We are glad to have you not just random mambo jumbo tuning things. Thanks for taking time and effort doing this!
@IshamPadron
4 жыл бұрын
Do you think it might be worth getting some logs from different people and compiling a datasheet of different TX and RX combos to see what the TX rate is? Would it vary with different RXs? I've got 7.69ms average Nirvana and X6B on Sbus
@peterowland1552
4 жыл бұрын
Awesome straight detailed info. I'm a basix qx7 accst amd and occasional r9 user. Already removed the adc filter from radio and it felt very smooth on my 3 inch toothpick.. will try the average 4 filter rc smoothing for sure. Thanks again Mark!
@jonathankrum3016
4 жыл бұрын
What kind of flysky where you testing? Mine sticks to 7.001 typically. I have seen it as high as 7.089 and as low as 6.88. Nirvana with the rs pro. No telemetry ibus.
@MUNKYFPV
4 жыл бұрын
Great video thanks. Do you think it’s possible to feel the R9 “heartbeat” while flying? It might be psychological but I thought I could feel a rhythm ticking along in the control link ever since I started to fly R9 and occasionally I get a random twitch out of the quad due to an erroneous signal being sent on a couple channels by the transmitter. BTW do you think turning off telemetry could get rid of those little twitches the R9 seems to cause? I like the much more reliable link when it comes to failsafe with the R9 over 2.4ghz but I don’t trust it for really long range and recently picked up one of the big crossfire modules from a friend who got himself a new Tango. I don’t think I’ll be switching everything over to crossfire yet though, just a couple of craft to see how I like it.
@uavtech
4 жыл бұрын
Every 200ms is 0.2 seconds. So IDK Turning off telemetry does not help. Tried it.
@MUNKYFPV
4 жыл бұрын
UAV Tech Thanks, yeah the logical part of my brain says there’s no way I should be able to detect that and it’s probably psychological. The weird glitch thing is my only big issue with the R9 and it’s not severe or often enough to cause problems as far as flying goes, it just sucks when filming smooth cinematic video and you get a little twitch during the flight.
@JimBoRamaWickyWo
4 жыл бұрын
Running ACCESS 2.1 on TX(Radio ISRM module v2.1.0) and R-XSR on 2.1 Opentx 2.3.7 I'm seeing Max high spike of 7.17 and lowest is 6.86 (Average 7.03). All running on the Taranis X9D Plus SE 2019. No variations through the whole flight even when pushing the range. So, yea. Seems to work. Thanks for the 4.2 explanation and making me finally getting around to confirm that this new stuff actually works!
@uavtech
4 жыл бұрын
Yes, looks like ACCESS is improved. Unfortunately, that doesn't help a lot of folks that don't have ACCESS radios.
@JimBoRamaWickyWo
4 жыл бұрын
@@uavtech Yea. I hear you. Kinda makes you wonder if they intentionally left ACCST hardware out in order to encourage people to upgrade, or is it a case of it just isn't capable. That said (And I'm sure I'll get flamed for this), I'm ok with them doing so in order to justify(read "pay for") innovation (developement costs).
@DonJohnsonFPV
4 жыл бұрын
Which hardware did you use for flysky ? From what I heared and feel deviation radio like t8sg have way lower latency... And last question, is it latency showing on betaflight cli when tiping "rc_smoothing_info"?
@uavtech
4 жыл бұрын
Yes, it is the same frame rate that is shown on the RC_Smoothing_info command. But of course the CLI command is not capturing during the entire flight. Folks were using OpenTX. I need to look into the latest OpenTX (2.3.7) and also see how ErSkyTX performs.
@DonJohnsonFPV
4 жыл бұрын
@@uavtech Thanks for your reply mark ! I just make some test with my jumper t8sg and latency is way lower than your testing on flysky protocol with constant 3.85ms no matter what the distancy between rx and tx. Same results for my friends on turnigy evo, 3.85ms on spi RX and 7.7ms with external RX. Frsky spi rx is 8.85ms fixed and crossfire vary from 6.6 up to 20ms on t8sg too. Just wait from my friend to report results with jumper t16 and crossfire vs crossfire shot...
