rrTV-PHOTO   New HD TV
HOME   rrTV-PHOTO   GALLERIES   MY GALLERY   HELP-FAQ
myHOME PM pmRR MEMBERS 544 ONLINE 45 EVENTS SEARCH REGISTER  START HERE
 
48 pages [ <<    <     41      42     ( 43 )     44      45     NEXT    >> ]74522 viewsPOST REPLY
CarbonXtreme . Midland Helicopters . HeliProz

.
.
Radio - Servo - Gyro - Gov - Batt > TX/RX eCCPM Latency Test Results
 
 
rexxigpilot
Senior Heliman
Location: florida

This thread has exhausted the issue of system latency. We all agree that faster is better, but is that the top criterion to use in judging one system over another? The answer is NO!!!

It's apparent that order of servo input/output from the receiver/transmitter combination is important in eCCPM helis to avoid a dancing swash plate and resulting precision loss. However, not many complained about the lowly, unworthy, slow Futaba 9C until we saw the JKos data on its latency. Why is this? It's because the limiting factor on our radio systems is the Human Interface Device, HID (the human brain, eye, nervous sytem and musculature, etc.).

We must put latency into perspective. There is such a thing in medical fields known as a P300 wave/response. This is the time it takes for humans to recognize/perceive something from first observation of a change in environmental stimulus. Put another way, it is the time it takes your brain to realize that you heli is going down instead of up when you lose orientation doing 3D.

The actual P300 time varies from about 300 milliseconds (hence the name) to 800ms. This time does not even count reaction time after recongnition. Total time to react can be twice the P300 time, or up to 1.6 seconds! Most of us in this hobby are probably closer to 0.7 seconds total to react.

Anyway, what all this tells us is that we do fly mentally far out in front of our helis. Combine this with the other factors that affect radio system transmission and reception such as signal strength, interference, error correction, etc. and we start to see the limited impact that latency has on our flying abilities. But we can discuss these other factors after another 43 pages of chatter about latency.

No offense to JKos. I enjoy reading the results of his testing just as much as the next RR member. Let's all just keep latency in perspective when discussing which radio system is superior. There are so many more important factors to consider than just latency. Sorry if I bored anyone.
04-03-2008 06:26 PM
 
 
JKos
Elite Veteran
Location: Town of California in the state of Maryland

rexxigpilot,
> Total time to react can be twice the P300 time, or up to 1.6
> seconds!

How do you explain test results which show more like 200 ms? There are computer based tests which have come up in threads here on RR and guys are routinely getting in the low 200 ms range.

Also, keep in mind that the radio is part of a continuous control loop not just simply receiving an input from the human reaction to an event (like clicking the mouse in the tests I'm referring). Any amount of delay (i.e. latency) which can be cut from that control loop is a good thing. Consistency in the latency is also a factor. There are differences between brands in that department as well.

> but is that the top criterion to use in judging one system over
> another? The answer is NO!!!

Isn't it each pilot's prerogative to choose what is important in their selection of a radio? Besides, the percent of people in the RC radio market who even know about my data is miniscule and even fewer of those who do know about it use it to choose a radio.

Razmo,
Yes, the 12X is a few milliseconds slower than the DX7. Even I would argue those few milliseconds are of no consequence. This actually plays right into rexxigpilot's coments. The differences in programming capability, 1024 vs 2048, radio feel, etc., are probably more important than a few milliseconds.

- John

Protos -- Logo 14 -- Logo 10
04-03-2008 06:39 PM
 
 
hootowl
Elite Veteran
Location: Garnet Valley, Pa.

There is the threshold factor.

When you can see interaction and slow response in plain view, on the bench with a 9C, and then observe the accuracy and speed of the newer radios, again, on the bench, you appreciate the large improvements they have made in the technologies.

Now we come to the point where "does it matter" how much faster it is. At some point going from -14 to +14 in .00001 (exaggerated) milliseconds may actually be a bad thing from an aerodynamics and mechanical point of view.

I have to agree that we are coming to this point, in my opinion, where lower latency becomes undetectable and/or does nothing for your flying capabilities.

Now we look at the equipment itself... Is it well made? Is it ergonomic? does it have all of the features you need and/or expect for the price? Is the transmission technology robust and reliable? Is it priced reasonable? Is it backed by a reputable company with a good service history?

For pure versitility, available channels, ergonomics and all of the above, I chose the 12Z with the FASST system. It has every capability available to man with the most important being it's ability to handle helicopter swashes.... all of them and the ability to fine tune them.

It's fast, precise, has a great screen and menu system and is about as versatile as you can get. I don't even think about latency anymore. John has given me all the data I needed and I thank him for his effort.

