Latency comparison on FPV HMD screens/goggles, part 2

First part is here: http://lag.lt/latency-comparison-on-screens-for-hmd-fpv-use/

Got new firmware for my HMD, purchased some equipment for testing, borrowed goggles and did another measurement. Entire chain:

Sony Xperia Z1 compact => HS1177 => BF OSD => VTX03 => HMD/Goggles – so, keep in mind that visible latency is from entire chain included. Results in short:

  1. LTD056EV7F (my HMD with 5.6″ screen and RTD2662, firmware from smth like 2016.06): min: 29, max: 55, avg: 44ms. Main issue – blur on screen itself (old type LCD screen)
  2. Headplays with low latency firmware: min: 30, max: 69, avg: 39ms. Slightly less blurry than old Toshiba 5.6″, but more blur compared to any goggles.
  3. Quanum Fatshark Genesis: min: 45, max: 89, avg: 64ms. Again, less blurry compared to my HMD, but still more than goggles.
  4. Fatshark Dominator v2: min: 31, max: 51, avg 38ms.
  5. Fatshark Attitude v3 (should be the same optics dom v2?): min: 27, max: 69, avg: 50ms. Maybe optics is identical to doms, but something is different in processing chain, because averages are clearly higher.

Photo samples if someone is interested (taken with SONY A6300 + Samyang 8mm F2.8, ISO 25600 EV -2.7, but allways at 1/4000 shutter speed, so you know what quality to expect =)

Continue reading

Small airplane – looking for best motor/propeller combo – some benchmarks

Compared to situation few years ago – now you do not need to use 50g+ motor and 15g+ ESC  to get 1kg thrust. Mini quadcopter racing mania made small motors/ESC much better, cheaper and better in terms of thrust/weight. They are high-kV (usually) but for small planes this isn’t bad – you still need speed and not only static thrust.

So, my requirements for motor/esc combo:

  • Still have enough thrust for unlimited vertical (so at least 800g of thrust)
  • Have enough pitch speed for penetrating headwind and flying around (so at least 100km/h at WoT)
  • Efficient setup at low throttle – need ~40km range endurance (total trip, not one-side).

Continue reading

VTX power comparison – quick run #3

all

Got several new 5.8G VTX’es and decided to test their power output. Quick table for reference, but see notes below:

VTX dBm mW Comments
Aomway #1 (fat pigtail)
8 6,3
22 158,5
Aomway #2 (thin pigtail)
5 3,2
24 251,2
Aomway 4″ copter
-13 0,1 Dead PA…
-15 0 Dead PA
RP-SMA DIP BG: http://www.banggood.com/Skyzone-FPV-TS5813S-5_8G-25mW-40CH-Mini-AV-Wireless-Transmitter-For-Mini-Multicopter-QAV250-p-1021364.html
10 10
Foxeer 600mW:
23 199,5
Foxeer 200mW:
24 251,2
Foxeer switcher:
11 12,6 (floating – pwr1)
13 20 (floating – pwr2)
13 20 (floating – pwr3)
600mW: http://www.surveilzone.com/58G-32CH-AV-600mW-Mini-VTx-for-DJImini-Multirotor-g-1303
26 398,1
TS58400: http://www.banggood.com/Skyzone-TS58400-5_8G-400mW-32CH-AV-Transmitter-Module-p-1004170.html
24,5 281,8 (stable)
ET526: http://www.banggood.com/Eachine-ET526-5_8G-40CH-25mW200mW600mW-Switchable-FPV-Transmitter-p-1075702.html
13 20
22 158,5
26,5 446,7 Hot
FT48X: http://www.himodel.com/FPV_Telemetry/FPV_5.8G_0.25-600mW_Adjustable_48-Channel_A_V_Transmitter_VTX_FT48X.html
13 20 (stable)
21 125,9 (stable)
25,5 354,8 (stable, veryhot)
-5 0,3

In short:

  • My equipment is not calibrated for absolute power output, but is precise enough (~1dB) for relative power output. Have that in mind when comparing power output between different measurements.
    In other words – relative output between VTX’es is precise. Exact power (for example, 300mW) may be off (may be somewhere in 3dB range).
  • All measurements done at A1 channel (5865MHz), 2 minutes after powering up, stock heatsinks and no airflow; 20x 1-second sweep iterations, max power average
  • Third switchable (25/200mW) Aomway which dies in the same way. No, it was not powered without antenna. Sometimes PA just dies for some reason and it becomes unusable (module is sold on taobao, but not on public-accessible places, so… Expensive and useless)
  • Foxeer 600mW sucks, but nothing new (see earlier post about it)
  • Foxeer switcher is useless – switching maybe internal heat generator, but not power output. But note – i have first version, with SMA on side. Second version (SMA in middle) may work better (and yes, i unpowered/powered it back after changing power setting). Also, it’s power is jumping in +/-6dBm range all time, really random-power VTX
  • TS58400 is something i like for mid-range mini-airplanes. It’s compact, no excessive heating (but i’m using small heatsinks on 2 of them), has very stable power output (all other VTX’es change power with time due to heating) and very stable frequency. Too much power for race copters, but nice stuff for mid range
  • ET526 is nice, starts-up at zero power, need to press button to power up. Heats up a lot at max power setting, airflow required (or at least heatsink). Two buttons for configuration and strange channel layout (A1 == C1 here, etc)
  • FT48X – VTX with 5.3G band! Startup at 25mW, which may be good and bad at the same time, depends ;-) Two minor issues with it – very hot at max power (the same as ET526) and “no RF power setting” is actually “Low RF power”, which is still flyable few meters around (~15m range with omni antennas), so you MAY kill someone on race with that “no power mode”. Other than this… Nice VTX with mic.