@rcfpver4335
4 жыл бұрын
Hi Mark, I had a resonance at 54hz in a build I was trying to debug. I threw in a notch and it cleaned up the gyro, and it eliminated it even from gyro scaled. Someone just told me that resonance can also come from Rc command?!?! If I filter Rc command more, I can remove the resonance thus removing the bitches I put in? Does that makes sense? Can Rc command jitter create a noise bad enough to smoke motors?
@brechtfpv7480
4 жыл бұрын
Great video, I always learn a lot from them! I noticed that the framerate from my FlySky setup (Turnigy Evolution TX and Flit14+ receiver) is around 11 ms (11000 in the debug mode). Do you know why there is such a difference?
@uavtech
4 жыл бұрын
Maybe Sbus vs iBus? Could just be the Turnigy or the RX it uses has a faster frame rate.
@Rejdukien
4 жыл бұрын
FS-i6x and X6B: 7.7ms (7691 in debug), both with sbus and ibus. The only difference is that the endpoints need to be calibrated with sbus, whereas with ibus they're spot on 1000-2000.
@brechtfpv7480
4 жыл бұрын
Rejdukien Just tested the iA6C receiver and have around 8ms. So it seems to be the receiver?
@jenshenrikson7693
4 жыл бұрын
I used Frsky EU LBT on Jumper with r-xsr before I knew better. It was 16CH with a pretty slow refresh rate that disturbed the PID response badly. Looked into smoothing but found LBT-setting with 8CH instead. Would never have known this without your videos. Thanks! 👍
@srossi88
4 жыл бұрын
THANK YOU for putting this out there. I've been trying to preach this to everyone since I started to see the first posts on the blackbox FB group back when 4.2 first started development. As I said in our FB conversation, this issue has been exaggerated by the evolution of FF over the past year, and being addressed more fully with the new FF interpolated settings, IMO, is the biggest contributor of why 4.2 flies so well.
@CapnBry
4 жыл бұрын
FlySky is only 50Hz? Which FlySky protocol is this, because I fly all AFHDS2A with Flit10/Fli14+ and a Multiprotocol module and I see very consistent ~7.7ms frame times.
@IshamPadron
4 жыл бұрын
Just logged Flysky Nirvana running X6B does 7.69ms average.
@litovillarin7525
4 жыл бұрын
In your tests done, Is R9 at SBus or FPort mode. Thanks.
@MrSmithersFPV
4 жыл бұрын
your content is absolutely amazing!!!! thanks so much for all the hard work you put into this. Its like a supercharged JB channel!
@JusticeFPV
4 жыл бұрын
Have you tested ACCESS 2.1? the biggest steps I've seen in my testing so far are between 6.8 and 7.2ms with an consistent average of around 7ms.
@uavtech
4 жыл бұрын
I have not. Do you have a log of it? I really don't want to go through the trouble of updating - plus I have a damn QX7 and don't have damn R9M 2019 edition - but if I would go through the trouble, I don't want to be disappointed.
@JusticeFPV
4 жыл бұрын
@@uavtech Yes i have logs for ACCESS 2.1 with xrsr and opentx 2.3.7
@uavtech
4 жыл бұрын
@@JusticeFPV , if you could send to me at spatzengr at yahoo.com that would be awesome!
@JusticeFPV
4 жыл бұрын
@@uavtech On its way
@Tw1dg1t
4 жыл бұрын
You beat me to it! I wanted to see if the new ACCESS updates cleaned up the signal as it's been updated quite a few times.
@dududu7694
4 жыл бұрын
Thanks !! I just upgraded my Taranis to OpenTX 2.3.7 after seeing your video (was on 2.2.1). I'm on D16 EU/LBT, by the way. Do you know if it is better or worse than D16 FCC ?
