ 6" Tri-Blade

* I'll post the prop pitches once I go through my boxes. most between 2p and 2.5p, I wish I could get 4" or 5" with 4p 5p

Flight Controller:
- Flywoo F745 AIO modified for low voltage

Video System:
- DJI Caddx Vista (Nano Cam, Polar Cam) Stock Antenna, Polar Antenna, TrueRC Singularity
- OpenHD, Stripped Rpi Zero, Asus AC56, IMX290 Camera
- I don't have a Frostbyte system to test with yet

RX:
- Crossfire Nano RX / Immortal T


Battery:
- 2S 18650 VTC6 Pack 99g (plan to use this pack until a battery pack capable of more W/g comes out)

Extra:
- To keep the DJI VTX with reliable voltage a Pololu 9V boost regular was added. total weight of about 1.5g.


Initial Flight Tests

Everything worked well, tuning with the 2500kv motors and 5" blade first, P to 80, D to 55.

First issue that I thought I solved with using a BLHeli_32 ESC board was getting what felt like LVC and motor power being reduced on my original BLHeli boards. However with the new board, as soon as I hit 3.3V/cell same motor power drop and then would not arm after a few minutes of this.


** Solution to the LVC drop-out of the 2S-6S FC ** copied from my post in the challenge thread:

Looks like bypassing the regulators on the FC is not optional, at least on the 6S capable boards. About the time I submitted my question above, I was staring at the Flywoo datasheet and realized it's a 6S capable board. Sure enough after some probing with a meter I found:

VBAT -> Protection Diode -> 12V Regulator -> 12V buss
VBAT -> 9V Regulator -> DJI 9V Bus

12V bus -> 5V Regulator
12V bus -> 3.3V Regulator

Bypassing the protection diode and 12V regulator I'm able to take the FC down to 4.45v and still operate the ESC's.

So for anyone using a 2S-6S Lipo FC I would guess this is required. I'm surprised that the board was capable of running under 5V!

Plugged the battery in and was able to jump back in the air with the battery that previously cut-out at 3.3V.

Now the bigger issue is lower than 2.8V/cell on 2500kv, on 5x2.5 Bi-blades it's close to 100% power lol 


Some pictures of the copter:








Frame Updates Sept 19th

I'm using the Airblade Transformer 5" LR frame. I find this to be a horrible long range frame. Built as is with battery strap it is over 50g+, it's really heavy. The bolts to hold the bottom arms are single 3mm bolts with heavy press nuts. There are no double bolts to hold the arms, they are snug/angle fit in the middle to not move around, this is already loose new and I expect it to only get worse. There is a "locking" upside down middle bolt, with also a press nut. Being 3mm bolts and large press nuts this forces you to offset the FC a significant distance to not short against the nuts. There are no TPU or mounts for the GPS on what is suppose to be a LR quad, no good place for the Crossfire antenna, and the cage up front is not well designed. To see a very well made, positive locking camera cage like this check out the AOS 3.5 frame. So I ended up just removing the whole cage.

Hi Littleflyer,                                                                                          Last Post

I have added this thread to the list of Challenge PARTICIPANTS.

Now, you potentially have a lot of variations of this build.
So, please post as you go and provide pics and data at each point.

As you work through the various configurations, list the components for that particular version
{include in the post} and perform the flights to get the results and post. Record the flights 
including the final stats screen. If criteria are met, post the flight video as evidence. 

Regarding Frames:

It is a challenge to find light weight 5 inch frames...not impossible, but depends on the goals.
The BetaFPV X-Knight is the lightest 5 inch frame that I have found at 26.44 grams. Now, this frame
is not a heavy duty frame, but might be ok for a Challenge build. 

I have been working on a custom fabricated frame for my build, but still coming in at about 33-35 grams.
I will likely try this frame, but if it doesn't work out, then I will likely go with the X-Knight  or the PyroDrone
Hyperlite UL5 V1.1 which weighs 30.5 grams. This frame appears to be a little sturdier than the X-Knight.

>>>>>>>>>>

It will be interesting to follow your progress with the different variations and observe the results.

I am particularly interested in how the different motors perform. 

By the way, if you have not already done so, you might want to download my PARTS GUIDE  <-- click to download.

