• If you are new to GiantScaleNews.com, please register, introduce yourself, and make yourself at home.

    We're 1st in Giant Scale RC because we've got the best membership on the internet! Take a look around and don't forget to register to get all of the benefits of GSN membership!

    Welcome!

Radio Wars, Part 2

gyro

GSN Contributor
Guys, I'm horribly sorry. When I was promoting SteveB's excellent post showing how many feet and airplane will travel with slow latency, I accidently deleted this whole thread. I was able to recover the below, which should keep things going for a bit, but again, my fault, I really messed up.


  • Yesterday, 08:49 PM #45
    Joe Smith
    Originally Posted by 3Daddict
    It is not slow, I had a spectrum dx7 before the A9 and the A9 feels more locked in than my spectrum dx7 ever did. However I guess that is what a futaba fan would say. I know a fairly good flyer that flys nothing but futaba at my field who flew one of my airplanes with my A9 and told me that my A9 felt the same as his futaba if not better. not to mention Hitec is a lot cheaper than futaba and just the same quality.

    Fortunately, the amount of latency in our systems can be found through tests so we can acutally prove which radios are faster than others.

    Take a look at this.
    http://rc.runryder.com/helicopter/t172571p1/

    As you can see, the Aurora 9 is up there with the slowest radios on the market. I've had the opportunity to put many months on a dx6,dx7, jr12x, aurora 9, 8fg, and the 18mz. I can honestly say that performance alone is the only reason I am still flying futaba.

    Also, what information do you have to back up saying that the quality of hitec products is the same as futaba? All of the testing I have done would easily disprove this statement but I won't say you are wrong. I am just wondering where you got your information?

  • Yesterday, 08:52 PM #46
    3Daddict
    I am just messing around also, I have to say I am a fan of futaba and may get one as my next radio in a few years. I just figured since this is the radio wars thread I would stir the pot a little.

  • Yesterday, 08:54 PM #47
    3Daddict
    none, just the feel of the radio. Maybe I came down to harsh on futaba. But I have to say that futaba and hitec both make some of the best radios on the market.

  • Yesterday, 08:54 PM #48
    Jim Smith
    Originally Posted by 3Daddict
    I am just messing around also, I have to say I am a fan of futaba and may get one as my next radio in a few years. I just figured since this is the radio wars thread I would stir the pot a little.

    ROTFLOL
    Score

  • Yesterday, 09:00 PM #49
    3Daddict
    dyAwijazqEweSeooCqmBIAAODEBVKtXNXHAAAAAElFTkSuQmCC

    Ok, I will admit that futaba is the best. However I still maintain that the A9 feels more locked in than my old dx7. You win Jim. I bow out of this radio wars.

  • Yesterday, 09:07 PM #50
    GYRO
    My Dx8 has been pretty reliable so far, and in 11ms High Speed mode, is plenty fast for my weak skillz.

  • Today, 05:30 AM #51
    Manta
    LOL Tell you guys a secret,my A9 with the XPS module and nano receivers feels the same as Losi's futaba 8FG ( which I really liked except for hand position on radio) but as far as the "locked in" feeling,same.. specky stuff doesnt "feel" as locked in as the futaba and the A9 with xps.. the hitec module and receivers tho felt goos,but the lag is noticeable for the high alpha stuff. I felt like I was chasing the plane..but this coulda been lack of skills too but it changed when i went to the faster XPS system.
    The 8fg I could get used to hand/finger/switch position with a few flights. I call my A9 better because of the touch screen and the rx are 30 bucks..more bang for the buck. and I love my touch screen. But i wouldn't turn an 8fg or higher end futaba down either. just the prices are ridiculous.
    RJgAJZdlsYzbjOrAAAAAElFTkSuQmCC
    Team Mascot for Motrolfly

  • Today, 06:12 AM #52
    3Daddict
    I have to disagree about it feeling like you are chasing your plane in high alpha 3d, all I do is 3d and I have never once felt behind my plane. I always feel locked in.

  • Today, 07:11 AM #53
    Steve_B
    This talk about letency misses what is the biggest contributer to latency.. the servo itself! Latency is defined as the time taken between a command and the control surface reaching it's commanded position so you really need to take the servo into consideration.

