The 4S LiIon VT6 packs that I have for this are 208g. I think you need something substantially smaller all round to get under 250g. Under 250g was not the goal 
That was not your goal, I know, but it is mine
You do not need such a heavy battery (each LIIon cell weighs 50gr, so 200gr for 4S is quite heavy). I intend to test a 3S and a 4S at max 100gr.
Here is my custom 5" frame (can fit even 6" props) : weighs only 44grams and is very strong (it is of course not meant to be a basher quad)
2 Likes
http://queenbattery.com.cn/index.php?controller=attachment&id_attachment=38
So each VT6 cell is 46g. 3000mAh which is what you want in a LR build.
To 46gr you have to add the wiring, solder, packaging, balancing wires, so you get closer to 50gr/cell.
Li Ion is what is required indeed for max autonomy.
If the objective is max autonomy, you do not fly a 5" which is overkill in size and weight. You can fly a 4k camera or even a Gopro on a 3" or a 3.5" on a 2S LiIon.
In fact there is an optimal design for each specific objectives you require from your build : do you need a camera ? do you need a gopro ? do you need autonomous missions ? (which require a GPS, a magnetometer which adds weight, etc)
Hi @Hugues ,
we were talking here about a sub 250, but it seems that everything is at a standstill.
Meanwhile, with a friend we are testing a 3" build made with the material indicated in my list (including lidar and optical flow) and at 2s Liion we are at 249 grams. However, I think that the choice of 3 "is the best way. I remember your 28 minutes 2s Liion 
Nice thread on the sub 250g project.
We (as a community around Ardupilot) have not yet a viable 3" solution that is sub-250, can use all of Ardupilot’s features (mostly autonomous waypoints), carry a digital fpv camera and possibly also an HD camera such as a 360go, SMO 4k or Gopro without jello/vibrations.
Ardupilot is on the 3" size copters not as good or not as easy as betaflight. I do not know why. it should not be the case. I assume it has to do with a design and internal algorithms differences between Arducopter and betaflight ?
1 Like
My sub 250 (219 gr exactly) has been flying for two years and has all the basic functions of arducopter: RTL / smartRTL, waypoint (also in the field), loiter, GPS poshold, FPV, pos hold optical flow, 15 min flytime ( without modification esc for the moment), Yappu telemetry on the radio (thanks Andy).
On the next there will be DJI digital FPV (which can record quality DVRs), esc modified to work at 2.5V, front lidar for obstacle avoidance. With the plugin for the initial parameters, notch filters, autotune for the PIDs, I don’t see great difficulties more than betaflight, on the contrary it seems more flexible to adapt to any configuration, unlike the Inav presets.
Good to hear that it works. Could you post your build experience, results and tuning ?
1 Like
Some information can be found in the thread sub 250 already posted, also in my youtube channel you will find some tests. However, going into the details takes time, I had intended to do it but the ardupilot funding committee at the moment does not find it useful, I don’t know. However even my 5 "has been flying for almost two years (2 inav + 2 arducopter you can see it in the sub250 thread) and it was relatively easy with the sequence: initial parameter plugin, notch, autotune.
The issue is the frame. The intent of this series is to reach as many potential users as possible with standard well made components that give them a good experience. We are not targeting power users who will be more comfortable with esoteric parts, builds, complications etc. I think this build is at the limits of what complexity is worth it - obviously you guys will be able to go further but you are not the target market here
TBS have some new superlight frames, it might be worth exploring in that context.
Andy,
i think the frame is not really a problem. In the thread of the sub 250 I have proposed three, at low cost, and two of these already fly with arducopter.
It can also be viewed here and there are also wonderful open source files (with DXF files). Look here for example … 3 “to 5” by changing arms
In the spirit of reaching as many people as possible with this build. It might make sense to do the exact same thing but with digital. Just switch out the sixty9 vtx/rx and cam. I know that difference sounds trivial but it could be an addon video. Therefore building off all of the videos you’ve already done so you can reach many more people for a small amount of cost and energy.
I think this concept in general though is a great idea. I know that my first ardupilot project felt much more difficult compared to betaflight. In reality it ended up not being any more complicated but that appearance of difficulty can cause more apprehension than actual difficulty every could.
Are the new lightweight TBS frames you mentioned already out or something coming out soon?
I’ll probably do this for the 7" as I already have that on digital
What is the weight of a 5" MOD-L frame ? On their web site they do not publish weight data
In the meantime I ordered on cncdrones the ulatr light AOS 5" frame at 55 grams (all hardware included) which can carry two 20x20 stacks (compatible to carry a caddx vista). I will try a light build on that frame.
Do you know a link to these frames ? I checked 5" TBS frames but they are super heavy at 145 grams
Hyper Light, SniperX, they are on the website.
The hyperlite is 108grams, way too heavy (versus the available 55g which is so far the reference of the lightest 5" frame that can carry two 20x20 stacks compatible thus for a caddxvista)
The sniper X is 61grams and can only carry one 20x20 stack (thus not be able to carry a caddx vista)
So it seems the AOS UL5 at 55grams, hardware included is the lightest choice available on the market for a 5" frame than can carry a HD solution
The weight is not declared, but in the comments I read this which could be a good compromise:
mod-L SB 4" frame should be 30g.
Thx. And what about their 5" frame ?