@uavtech
4 жыл бұрын
Unsure. Need to do testing on OpenTX 2.3.7 as well. If not better, I'll try ErSkyTX as well.
@uavtech
4 жыл бұрын
Updating to OpentTX 2.3.7 did NOT fix R9m (ACCST version) at all.
@dududu7694
4 жыл бұрын
@@uavtech 😥 it is probably related to frsky rf protocol, rather than Taranis itself. I guess at least all frsky accst product will show the same behaviour...
@sketchpv3080
4 жыл бұрын
Mark- do you think the recently announced R9 SX and MX receivers will have more consistency in frame rate vs the current receivers? Or is perhaps the “heartbeat” frame rate pattern inherent in the hardware of the transmitter itself?
@uavtech
4 жыл бұрын
I'm not sure. Some are saying ACCESS 2.1 on OpenTX 2.3.7 doesn't have it, but I want to see a log before I invest in a 2019 Edition R9M module.
@Jesse-Rodriguez
4 жыл бұрын
How is the signal quality on Spektrum? Thanks for all the great vids. I'll try to see my logs. Just got to that part of the vid. Thanks bro
@uavtech
4 жыл бұрын
I have to look it again. But if i recall, it was consistent. Some of my issue is i only havr a DX6i, so frame rate is 22ms. If you have a better Spektrum TX, i would be interested in a log of the frame rate from the girl.
@BrandonBeans
4 жыл бұрын
@@uavtech @UAV Tech my dx6i's are both 22ms dsm2. I've set to average 3 just because I assumed my rc signal was crappy. I guess I could just lower it and check for unwanted movement.
@fpvattitude8163
4 жыл бұрын
Love what ya do man you make it so much more manageable ..plus ya give bardwell the opportunity to dumb it down even more but i enjoy being able to tune my kwad with help help of someone that i consider a pro!
@uavtech
4 жыл бұрын
Thanks. Yeah, i figure there is room for someone to take it to a little deeper level.
@oyvinla10
4 жыл бұрын
Have you tried logging and see if there is any differences in consistency running 2.4ghz frsky like the r-xsr with ACCESS over R9?
@uavtech
4 жыл бұрын
r-xsr on the ACCST suffers from the same issue. I have not seen r-xsr on ACCESS.
@illusionfpv708
4 жыл бұрын
@@uavtech strange, I have no problems with r-xsr and accst, otx2.3+ has hearbeat for internal module
@grounddodger3212
4 жыл бұрын
@@illusionfpv708Any update on whether this effect with R9 ACCESS?
@poorhaas
4 жыл бұрын
Is there a difference between the tango 2 results and a radio with a micro crsf module in the back ?
@zoobydooby2302
4 жыл бұрын
I've successfully upgraded my taranis to 2.3.7 . In the 4.2 tuning notes. It said to disable the ADC.. What's that? and how? Maybe you can help? Thanks ahead.
@Tonynux
3 жыл бұрын
In the radio setup section (in your Taranis) there is a page called HARDWARE, and there you can uncheck the ADC_Filter option.
@edouardmalot51
4 жыл бұрын
Thanks mark. I had many issue with R9M when I set high FF. On logs you can see that the framerate : 6ms/6ms/13ms really add back effect on your flips
@mike216ism
4 жыл бұрын
I have a taranis qx7 and rxsr receiver. So should I be using 4 point averaging? I really don't want to buy a new radio to love my FF. I do feel the dropped frames. When I reduce FF it gets smoother. I prefer running higher FF. This might have been the missing tuning problem I've been fighting for at least a year Edit. Average 4 it is. Can't wait to try when I get home from work. I hope this works and I can bump back up FF
@skirmich
4 жыл бұрын
I am also an R9 User using 8 channel mode and F.Port. What filters do you recommend?? I am using input 15hz and derivative 20hz
@uavtech
4 жыл бұрын
Set all to Auto and use Aveeage_4.