    If you take the latency data linked previously by Jim, the fastest systems (example; Futaba 8FG and 10C, Spektrum DX8 and DX7se, ATX SD-10G) fall in a latency band 10 to 20ms. Compare that to the slowest systems (example; Hitec Aurora 9, JR 9303, the old type Spektrum DX6) are in a band 50+ms

    The average difference between the best systems and the worst is about 35-40ms

    Now lets look at a servo.. To travel 60 deg under no load here are some typical times (small sample of servos i have experience of):

    Savox SH-0257: 90ms
    Savox SH-0255: 130ms
    Hitec HS-65MG: 130ms
    Hitec HS-85MG: 160ms

    So taking the best to the worst of these times the difference is 70ms, so the contributed to overall latency caused by the servo is potentially significantly greater than that of the radio. It's perfectly feasable that the guy with the slow Aurora 9 will in fact have better overall responsiveness than the guy with the Futaba 8FG purely because the Hitec flyer has faster servos.

    Having said that, lower radio system latency is better and a difference of 35-40ms is quite significant and must be noticable in flight, all other factors (including servos) being the same. Hitec need to get it sorted!​
    a.k.a 'JetPlaneFlyer'

  • Today, 07:28 AM #54
    teookie
    yeah but,

    the plane doesn't wait for the servo to get all the way to it's commanded position before it reacts. As soon as the control surface moves 1 or 2 degrees towards it's commanded position the plane will start reacting.

    so, time from stick movement to the plane doing something is still largely dependent on TX latency.

    what fast servo's do is allow you to change direction really quick and get large deflections very quickly.

  • Today, 07:47 AM #55
    Manta
    Yea hybrid hitec xps man!!! Woot!!

    Jim drew needs to quit smooozing that movie guy and his aquatic quests and get his miracle tx to market.
    Lol
    Team Mascot for Motrolfly

  • Today, 09:46 AM #56
    Steve_B
  • Originally Posted by teookie
    yeah but,

    the plane doesn't wait for the servo to get all the way to it's commanded position before it reacts. As soon as the control surface moves 1 or 2 degrees towards it's commanded position the plane will start reacting.



    That's true but the plane with the faster servos will always have less lag in it's response to stick input because the surface will move faster compared to the plane with slow servos (all other things being equal). Faster reacting control surfaces means faster reacting airplane so servo speed is part of overall latency, no getting away from it.
    In practice i bet it would be impossible to tell the difference between a high latency radio and a slow servo, they would both make the plane feel sluggish, especially so in 3D where we need the servo to move a long way.
    Last edited by Steve_B; Today at 09:49 AM.​
    a.k.a 'JetPlaneFlyer'

  • Today, 10:26 AM #57
    Jim Smith
    Originally Posted by Steve_B
    That's true but the plane with the faster servos will always have less lag in it's response to stick input because the surface will move faster compared to the plane with slow servos (all other things being equal). Faster reacting control surfaces means faster reacting airplane so servo speed is part of overall latency, no getting away from it.
    In practice i bet it would be impossible to tell the difference between a high latency radio and a slow servo, they would both make the plane feel sluggish, especially so in 3D where we need the servo to move a long way.



    Built 2 identical planes. Same servos ESC, etc.
    1 Flown with A9
    1 Flown with 8FG

    Lag was noticiable enough that a random spectator could tell which radio was being used. Didn't even need to be using the radio, just observing how the plane reacted in the stunts being flown.
  • Today, 11:56 AM #58
    3Daddict
    Question for you Jim. I got to thinking about it and I only fly 57'' and under planes which are quick responding as it is. Perhaps the lag is not noticeable to me because of this, and maybe it is more noticeable on a larger plane. What do you think?​
  • Today, 12:18 PM #59
    hone1er
    Originally Posted by Jim Smith
    Built 2 identical planes. Same servos ESC, etc.
    1 Flown with A9
    1 Flown with 8FG

    Lag was noticiable enough that a random spectator could tell which radio was being used. Didn't even need to be using the radio, just observing how the plane reacted in the stunts being flown.