I did NOT see your parts guide and detailed notes! That is some great info...

I does look like I'll be going down the custom frame design rabbit hole on this one, there just isn't anything that goes up to 6" in the light weight design that I am looking for to do my testing.

Been working on many calibration flights and some tuning, prop-wash is bad, still a bunch of wobble, but we are 90% better than stock inav PID's.

Also, tiny power consumption preview, @ 249g using 18650VTC6 2S pack, pulling about 850mah out of the pack when hovering for 10 minute flight +- 25ma

Hi littleflyer,

After much work on a custom frame that I could actually fabricate with my meager means, I settled on the
design as shown in the picture below. It is pretty simple and easy to fabricate and put together. In my case,
I fabricated the arms from a fiberglass sheet, a small base plate from thin carbon fiber, and a "top plate" from 
lightweight clear plastic. This was the most workable solution for me. Now, I am using steel bolts and lock nuts. 
I will use nylon standoff columns in front to mount the camera and in back to mount the GPS, VTX antenna, and 
Rx antenna. Yeah, I will be using a 26x26 whoop format AIO FC board.

I could probably have made the arms a little narrower to reduce the weight a bit, but I wanted to start with arms
that might actually be strong enough to handle the load of the drive train. 

PCB material should be just fine! I build a lot of frames from that before I started to build from CF or purchased "real" frames eventually.... I scratch-built a TBS discovery with PCB material and it worked, and lots of small quad test frames. I even sometimes up to 5" frames 3D print them from PETG, I just up the thickness from what the CF will be, if its 4MM cf, I go 6MM printed and they hold up good for prototyping.

I'll update a power/flight-time table to the original post, but I finally calibrated my current measurement very accurately and went for a full flight, I only got to 22 minutes before the craft ran out of throttle.

So a couple things, I'm planning to always load as close as possible to 249g with each test of frame sizes, props, motors. A big goal now is to optimize for 249g, ie: best flight time and performance possible at maximum weight. As we already have proof if you make it 200g you can fly for almost an hour on 2S VTC6 cells.

What I think I've now hit and was not expecting is a motor KV for 1404 motors @ 249g of 2500kv is not enough to keep the aircraft flying sub 2.9v/cell.

My flight: 249g - 22minutes - 5x3 bi-blades t-mount - 2500kv 1404 - then not enough thrust to continue the flight... plenty of battery. consuming apx 88mah/minute, 900ma at 10 minute mark.

The next attempt I need my new larger frame, and that will be testing 6" props. Then I have some 2750kv motors and 3000kv motors. If you build waaay under 249g then these motors should be super efficient and fly for maybe over an hour with 5"/6"+ props? However at 249g on 4" or 5" you need more than 2500kv to extract all the go juice from these VTC6 cells.

I'm also running the Caddx Vista at 700mw output and I need to test how much power this is actually consuming during flight, does using 25mw change how much power it uses in close range? and as well going through the BLHeli_32 settings to optimize for LR duration efficiency.

But first to determine the minimum KV and maximum prop size/efficiency @ 249g , maybe its 2004 motors?

Hi littleflyer,

It is interesting that the 1404 2500Kv motors did not pull as expected. Presumably, that was a 2 cell 18650 pack.

I am still going to try the 1404.5 2722 Kv motors and see how they do; they are a bit bigger...we will see.

Is the PCB in sheets? Where do you buy it??

I got the FR4 sheets from American Composites:
https://www.americanepoxy.com/g10fr4sheets

I got the black in a few thickness's, its really awesome for making small parts from, frame parts, wing surface horns, all kinds of stuff, and the black almost looks like carbon.

Yes that's with the 2S/18650 VTC6 cells. It's not an LVC issue anymore, I fixed that and can go down to 4.5V. I just end up pushing in more and more throttle, and at 23min it just settles into ground effect, I could probly hover there another few minutes but it no longer has enough power to fly upwards out of GE.

Got some interesting results when I removed the DJI camera and a few other items and dropped the weight to 100g before battery pack. 2S-18650 I'm able to hover about 600ma/10min. Using 3S-18650 around 525ma. So for anyone that builds this with 2500kv motors and keeps it sub200g AUW you should be nearing 55minute flight times.

As an analog build its tight be doable to keep this under 100g before battery that would be also capable of 5km ranged flights.

