Thanks, bro. I hope it can help people to build their own… 
2 Likes
Hi Maroquio, could you post your param file for your sub 250 ? Thx
Of course, Hugues.
Regards.
1 Like
Thank you. What firmware version does it correspond to ? and in 2S voltage, right ?
Do you get a flyable ship (external flights, slight winds) and poshold modes working right with these parameters ?
Yes… it is working nice. I’m using firmware 3.7 from June 15 and 2S battery. PosHold and Loiter working with and without GPS.
I can post my latest param file later, too… I fly it daily. Heavy winds (Wind doesn’t seem to bother it as much as a larger copter), all modes, auto missions, etc. Everything works great. 135mm frame. 3" props. 2s.
6 min flights with my 950mah lipo and fast flying. 18 minute hover w/ 3000mah 18650’s.
Have you tried althold based on baro only (no lidar, no optical flow) ? I can’t seem to have any alt hold working on mine : Kakute F7 with its integrated baro (with or without covering foam), 2S, firmware 3.7
nice performance. Would be great to check your params too. This would compile into a useful database for people building 3" arduships. Good flying time too with 18min. Your total weight without battery= ?
Yes… It was working with baro in AltHold, but always with vertical step up/step down variation (around 40cm). I think that it can be configured (parameters) to discard very subtle measures, but I did not try.
Total Weight w/out battery now is 123 grams… It has gotten heavier over the past year… I was right around 100g when I first built it… (Now using diversity RX, different GPS, glue and various repairs, etc).
That 18 minute time is literally just hovering and the flight ended when the ESC cut out… I have flown for about 12 minutes and landed before the battery died. I’m not super comfortable w/ the 18650’s yet, because the voltage sags a lot, and I don’t really know the limits. I know I was flying around at about 5.7v when I decided to land.
I agree about compiling all the info we can… These small copters take considerable tuning… More than just PID’s. I think Andy has a well performing 3" copter, too, so maybe he’ll share his config file if he sees this.
My baro only alt hold performance is good… If I’m above all the ground disturbances it seems to hold its position quite well. I’ve not measured it, but I’d say 40cm at worst. Maybe even more like 20cm vertical drift. Basically, nothing bad enough to notice or worry about for what I’m doing. I did tune those parameters… There’s a filter in I think the PSC settings that needs adjustment, and the related PID’s. I don’t remember the names… I’ll get a parameter file uploaded after work. My first flights were not successful! I basically destroyed my copter the first time I tried alt_hold! It flew up and hit my ceiling hard. It took me a couple days of tuning to get alt_hold working well.
Here’s a flight w/ this particular copter. Just to show that it really works, I guess :). It’s not the same copter as in my first post in this thread, or even the same copter as the video from December I posted… This is iteration #3.
1 Like
Mine is 4", I am in the process of ordering the parts for a 3" Gecko, but also doing a lot of heavy hacking with my harmonic notch PR and FFT support (which is where all the time goes!). Since I last posted params I have been able to make some fairly significant change using the harmonic notch, so don’t expect these to work on something without special firmware. Working on removing oscillations at full throttle. The most pleasing change was ATC_INPUT_TC = 0.08 which improved responsiveness in acro a lot
QAV180-TunedDevFFT.param (14.7 KB)
Hi all
I have a MicroArducopter with 3" props that I tuned with my experience and following the hints on this thread and the wiki page from @Leonardthall .
I obtained a good flying Microcopter both in STABILEZE, ALTHOLD, LOITER and GUIDED (also indoor with aruco marker http://ardupilot.org/dev/docs/ros-aruco-detection.html).
Recently one of my ESC (DYS XM10 non_S non_32 maximum Oneshot125) died and so I changed all 4 with 4 in 1 Turnigy Multistar BLheli_32.
Now I have a BIG problem: with the same frame, same motor, same everything except the ESC my quad is OK only in STABILIZE mode but if I try to switch in ALTHOLD the copter goes to the roof like a rocket.
Unfortunately I only have a tlog and no dataflash log. I attach this tlog, from line 238500 to line 242500 (no time in this tlog) you can see a little indoor flight: takeoff in STABILIZE, a little bit of hover with motor around 1500 - 1600, then when I switch to ALTHOLD the motor goes immediately near 1900 till I switch back to STABILIZE.
I did several controls on the various parts of my copter, connections, firmware, parameters but I did not find any difference beetewen before and after the ESC change.
Link to tlog
Lets start w/ the basics… Any chance something physically changed w/ your barometer? Or maybe the ESC is larger and funneling more air into the barometer? On one of my copters, I had to put foam around all four edges of the FC, between the ESC and FC.
Tanks @wicked1
The FC is an openpilot revolution within its case. The case enclose well the FC and there is foam over the barometer. All this is in the same place as before and with the same relative distance to frame, ESC, props.
EDIT
what I don’t understand is that in the past I had problems with stability in ALTHOLD, because of vibration or barometer with too little foam, but I never see the copter rocketing to roof.
You were lucky you had a roof. I almost lost my big 650mm quad on the first test. From 0 to 120m in 3 seconds… take a look… 
Cheers!
I had the same w/ a 650mm copter… I had a motor kill switch programmed and I panicked and killed the motors… I tried to restart them when it was close to the ground, but it wasn’t soon enough and the frame was destroyed.
And like I mentioned earlier in this thread, the same happened w/ my very first microcopter and it destroyed itself slamming into my ceiling and then the floor.
So, common issue
It’s usually wind over the baro or vibrations. If you’re sure vibrations are OK, I’ve found increasing this filter helps: PSC_ACCZ_FILT
hi, i did not look at your logs, but, if the symptom you have is the uncontrollable rapid ascent in the pos hold mode - it is usually a side effect of extremely spiked vibrations that prevent normal operations. the threshold from normal to crazy is difficult to establish, but, if you get a situation when a single input from stationery results in the drone suddenly darting up into the sky - that is it.
solutions are to use motors that are less notchy, run 3 blade props and always mount FC on 5mm rubber posts. also check if any new wiring now connects or transmits extra vibrations. motor wires would be the primary suspect.
Thanks a lot, I will check for vibrations, I will do other tests and I will report here.
Okay. I will try to make some tests with this param when possible.
My problem was the PWM learning for hovering flight. It was too high for my setup, because that was the first flight and I setted to AltHold without waiting a bit for the learning process. Anyway, it’s good to know about this param too.
Thanks. Regards!