    That's one skilled spectator. They must watch a lot of flights

    The A9 has been working great for me but i would like to try a Futaba radio someday as well just to see if i really notice the difference.
    Last edited by hone1er; Today at 12:22 PM.​
    We're all here for 3D :cool:

  • Today, 12:54 PM #60
    Steve_B
  • Originally Posted by hone1er
    ... i would like to try a Futaba radio someday as well just to see if i really notice the difference.



    Doesn't have to be Futaba. The spektrom DX7se or DX8 in 11ms mode are pretty much as fast as Futaba (a bit faster than some Futaba models marginally slower than others). To be honest though I bet there is some considerable element of subjectivity in what 'feels' faster especially when we are talking about differences of a hundredth of a second or less.

    Bear in mind that the average human reaction time between an event and the finger reacting is 215ms: http://www.humanbenchmark.com/tests/...time/stats.php , and that's when we are concentrating 100% and expecting the 'event' to occur.
    a.k.a 'JetPlaneFlyer'
    okQAAAABJRU5ErkJggg==


  • Today, 01:04 PM #61
    Jim Smith
    Bear in mind that the average human reaction time between an event and the finger reacting is 215ms.
    But Average Humans don't fly like this


  • Today, 01:23 PM #62
    Steve_B

  • Originally Posted by Jim Smith
But Average Humans don't fly like this



Indeed not but I'm not sure how much of that outstanding ability is due to reaction time. Seems to me that most inputs during acrobatics are planned in advance, even if only a fraction of a second in advance. It's maybe more about predicting when the plane will do and planning ahead than it is to reacting to events.. No?

Having said that having super fast reactions must help a bit; so how does Joe score on the human reaction test: http://www.humanbenchmark.com/tests/...time/index.php
Last edited by Steve_B; Today at 01:24 PM.​
a.k.a 'JetPlaneFlyer'
okQAAAABJRU5ErkJggg==




  • Today, 01:32 PM #63
    Steve_B
    FWIW my I cant get human reaction score much below 250:( That might explain why my 3D sucks:(
    a.k.a 'JetPlaneFlyer'
    okQAAAABJRU5ErkJggg==


  • Today, 01:34 PM #64
    Joe Smith

Originally Posted by Steve_B
FWIW my I cant get human reaction score much below 250:( That might explain why my 3D sucks:(



I just tried it too. All I will say is that you can't use 250 as an excuse. :rolleyes:






  • Originally Posted by Steve_B
    Indeed not but I'm not sure how much of that outstanding ability is due to reaction time. Seems to me that most inputs during acrobatics are planned in advance, even if only a fraction of a second in advance. It's maybe more about predicting when the plane will do and planning ahead than it is to reacting to events.. No?

    Having said that having super fast reactions must help a bit; so how does Joe score on the human reaction test: http://www.humanbenchmark.com/tests/...time/index.php


    I cut a lot of it out of the video, but left some in. Reaction is when your flying your plane in a direction and all of a sudden a plane is in your predetermined aerobatic maneuver. Reaction time keeps you from building a new plane. Or when your torque rolling a few inches off the deck and a gust of wind hits. People think that once you learn the stirr correctly, reaction time isn't necessary. Try a rolling loop flying both directions. Does both of yours look the same. Only if you are controlling the plane by reacting to minut variations will they both be the same.
    I calculated it once. But with the A9 the plane would fly 6' before reacting to stick input, versa 1.5 feet for the 8fg. Have you seen a heartattack maneuver. 6' would mean a rebuild.

  • Today, 02:20 PM #66
    Steve_B
    Jim,

    how far the plane would fly between the stick moving and the servo receiving a signal to move is an interesting point and maybe helps put some reality into pretty much meaningless talk of a few ms here and there.

    Of course it totally depends on how fast the plane is flying! 3D model aren't really so fast so what about if we take 30mph as an average speed half way between walking speed harrier and WOT high speed manoeuvres?

