3 Inch "Scratch Build" Log


#21

Yikes! Sounds frustrating, @NotFastEnuf. You tried what little I would have known, e.g. the ImpulseRC tool, further up already. So I’m afraid I’m at a loss. :confused:


#22

Thanks man. Oddly enough this is coming down to a short that for whatever reason doesn’t show up on USB connection. But it’s there. The meter shows it. That’s what I get for trying to do just one more thing at 2 am without checking my work as I go. Maybe the problem identified itself before causing any damage (other than the arduino).


#23

So next steps are disconnect 5v power from the pdb - recheck for short. If it’s still there, it has to be on signal wires. This was a known good board.

I really hope it’s not there, these cheap racerstar signal wires have the insulation that burns back like a fuse whenever the iron touches the wire. As you all know in a micro build - we don’t leave a lot of extra wire to “try again” with. I’m dreading going back into the signal wires with the iron.


#24

Just wondering… anything you can paint on the ends of the insulation to retard the melting? Or, just rest assured at using liquid e-tape to cope?


#25

I’d like to know that too.

Update, linear 5v regulator is my short on the pdb. Time to look in spare parts and replace.


#26

Not sad to see it go, it was always too noisy anyway. Never could get clean video off of it.


#27

Isolated the linear very reg, short was still there. Figured it was the capacitor. Un soldered it. Short was still there. Re melted positive pad - short gone. Re melted the re melt and pulled the solder around … short came back. So somewhere very close to the positive pad a trace has lost its protective coating. This is what you get with (over) used parts.

Solution: I should replace the pdb … but I’m not going to. I re soldered the vreg, found another ceramic cap and soldered it back in. Then re attached power and ground from 5v to fc. I checked for shorts along the way and readings look good now. It should be fine - time to get a deans on there and test with a battery (and smoke stopper).

Update: no smoke. Back to the building.


#28

So @Chaotix is recommending latest betaflight.

Possible loop times are

2k/2k
2.67gyro/1.33 pid

If I go back to 2.91, I can get 4k/4k. That’s still really tempting - but I will follow advice in the end.


#29

All I’d add is check your cpu % hopefully it’ll stay below 50%. The newest Betaflight seems to run on most of my F1 boards fine but I do have everything turned off, acc etc. Earlier 3.0 targets tended to be hit or miss trying to run 2/2K.


#30

Yes and thank you. The above speeds are fine. Very low cpu percentage - like 14%. Anything faster and it slams into 100%. Lol - no middle ground. How do you feel about over sampling gyro at the expense of a slower pid loop vs 2k/2k?


#31

Well here is my next snag. I can’t seem to get these blheli_s racerstar escs ti take a throttle calibration. One of them will go but the other 3 just keep locking into a range of like 1487 to 1608 or something ridiculous like that. They calibrate fine on oneshot125 but just won’t go on multishot. Argh. Knew I went out on a limb with the liteS version. Freaking oddball just to save a buck.

Uodate:
Multishot wouldn’t work calibrate on motors 1,2,3 for betaflight 3.1.6 but one shot was fine.
Downgraded to 2.91. Calibrated and worked fine
Upgraded to 3.1.6 - remained working fine.


#32

So this one is wrapped up … just have to find time to maiden and tune it. Build summary as follows:
Cc3d running betaflight 3.0 - nothing newer will run a pwm receiver and multishot properly
Dal 3045 tri blades spun by 1806’s on a 4s 860
It’s a fat pig at just north of 300g with battery.
More could be done to trim by updating parts … but at least all this stuff isn’t in a drawer or still on an under performing 5". Switching to a 650 pack, lighter vtx & antenna, aio fc would all get it under 250grams and it would still probably fly for 4 minutes. Total amp draw for 4 motors together is only like 28 amps. These 1806 motors are really efficient at the low kv but have the torque to get close to theoretical rpm which is probably 90 % of what the high kv 1306 give at way more amps. I’ll come back to the and post a flight video when I can record one.


#33

@NotFastEnuf: Good to see this completed and the 1806 motors successfully re-used. Obviously heavy for a 3" quad. But I’m sure it will fly just fine. Good luck with the maiden flight!

You note the efficiency of the 1806 motors. A shorter version, 1804, is very efficient too, a good bit lighter, and does well with 4" and 5" props. 1804s have been around for years, but somehow aren’t popular anymore.


#34

I love the application of your camera mount. It looks solid… definitely looks like a factory precision made quad. AMAZING job on the flames, looks outstanding!


#35

Thanks man. To be honest … it’s really “good from a far but far from good” hahaha… lots of little imperfections you can see up close. But I knew it was a chubby build going in so I didn’t give it the love it deserved in the process … just kind of banged it out. I have another 30 gram 3" frame that I indend to build with better components that isn’t starting off behind the 8 ball heavy (frame featured here was close to 60 grams). This one is more of a practice quad for that one ultimately when it gets built and a way to eek life out of a lot of dated components. In other words - I intend to crash this one a lot so i didn’t care if it was perfect… I may even name it the simulator.
Hmmm…wonder if I can matty flib backwards under that gap … better practice in the simulator first!


#36

I’m seriously considering those for my other 3" build. Nice weight savings. There is such a trend these days for ridiculous high kv motors that can’t get anywhere near their rated rpm once you put a prop on them in the 3" class. For the 11xx series it makes more sense cause there is very limited torque coming out of such a small diameter. The 1306 sized motors are kind of are medium soda … A little more torque but still rely on hot winds, high kv, and ludicrous inefficiency to get thrust. If I remember correctly, a 4000kv 1306 only actually makes 35000 rpm on 4s with aggressive props. 1407’s do a few thousand more but under 40000. 4000kv gives you a theoretical rpm of like 60000 - so that’s a big spread lost to heat and excessive amp draw. Those motors pull upwards of 16amps each compared to my 7amps. But, I benefit from the greater torque of a wider stator and get much closer to my theoretical rpm of 34000. I hovered it and did a few punch outs in the front yard. It holds its own even with the lower kv. Hover at under 40% and a disgusting punch out. Probably no speed demon in a big open field though.

Anyway, 1804 is on my radar for the next build. So is a 2700kv 1806 too though. :slight_smile: Nice suggestion - you’re thinking just like me!! I like it!


#37

Recent flight video showing proof of concept at 306g AUW (oh my thats heavy!!). 5min flight times and decent power. Next up - plan a similar build on 1806 trimming weight to as close to 200g as possible.

Sacrificing some flight time for weight savings with a 600 or 460mah battery will save a ton - 30g maybe
Lighter escs will save about 25g
Lighter frame will save at least 30g
Lighter FC / PDB will probably save 5 to 10g

There is room to improve here!!


#38

Feeling a certain attraction towards the brother hobby returner r4 2850kv 1806 for this project.
Not cheap, but could be repurposed on a super light 4" or 5" if it doesn’t turn out how I hope. Only problem is that I’m at a complete loss to anticipate current draw numbers on this motor for 3" and have no idea what esc to use. Originally I had planned the 1.2g 10a racerstar. This motor may pop that one.


#39

What would happen if one were to “over esc”?
Besides added weight, would there be a down side?


#40

Weight is the only down side. I’m still gonna go for it on the 10a 1.2g racerstar. Next lightest option for a higher rating costs 15g. That’s not nothing!