@skirmich
4 жыл бұрын
@@uavtech I am still on BF4.1.5 though 10" are really hard to tune and I am really not in the mood to try a new target.
@Fumes74
4 жыл бұрын
Damn this is good. Nice explanation, thank you!
@prozay6856
4 жыл бұрын
i was using interpolation... thanks for the tip.
@anthonyshort8957
4 жыл бұрын
Very helpful, thank you.
@h2o-fpv623
4 жыл бұрын
Thank u for a great video.
@nyx574
4 жыл бұрын
I thought the Tango 2 was running CRFS shot, should we still lock it at 150 hertz?
@uavtech
4 жыл бұрын
If you have, check it out and see what you get.
@Dorkyand40
4 жыл бұрын
turn off tele too. remember if you lock 150hz you will be losing range. for freestyle i would leave 150hz off and just roll with it but def run 8 channel only and 150hz off per TBS
@genivos
4 жыл бұрын
@@Dorkyand40 so, what's the recommendation now? As I understand for my situation, atm no crsf possibility (FlySky NV14 with Crossfir Micro TX) and solely freestyle flying. From your note I see that I need to set it to: 50Hz locked and 8ch (tele can be on or also off? what's the catch if either on or off?). I had it all the time dynamic (power and Hz) and 12ch. Last weekend I set it to dynamic power and 50Hz locked but still 12 channel. I'm trying to find out what's best but I can't find a clear instruction on this anywhere so far. I'm planning to go crsf shot in long term but for now I'm limited to crsf v2.
@JavisHD
4 жыл бұрын
So which ff_interpolated_sp is recommended for Dji fast? None? Thanks!
@uavtech
4 жыл бұрын
Average_2 seems to work well. That are ON. ON is just interpolation without Averaging.
@ehunter2010ed
4 жыл бұрын
I'm flying astro x5 with caddx vista katoka holybro f7 32 b esc 2700kv brother hobby deadpool motors. I'm getting horrible wobbles & bobbles running 4.2 newest FW where should I start??? I've refreshed the FW no adjustment yet where should I start??
@uavtech
4 жыл бұрын
Tune it. I would start with moving up PD Gain and PD Balance sliders. One and then the other. See tiny.cc/pidtuning for the BEST tuning advice to get a well performing craft.
@ehunter2010ed
4 жыл бұрын
@@uavtech can u make video on tuning like ur 8 page tune!! If you do already let me know!! I rather listen to podcast, video vs reading!!
@MattVentureNL
4 жыл бұрын
First time I see ersky - would this kinda of transmission latency be an improvement over the access protocol?
@uavtech
4 жыл бұрын
Unknown.
@MattVentureNL
4 жыл бұрын
@@uavtech could you please share the link/source of where you got the diagram with the transmission latency comparison from?
Thanks for your video. Btw I have tried 4.2 and there is badly propwash any solution ?
@uavtech
4 жыл бұрын
Tune it. Tune up PD Gain and/or PD Balance slider.
@hateeternalmaver
4 жыл бұрын
Sucks that R9 is so dirty... I mean I can't compare it to anything else but I'm still very satisfied with it and I can't really say it's flying as horrible as it's looking in the graphs... ;) Thanks for showing the actual data.
@flo-ridafpv5713
4 жыл бұрын
hateeternalmaver agree its butter for me, but i run low FF, i hear that helps. & his vid after this shows how to average r9 spikes in cli, setting #4
@mike216ism
4 жыл бұрын
I think this is why I get occasional bounce back with default FF values. No matter what I've done it occasionally happens. I'm wondering if my really old open TX software is the problem. Incredible find. Thanks for another excellent find
@uavtech
4 жыл бұрын
Could be this or motor just getting saturated some times too.