Daugiau – jokių viešų renginių….

Kaip ir ne visai į asmeninį blog’ą reiktų rašyt, bet LT-FPV yra komanda, o tai kas yra žemiau – būtent mano asmeninė nuomonė.
Gyvenime buvo visko – IRC pjaudavomės iki pat aiškinimųsi tarp barakų (paprastai statybiorai laimėdavo, gaila =) , savaitėmis pjautynės niufsuose, ilgamečiai nesutarimai forumuose…
Paprastai tai viskas baigdavosi taikiai, kaip ten bebuvę – visi kaubojiški įgūdžiai išgaruoja momentaliai vos atsitraukus nuo klaviatūros.
Šįkart viskas kitaip. Deja, tenka pasakyt, kad pasinaudosiu klasika – “duok durniui kelią” ir atvirai pasakau – nebeorganizuosiu jokių viešų renginių.
Jau žinau, pažįstami paklaus – “tau tai nebūdinga, negi pasiduosi dėl tokių smulkmenų???”. Turiu aš tam atsakymą – pakol kitam “laido gale” yra kažkoks adekvatus oponentas, tegu ir visai bambukas, bet besivadovaujantis minimalia logika, ar turintis suvokimą kas yra gerai, o kas yra blogai – galima ginčytis. Galima diskutuot. Galima kovot, galų gale, galima ignoruot ir tikėtis adekvačios reakcijos.
Bet kai kitam gale yra psichiškai nesveikas žmogus, kuris normaliai net nesuvokia visuomenės elgesio normų, kuris yra nekontroliuojamas (tiek kitų žmonių, tiek savo sąmonės, jei ji ten yra), tai nebegalima žaisti. Yra ir dar kitas niuansas – aš kaip aš, bet yra ir man artimi žmonės, draugai, šeima – maža kas susišvies nepakaltinamam su ginklu….
Tai kur kas paprasčiau yra tiesiog pasitraukti ir tiek. Juo labiau – įsipareigojimų nėra, o ir pats procesas patapo kažkokia neaiškia, neaiškiai suvokiama pareiga. Aš visiškai neturiu patirties, kaip reikia bendrauti su nepakaltinamais ir elementaria logika nesivadovaujančiais žmonėmis – ir, tiesą sakant, neturiu noro tos patirties įgyti.
Continue reading

LT-FPV organizuojamos lenktynės

Aš aišku smarkiai stabdau su šiuo postu, nu bet geriau vėliau nei niekad.
Taip jau gavosi, kad suorganizavome mini-kopterių lenktynes:
DSC00382_visi
Ir ties tuom reikalai nesibaigė, bus dar vienos lenktynės (ir gal dar vienos… Bus matyt)
Na, bet projektas turi savo puslapį, kam įdomu: http://fpv.team/
Tam kartui – trumpas (dar nepadėtas archyve) filmukas iš buvusių lenktynių:

Prisijunkit, dar liko laiko užsisakyti įrangą, užsiregistruoti ir sudalyvauti =)

Surveilzone (Foxeer)/Aomway VTX comparison

Got ~10x Surveilzone (Foxeer) VTX’es around, from 25mW to 600mW, also 25/200mW Aomway and decided to check what’s inside.

Testing setup – RFE + attenuator chain. 60dB attenuation + 3dB offset in RFE, so in total you need to add ~63dB to visible averaged result on graph. All tests done on C1 channel (5705MHz).

Absolute power ouput is not precise, relative – much better (0.5dBm or better).

Cold testing – ~10s after powering on (already full averaged cycle), hot – after 5 minutes on my desk.

This looks like:

sz_tsetup

Continue reading

Latency comparison on screens (for HMD-FPV use)

I noticed slightly increased latency when flying with my HMD’s, compared to my old cinemizers (plus, not OLED). Not critical, flyable – but noticeable on corners and flying over obstacles at high speed.

Had few screens around, measured latency (just to remind – @25FPS one frame takes 40ms; at 50km/h you fly ~1.4m over 100ms), only few screens provided, i did > 100 photos and averaged:

Continue reading

5.8G analog receiver comparison (RX5808, RX5880 modules suitable for FPV and FR632 RX)

While 1.2G video is working very well for flying around obstacles, it also has own limitations. 2.4G video is somewhere in middle – small VTX, smaller antennas and good enough penetration, but unsuitable for flying in city (WiFi everywhere) and in meetings (2.4G RC adds a lot of noise even on 2370/2510 channels).

So, i moved back to 5.8G with some of my copters. Did simple RSSI-based diversity, added RX5808 modules to my HMD… Works good enough for everyday flying, but i noticed, that my friend can fly further using FR632 RX. LOS Range is acceptable for both, but he was flying further over obstacles  and in buildings.

Continue reading

Antennas for mini-copter – what we can do?

Long story short – started with 5.8GHz on mini FPV copter (pocket rocket quad), soon got another, tried 1.2GHz (see earlier videos here) and finally picked 2.4GHz for my video.

Things change with time, Lithuania now has #1 Wi-Fi coverage/speed in the world and now i found that 2.4GHz for video usually has lower range than old 5.8GHz setup. It still works fine in forests or far far away from populated places, but flying in city is next to impossible (for example, Vilnius Old Town, near Užupio Kūdrų parkas – loosing video in 20m away from me, using 2370MHz or 2510MHz which should be away from common Wi-Fi range).

Continue reading