    30mph is 44ft per second (5280/60^2)

    Based on the data shown in the linked thread:
    The Hitec Aurora 9 being about the slowest of the current systems is 55ms latency. Thats 0.055 seconds so the plane travels 2.42ft
    The Airtronics SD-10G is the fastest at 10.3ms. that's 0.01 seconds so the plane travels 0.44ft or 5.3"

    So I calculate the difference between the fastest and the slowest radios at 30mph flying speed, everything else being exactly equal, to be near enough 2ft (24").. That could very easily be enough to make the difference between a crash and a near miss. It's hard to see how a 2.42ft departure from the desired flight path would not be noticeable:(

    The Futaba 8FG (in high speed mode) plane would travel 0.62ft (7.4") and the Spektrum DX8 (in 11ms mode) plane would travel 0.84ft (10.1").. so not much in it at the faster end of the scale.​
    Last edited by Steve_B; Today at 02:25 PM.​
    a.k.a 'JetPlaneFlyer'

  • Today, 08:33 PM #67
    Jim Smith
    Originally Posted by Steve_B
    Jim,

    how far the plane would fly between the stick moving and the servo receiving a signal to move is an interesting point and maybe helps put some reality into pretty much meaningless talk of a few ms here and there.

    Of course it totally depends on how fast the plane is flying! 3D model aren't really so fast so what about if we take 30mph as an average speed half way between walking speed harrier and WOT high speed manoeuvres?

    30mph is 44ft per second (5280/60^2)

    Based on the data shown in the linked thread:
    The Hitec Aurora 9 being about the slowest of the current systems is 55ms latency. Thats 0.055 seconds so the plane travels 2.42ft
    The Airtronics SD-10G is the fastest at 10.3ms. that's 0.01 seconds so the plane travels 0.44ft or 5.3"

    So I calculate the difference between the fastest and the slowest radios at 30mph flying speed, everything else being exactly equal, to be near enough 2ft (24").. That could very easily be enough to make the difference between a crash and a near miss. It's hard to see how a 2.42ft departure from the desired flight path would not be noticeable:(

    The Futaba 8FG (in high speed mode) plane would travel 0.62ft (7.4") and the Spektrum DX8 (in 11ms mode) plane would travel 0.84ft (10.1").. so not much in it at the faster end of the scale.


    I really hope everyone reads your post carefully. Its exactly what we are discussing. Latency is = to distance traveled.
    2.4' versus .62' is a bunch when your flying 6" about the ground. I really hope people understand that latency is a real number. Not imaginary at all.
  • Today, 08:43 PM #68
    GYRO

    excellent posts and information Jim and Steve. That is exactly the way to sort through this type of stuff, plain math and good explanation! I liked it so much I made it an article!
 
Gyro you should know this it's drone 101 ;) or did you skip this bit :tongue:

Steve is quite correct with what he is saying - there has to be a servo - tx - rx latency compatibility to get the best and fastest response...

As for travel distance - boffins where i used to work found that this is something that is so so so hard to predict

1: response depends on the correlation of the signal train - and where the rx recognises and picks the signal and reacts (hope that makes sense - NJSwede would probably explain better)
2: response depends on servo reaction - which is not always the same...several tests on the servo's on a prototype drone proved this out...and so many factors played a part in this to put down at present - but most dependant factors being voltage and heat...

This is something that was key to drone advancement and the servo's (actuators) on drones were uber fast and sophisticated....and stupidly priced...but still threw up unpredictable issues - but in the most part operated within the bounds of the specification....issues of noise was the main irritation to the accuracy - which I don't think would be an issue to small servo's but I could be wrong...

excellent posts on this subject and does help show that Spekrum isn't as good as futaba ;) :p
 

Steve_B

70cc twin V2
As for travel distance - boffins where i used to work found that this is something that is so so so hard to predict

1: response depends on the correlation of the signal train - and where the rx recognises and picks the signal and reacts (hope that makes sense - NJSwede would probably explain better)

Space.. I don't think there is anything hard or mysterious in the prediction if we restrict the variables to radio latency alone . If we accept the measurement recorded here: http://rc.runryder.com/helicopter/t172571p1/ (which includes signal processing time etc etc) then distance travelled before initial servo response is simply flying speed x time. Couldn't get much simpler.

In the real world servos, the airframe and human reaction time all make a difference too, of course. But for the sake of this comparison the (reasonable) assumption is that everything except the radio is identical.