@mike216ism
4 жыл бұрын
@@uavtech Im using t motor f55 pro 2 f3 mcu 4-1 esc and f60 pro 4 1950kv 6s motors. Even these strong motors get saturated?
@uavtech
4 жыл бұрын
If a heavy quad, sure. If really light, not likely.
@mike216ism
4 жыл бұрын
@@uavtech really light. I'm using raging droner. No go pro. Around 350gram. Appreciate your time as always
@TheMadmacs
4 жыл бұрын
i vaguely recall joshua mentioning flysky is doing something different with rc interpolation that makes it smoother, i also thought drone mesh measured flysky as fast as crossfire almost?? all this makes me wonder if those days (or first few packs) i feel less connected to the controls could be a matter of interpolation
@uavtech
4 жыл бұрын
I don't know about those things. I just know Frsky has has inconsistent packet trsansmittion for YEARS!
@IshamPadron
4 жыл бұрын
10:34 is the flysky running in ibus or sbus? BTW in the intro it says Frsky but then at 10:34 it says flysky.
@uavtech
4 жыл бұрын
IBus. Yeah, saw the mistake in recording and fixed it. Good eye though.
@IshamPadron
4 жыл бұрын
@@uavtech I wonder what the difference is with Flysky sbus. I mostly run my flysky rx in sbus. I guess i'll log it and see for myself.
@PIDtoolbox
4 жыл бұрын
4:00 something that always confused me here. how do we know for sure these differences in stair duration are due to packet delay vs just slower stick change at that moment? I mean if I moved the stick slow but at an inconsistent velocity, it would produce a similar pattern because the "bit depth" if you will is very coarse so it's gonna step up depending on how long you hold a stick position at any given moment, or how slow the movement transitions from one level to the next. I guess what I'm saying is this is kind of an indirect measure of actual packet times. I hope this makes sense. You can imagine for example in an extreme, if you hold the stick still, it's a straight line. how do we know the user didnt just hold for a moment producing a slightly longer stair?
@uavtech
4 жыл бұрын
RC_Smoothing_Rate Debug measures frame rate, which with the ADC filter off in OpenTX, is should be consistent regardless of value. We can see this when at 0 stick and the frame rate comes in consistently for some TX's and not others. So you can see, with the debug, it don't matter if you are moving the sticks or not.
@TaaviTuisk
4 жыл бұрын
Any differences between R9M and R9M lite? Accst vs Access?
@uavtech
4 жыл бұрын
I don't have all those variants. If you have one, run it and let me know what you use. This was R9M , ACCST on OpenTX 2.3
@TaaviTuisk
4 жыл бұрын
@@uavtech Any hints on how to test this? I will gladly do the tests.
@uavtech
4 жыл бұрын
Just log RC_Smoothing_Rate for a normal flight.
@TaaviTuisk
4 жыл бұрын
Mark Spatz posted a comparison between r9 accst and r9 access in r9 group in facebook. Can't post images here unfortunately. I'll do my own testing once i can fly this weekend.
@uavtech
4 жыл бұрын
I am Mark Spatz. Geny below commented he saw better results in R9 ACCESS and sent me his log. I only have R9 ACCST.
@sorenlike5242
4 жыл бұрын
Hey, can you show us how to change our vTX via lua Script on the transmitter? I can't find anything on the internet :/ and I just want to change my vTX Power via a Switch
@TheXanUser
4 жыл бұрын
the main reason 3.5.7 is used so often as a reference is its a bit harder to set up wrong. theres less ways to set it up so it damages something when youre a dummy/noob. especially over a large range of different size gear. Also there are so many decent tunes and working stock configs out there for so many quads that can just be copy/pasted. especially for those new pilots that find out the hard way what the BF devs really mean by "reset". IMO at this point betaflight really needs a simple mode for those that haven't been around and seen all the changes since ~3.2. the sliders are a big improvement. could even use the old moron threshold as a tribute. a whole fork called moron flight! Hopefully someday a BBlog or 3 can be read by a comp and then just spit out a decent fixed CLI to paste. (locked up at home drinking and smoking)
@uavtech
4 жыл бұрын
Folks should check out Cleanflight. Old and never changes. I will do a video on why 3.5.7 is subpar. Stay safe amigo!
