Depending on the quality of filters in the GPS chip, you would probably want to increase the distance to the video antenna somehow. To me, it looks almost too close (but only real tests can tell for sure).
1 Like
I will fold it down for first tests, i dont need it to go far, i might extend it down under the quad if it becomes an issue.
I went to do a test flight of the ultralight v2 and it started cutting out, a check with the thermal camera shows one motor getting hot so its probably damaged, i have ordered a replacement.
I’m changing the v1 back to an analogue 5.8ghz video system and over to a 2.4ghz mLRS receiver, as the antenna is much smaller and it has a much faster telemetry rate.
Fyi, regarding batteries, while the vapecell are the highest capacity cells available, due to high internal resistance, they are more suitable for very low c rating.
From what I understand the drone consumes around 5A or 6A.
So from this graph of battery testing, the EVE 58E seems to be the best option for this application.
https://www.patreon.com/posts/updated-21700-130567536
2 Likes
I have bookmarked those cells, I will order them once I have done some testing with the 6250mah cells.
1 Like
i was looking at this for an autonomous security drone, i wanted to know hows the wind resistance on these builds, can they handle a small to medium gust of wind?
I’m not sure, it probably will be ok in light wind, but it will shorten its flight time.
Thanks to all for the instructions and tips. This is a rebuild after I crashed the first one. It is set up for a FPV demo now. 221g. I went with Samsung 35E batteries for low internal resistance and availability.
Some things I learned building and flying it. Keep the compass away from the power wires. I need to redesign the top at some point to move it away, right now it is directly above the ESC input and capacitor. I constantly have to recalibrate it.
I had originally intended to use some spare XBee modules for telemetry but those didn’t work out so I put ESP8266 boards on the quad and controller. It is not good. Range is terrible and connections are flaky. I’m working on building a mLRS radio for the controller.
I had to disable GPS auto-detection and set the type and baud directly. Set to auto it would take 10 minutes or more to detect the GPS.
J-B Weld clear would not stick to the carbon fiber sheets even with sanding. PlasticWeld has worked much better.
The Cicada 6A was not available. I used the Hakrc 15A and that seems to work fine.
4 Likes
Yea, I use those and use them like a very long USB cable…
1 Like
you will struggle with range on a esp8266 as its running on the same frequency as your 2.4ghz RC system.
The motors are from something similar to a dji mavic mini, it’s rated to 8 m/s (17.8 MPH) wind, so this should be safe up to those limits.
Making some progress with my build.
One key design point, is that the battery sits direct on one of the tubes (45deg to line of flight). The FC (GOKU F405 1-2S 12A) is strapped to the battery with the goal of minimising the vibration that the FC experiences, because the battery mass is 60% of the drone and will damp out the high frequency vibrations.
Frame tubes are 8mmX0.5 wall. Bonded together with epoxy coated carbon “thread” and braced with 1mm rods. Arms are long enough to upgrade to 8" props later.
Camera and GPS mounting not decided yet.
Motor mounts are tied in place with kevlar thread and a spot of hot glue (easy to remove and change later for different motors). wires run inside tubes.
1 Like
thats very cool. how much does it weigh?
Thanks. It is 258g with the gps (not too bad considering Walksnail video, the long arms and larger tube dia). I can see some potential for reductions like mounting FC and video as a stack, but I’m happy with it as my first ever build. I wanted it a bit stronger to cope with my likely crashes as a beginner (like the tube extending beyond the motors to take an impact to save the motor from a direct hit). A “V2” version should be possible under 250g.
1 Like
This. Is. Amazing! You identified the problems and solved it, the greatest ones being flight time, autonomous and price. This is engineering.
But I have a question. Would you possibly get more flight time if you use a 1s2p battery instead of 2s1p and power the FC which requires 2s through a step-up converter? After increasing the kv of the motors, of course.
Another question:. I would have thought 2750kv was too low for a nearly 5” prop. Would this actually help with efficiency or did you choose it because of cost?
IIRC the idea with using DJI propeller and motor assemblies is that they are highly optimized for the intended operating point which is 2S and 250g takeoff weight, ideally we would also use DJI ESCs but that isn’t practical.
If you change voltage too much you need to redo the entire aero-electro-mechanical optimization from scratch. Lower supply voltage typically means higher losses as P=I²R
4 Likes
Yes, redoing the propulsion is what I intended. As for losses, I think the extra capacity would have greater effect than the losses. Can’t really tell unless calculated
2p1s is still two cells, energy doesn’t particularly care about voltage you deliver it at.
1 Like
1S2P (One Series, Two Parallel)
Cells: 2 cells.
Voltage: Same as one cell (e.g., ~3.7V for LiPo).
Capacity (mAh): Doubles the single cell capacity (e.g., 400mAh becomes 800mAh).
2S1P (Two Series, One Parallel)
Cells: 2 cells.
Voltage: Doubles the single cell voltage (e.g., ~7.4V for LiPo).
Capacity (mAh): Same as one cell (e.g., 400mAh remains 400mAh).
P =VI, V increases, I decreases
My experience suggests that if ESC and motor can handle higher voltage, it is best to use higher voltage. The wires can be thinner, which helps reduce some weight. Typically, the ampere is lower, and the spin is at a lower speed for the same thrust, resulting in reduced power consumption.
1 Like
Hover RPM is fixed for the given propeller and weight combination.
2 Likes