Frame - 5" -> 30g (lighter is possible like the x-knight 5" @ 26g)
1404 Motors / 5" HQ Props / bolts -> 45g
Flywoo AIO FC, Gps w/Compass, Crossfire 15g
Caddx Ant Nano Camera, 400mw VTX, RHCP Antenna -> 10g

If the 2500kv motors can fully utilize all 3000ma of the pack this should yield a 40~45 minute flight time. It still might be a minimum of 2750Kv @ 2S

How heavy are your 1404.5 motors? have you tried them yet? Going 2S, 4" or 5" ? which props?

Hi littleflyer,

First, thanks for the link. Good information. I will likely get some since I like having raw material on hand.

Now, I don't do digital so my craft is analog using the Ant camera as well.
The motors weight 10.9 grams each so that puts the set at a little under 44 grams.

The build is not yet complete. At this point it weighs 100.5 grams without props, Rx receiver, camera, and antennas.
So, I figure it will end up with a dry weight of about 115-120 grams when all together. This configuration will be
my "baseline" and likely have the lowest kv motors with the largest props. Each configuration, will be testes with both
5 inch and 4 inch props and an assortment of batteries including 18650 packs and LiPo batteries. 

Also, there is a Low Voltage Cutoff modification for the FC that I am using, however, I have decided NOT to make
this mod up front. I will see how the testing goes. Although pretty good, there are places that I could reduce the
weight and the LVC mod is another place to gain a little more "battery" margin. 

My building and flying has sort of been put on hold due to Real LIFE priorities, but I hope to get some time to finish 
this one sometime soon. I am building a house for our son, so that has taken the top priority slot. Since I am doing 
almost all of the work by myself, it is taking up a good share of my time. Yeah, I work on it almost every day, most 
of the day, and I am pretty tired at the end of the day.

To complicate matters, I have also lost my two flying locations...so...now I am looking for a decent place to fly.

On the positive side, I have faith that everything will come together and work out fine...all in its own time.

Your interest in the Challenge has encouraged me to work in some time to get this build finished. 

Thank You.  

Sorry to hear about loosing your flying spots, had that happen many times, really sucks to have to find a new spot. great thing about the sub250g is it does open up some doors for flying for a lot of places. The house stuff also seems like a lot of work, hopefully an end in sight soon

----

I think I'm able to fairly confidently call some hard limits in terms of motor/prop selection, until someone proves otherwise with a real flight.

I've assembled my 6" platform that I can now test 4", 5" and 6" props, I may also want to use the frame later so I didn't opt to build it just to maximize prop-size weight. It is a LW-LR Design, not a Toothpick X. For all my comparisons I'll be keeping the sub250g 6" and below for now.


FLIGHT TESTING
Motors 1404/2500kv / 2S 18650 VTC6 / 249g weight all tests

- On 4"x3x2 is not enough thrust to fly on 2S except for the first few minutes of the battery
- On 5"x3x2 is enough thrust to fly for 20 minutes, but then there is not enough thrust to continue to fly, however the PID controller and torque is OK
- On 6"x2.5x3 at 16 minutes the voltage too low for this size of motor to maintain enough torque to compensate and fly even with loose or strong PID loops (very floaty and quiet though)

I did a 4" prop test on 2750kv/1404 for 10 minutes, the draw was around 900ma, and not much less efficiency over the 2500kv motors. 

Key takeaway:
From this I now know that 1404/2500kv is too low a KV to use the full capacity of the 2S pack. It is also too small to swing a larger than 5" prop on 2S.



2 directions to go: UP or DOWN -> maximize and be OK with 30~35min flight on 2S < 249g digital VTX quad or push to larger motors to gain flight time at the cost of a quad that can't handle as well.

Sacrificing a "LR" style frame for a toothpick X unibody, and stripping the Caddx Vista would makes it possible to add larger motors, props, and this is the direction that is needed to fly much longer than 30min @ 249g.  Going with possibly 1404.5, 1804 or 2004 motors and trying motors starting at 2500kv on 7" props, it may also be that 2750+ is required. Mathing out the weights it should be OK to get up to around 8" before you have to go back to analog due to the current weight of the Digital systems.