@tweekFPV
4 жыл бұрын
👍👍
@Nikita-zd1gc
4 жыл бұрын
so... just bought r9 lite for xlite . Pls make a video about r9+averaged_4)) as it the most unconsistent protocol....
@TaaviTuisk
4 жыл бұрын
You can load access firmware on the r9m lite.
@mistral7280
4 жыл бұрын
thank you for this very interesting video. Maybe the R9M sends the frames every 6ms. approximately every 10 frames, the R9M does not send a packet so that the R9MM sends a telemetry frame to the R9M which takes 2 cycles of 6 ms. then the R9M sends a frame again. that makes 2 frames not transmitted. therefore 20 ms between two frames. I will test by deactivating the telemetry in my QX7 and in betaflight to check if what I say is true.
@tstrrtstrr745
4 жыл бұрын
Turning off telemetry doesn't help. This got fixed with ACCESS though, but I would have to buy a new transmitter and the new R9, so I will just buy Crossfire and be done with it.
@Nikita-zd1gc
4 жыл бұрын
@@tstrrtstrr745 so as yu say access firmware dont have dirty wobbles in rc signal?
@uavtech
4 жыл бұрын
If actual fixed in ACCESS that is good, but I want proof. I think i would be in the same boat with a QX7 and R9M prior to the 2019 edition. I have tried with the Telemetry ON or OFF and still see the inconsistency.
@azfpv5967
4 жыл бұрын
Been using ErskyTx, formerly Ersky9x for five years now. Ersky ftw. In my view, much easier to use than opentx, I'm biased though. I have a playlist on using ErskyTx on my channel in case you want to check it out
@uavtech
4 жыл бұрын
I will. Maybe with a switch to see if it helps with R9.
@azfpv5967
4 жыл бұрын
@@uavtech just an FYI, I inquired with the developer of ErskyTX and this is what he had to say about ADC filter on ErskyTx, www.openrcforums.com/forum/viewtopic.php?f=7&t=4&start=2010#p150058
@RobertLeclercq
4 жыл бұрын
Mark 3.5.7 produced better footage for 6"+ quads. When I say better I am talking about straight line, no bobble, jitter, bumpiness, whatever you want to call it. Mid range and long range groups on facebook all love 3.5.7 for the same reason, the HD footage. 4.2 is fixes much of the problems 4.1 introduced that 3.5.7 flyers hated, but it still can't give us that butter smooth straight line HD footage 3.5.7 can. For you to dismiss them shows a real lack of understanding. I've learned tons from you, but man if you listened to those haters you would probably learn a thing or two as well. The people flying 6"+ producing amazing footage are saying different things from the you and the devs, both are right in there own regard. Come together and fly(metaphorically), don't dismiss them/us with 'your just wrong'.
@RobertLeclercq
4 жыл бұрын
PS I prefer 4.x only because the flight feel was better and I didn't mind using stabilization. Some folks hate stabilization and 4.xx to go along with.
@uavtech
4 жыл бұрын
That is not a version differential, that is just a default differential. Knowing the inner workings of the firmware, 4.0, 4.1, and 4.2 are just better in so many ways. Folks should open up to putting 3.5.7 settings in a future releases to prove for themself; or prove me wrong. It is just PIDs and Filter settings making the difference.
@flo-ridafpv5713
4 жыл бұрын
I felt same way....until i learned to tune 4.1, now its soo money ...and on 1 rig i fly 6in bi blades, almost same tune as 5.1"
@stevenfpv2345
4 жыл бұрын
I checked my radiolink SBUS 10CH.. 13ms constant! LoL
Пікірлер: 114