As for the tests proving Futaba is better...Well I could have done the comparison of the 'mega bucks' 14MZ vs 'cheap as chips' (relatively) DX8, on latency alone the DX8 wins hands down;) So Spektrum's obviously much better than Futaba:D
 

gyro

GSN Contributor
Steve, if you don't mind, could you repost your data and explanation on latency? I'd still like to make it into an article... Or even better, you can make a wiki for it!
 

wedoitall

Member
yea steves information was quite helpful. kinda helped me make up my mind on what to buy next. gonna get a new radio really soon. cant decide between the dx8, or futaba 8fg or the one coming out next month. if i go with futaba might as well spend another $100 or so to get the one coming out next month. thats just alot of $$$$ for a noob like me.
i know that you guys that have been flying for long time can notice the difference between the dx8 and 8fg but i wonder if a someone that hasnt been flying very long would. and also would a dx8 be sufficent enough to fly 3d. if so is it just barely sufficient.

im looking at my next radio purchase as an investment. ive been flying for around 3 months now and dont see myself quitting any time soon.

sorry if i jack'd the thread
 

Manta

70cc twin V2
Dude a dx8 will do you fine for a long time. I dont like the spectrum handling of the link but in all actuality the dx8 is a feature-rich radio.
Having said that, a hitec Aurora 9 and an xps module and 29 dollar nano rx's be the shizzle. :) way better than dx8 :)
 

Steve_B

70cc twin V2
yea steves information was quite helpful. kinda helped me make up my mind on what to buy next. gonna get a new radio really soon. cant decide between the dx8, or futaba 8fg or the one coming out next month. if i go with futaba might as well spend another $100 or so to get the one coming out next month. thats just alot of $$$$ for a noob like me.
i know that you guys that have been flying for long time can notice the difference between the dx8 and 8fg but i wonder if a someone that hasnt been flying very long would. and also would a dx8 be sufficent enough to fly 3d. if so is it just barely sufficient.

im looking at my next radio purchase as an investment. ive been flying for around 3 months now and dont see myself quitting any time soon.

sorry if i jack'd the thread

As noted above, the DX8 has all the features you would ever want to fly 3D, triple rates, throttle curves, tons of mixes, low latency. I don't think there is any feature that the 8FG has that the DX8 does not (no doubt I'll be corrected on that if I'm wrong). for 3D there is nothing missing at all. Plus the DX8 has telemetry which is very useful as you can monitor the plane's battery voltage and other parameters in flight.

A lot of the buying decision comes down to brand loyalty. There are a number of very capable systems out there and Spektrum DSMX is as solid and secure a link as any that are available (arguably the best in fact). If you are free of pre-conceptions I would suggest you base your decision on bang for buck.

PS.. one feature that Spektrum has that for some inexplicable reason others haven't implemented is Model-Match. What that means is that each receiver is bound to the individual model memory. With Model-match if you try to fly a plane with the wrong memory programmed in your Tx then the receiver wont link, so your model will just sit there.
Imagine with other brands how easy it is with 20, 30 ore more models in the Tx memory to inadvertently set the wrong memory and take off with a plane that has maybe reversed ailerons and elevator:eek:.. with inevitable disastrous results. This is a very common cause of crashes but it's impossible with current Spektrum gear... Reason enough to sway a buying decision IMHO.
 
Steve the point is not the simplicity of speed v time - its more on the consistency of the legacy/latency and the reaction of the servo - not being a bang on figure you can just take to the FAA - when testing it was obvious that there was a band of time to work with so you had a band of variability...it wasn't a huge band...but for example if you have a band of 20ms where 10ms to 30ms is bottom and top observed limits where you record sufficient data you would have a varied distance...based on this time you can then proposed a top and bottom and mean distance based on this...now...when testing we found that at some point you may get a faster or a much slower response than the mean data - so for example you may get the odd 8ms or 44 ms reaction time..

Now as you say it's not totally out of the realms of figuring out - but you have to remember also that FAA and military regulations don't like gestamation so providing the most accurate data and predictability of result and providing the testing criteria to demonstrate your specifications are accurate...BUT PREDICTABLE...I its a military aviation thing...they are very intense in this area to the point of it being anal..
 
Top