The other direction is to simply build out a practical 5" frame design on 1404 motors with the full caddx w/heatsinks on, and likely end up under 249g anyways. You could then strip the vista and run a toothpick X like the X-knight just to see how much duration you could get, or put it on a LR-lite frame for a more practical flying machine.



----

Looking at the amazing build that is the Ultralight Ghost 4", it made 34 minutes on a 2S or 40min on a 3S all while keeping under 200g and 249g respectively. If we compare just the 2S build then the weight is 200g with a flight time of 34min. What this seems to tell me is that if this machine was loaded down to 249g that it would never make 30 minutes and that 1404/3800kv w/4" props is less efficient than a DJI mavic. What I have not actually seen yet is anyone competing in this challenge at the mavic weight of 249g. So my goal will be to continue to try beat or meet the DJI mavic mini's performance directly.


-----

FR4 stuff....





Here are two old copter made from F4 material that I got from American Comp. It's the 3mm/1/8" black sheet. These are both 10" quads, old MutltiWii gear, but can still fly. The material does work well and before getting into CF frames I made a half dozen different designs with that stuff.

The new 4 to 6" testing frame. 35g.




1404/2500kv motors (that are too low a KV and to small for 6" props)
HQ 6" props
Flywoo AIO w/Baro
Matek Compass/5883 - w/GPS 
Caddx Vista w/nano Camera, Polar Antenna
2S 18650 VTC6 cells
Crossfire Nano RX
iNav 3.0.2

Flew good on 5" props, just too low a KV. So for now 6" will need > 1404 motors @ 2S

Big update today on the Sub250 exploration build....

- A couple of up front information about this project. Not only is my goal to meet the challenge of 30min+ flights and 10km+ range, but I want to do it with digital systems, as well as find the optimal duration build, which can mean being sub250g, and the optimal 249g build, which maximizes power, motors, flight performance, range, but still manages the minimum challenge goals.


So with that this build log will have not one but 2 simultaneous builds going. Where things are specific I will preface the notes with Performance or LR, so just like Tesla's, where the LR goes further, more efficiently, and the performance models sacrifice some range for ludicrous mode acceleration and speed.

Both designs use the same unibody frame design. This very carefully engineered frame takes design aspects from the Explorer/Rekon quadcopters, as well as AirbladeUAV's Transformer, and Ummagods and others. I believe the interlocking camera plates of the AOS 3.5 are lighter, offer better camera protection, at the cost of a little durability. Due to the time to update the CAD for the frame this will be a V2 change.

Frame Design - 5" Arms

Arms
- Using the shape of the Dave_C Rekon/Explorer arms, +0.5mm wider and stretched to 5".
- Motor protection "triangles" that push outside the diameter of the arms have been lightened.

Angles
- Arm frame angles and placement optimized for 2S li-ion 100g CoG balance. 

Body
- 20x20 front mount
- 20x20 rear mount on body
- 20x20 rear matching holes on upper plate (allows "inverted" mounting of the VTX like the Babyhawk HD)
- AIO 25/26 front mount (the optimal FC for weight savings)
- Enough room between the front FC and Caddx Vista in the rear, unlike the Airblade Transformer
- 3mm front cam stretch to accommodate the AIO FC's
- Room and TPU mounts for both Caddx Polar or Caddx Nano camera's


Weight
- 3mm unibody frame design.
- This is one of the main features of this frame, nearly 50% lighter than most 5" frames, and lighter than almost all 4" frames.
- Since the goal is LR and duration, this is not a basher frame. If you break an arm, you replace the frame. However as most crashes flying LR will be more of a 
recovery effort, the cost of a $30 frame shouldn't be of primary concern.
- Although the front pokey out bits on the bottom frame look thin, they are 3mm thick and plenty strong

Can be built as BASE BOTTOM plate only, bottom frame is only 19.75g! 










next part, Performance build overview and testing results tomorrow....

That is sick man!

Hi. Where is the continuation of the story? I expected to see a series here for several seasons)

(07-Dec-2021, 06:12 AM)Navigator102 Wrote: Hi. Where is the continuation of the story? I expected to see a series here for several seasons)

Hi Navigator,

Good Question. I was also wondering what happened to littleflyer. 

I was expecting a video of the flights either here or on the CHALLENGE thread, but nothing yet.