"Any man who says he is an American, but something else also, isn't an American at all.
04-03-2008 07:13 PM
HOMEPAGE  
 
 
seattle_helo
Key Veteran
Location: Washington State, USA

John, I would agree that relatively few folks know about your data presented here- but I must say that I'm grateful for all your time and effort spent obtaining the info and presenting it. It certainly has had an effect on my purchase decisions and reinforces how nicely the Spektrum gear performs.

nick
04-03-2008 07:35 PM
HOMEPAGE  
 
 
Razmo
Key Veteran
Location: Chicago

Same here John, glad to have you!

Raz
04-03-2008 07:36 PM
 
 
human213
Senior Heliman
Location: malibu

John

Kudos from here as well.



This matter is analogous to those who cannot hear the difference
between LP and CD, or mp3 and 24 bit 192khz or higher audio.


I prefer better always, and as Einstein said, 'I am very easy to satisfy;

I am always satisfied with the very best"...


Thanks again for all your work, and even the hacks benefit, however
unknowingly....





michael

Not everything that can be counted counts, and not everything that counts can be counted.
04-03-2008 07:47 PM
 
 
rexxigpilot
Senior Heliman
Location: florida

I'm afraid my last post was totally misunderstood. I didn't suggest that John stop posting the data. As I mentioned quite clearly, I, like many, enjoy the information John so kindly posts regarding latency. I merely pointed out that the issue of latency has been discussed enough.

My intension was to encourage John and others on this thread to let this discussion evolve into more than it is now. Let's discuss other factors that have a profound affect on our radio systems so we can all enjoy a fulfilling dialog and learn what is truly important in a radio controlled system.

This is an important time in the world of RC as we move swiftly from the old fixed frequency systems to the new 2.4 GHz shared frequency systems. I wish I had the time to do much of the testing needed to allow our complete understanding of these new systems, but sadly I don't. John does such a great job with the latency issue, I believe he above all other RR members is probably best suited and capable of providing this much needed information.

As human213 came near to understanding, this issue is much like a discussion of high fidelity music and home theater systems/formats. As my pictures below reveal, I am a complete AV junky as much if not more than with RC helis. I do understand the benefits and drawbacks of the various features/formats in both AV and RC that have come and sometimes sadly gone. In my audio rack you will notice LP, CD, SCAD, DVD-Audio, VHS, Laser Disk, DVD and Blu-ray to name a few. My Martin Logan Prodigy's reveal every detail of music. I don't use MP3 because I can certainly tell it is low quality sound. It just doesn't cut it for me.

But tell me please how a garbage format like MP3 survived while SACD and DVD-Audio went away? Value is often found in more than just the superior specifications (read low latency). The vastly superior audio formats of DVD-Audio and SACD died out due to a lack of marketing, a lack of appreciation for quality by the general public and a cultural phenomenon for MP3's (aka fad).

I'm just glad at least that the format war between HD-DVD and Blu-ray has ended. Kudos to Toshiba for that and kudos to John for his work. Let's just bring this discussion to a higher level so we don't go back to 72 MHz. After all the best 72 system has less latency than every 2.4 GHz system tested so far.

04-04-2008 04:16 AM
 
 
GimbalFan
Elite Veteran
Location: Copter County, Nv

These 1 or 5 or even 10 ms differences are arguably insignificant once we get below 20 or 30 ms total latency, but those 9C numbers showed one of the most obvious symptoms of high latencies:

With my old 7C I could easily move the cyclic stick fast enough that the 9252 servos on my R50's swashplate would get so far behind the stick that their total travel dropped to less than a ¼ of normal. The tx couldn't keep up with what the stick was telling the servos to do because location commands were being processed too slowly by the tx for the servo's locations to be sequential. ATX 94758's (which are about twice as fast) appeared only slightly better.

But 94758's being pushed around by a Stylus in PCM are another story.

60 ms is a significant difference in average stick-to-servo speed, especially on a heli, even if you're only mildly-skilled like me. Even though I rarely if ever need to move the servos that quickly, given a save situation it's damn fine to know I've got the extra 5' of air (at 50 mph) which that speed gives me.

op-thwõp-thwõp-thwõp-thwõp-thwõp-thwõp-thwõp-thwõp-thwõp-thwõp-thwõp-thwõp-t
04-04-2008 06:01 AM
 
 
Four Stroker
Veteran
Location: Atlanta

I am all for keeping this thread what it originally was: Testing CCPM stick to servo latency. If you want to discuss anything else, start a new thread.

Gimbal Fan, you are describing slow servos more than latency. The two are related.
04-04-2008 04:51 PM
 
 
GimbalFan
Elite Veteran
Location: Copter County, Nv

Quote 
you are describing slow servos more than latency.
How then do you explain why this occurred to a much lesser degree with the same servos (both 94758's and 9252's) -- but using a 60 ms faster tx/rx?

The servos weren't slow to duplicate the stick movements -- their movements actually became attenuated if you will -- they traveled less and less the faster I moved the stick, while moving the stick the same amount and in the same pattern. 'Attenuated' is the best word I can think of to describe it -- but I'm nowhere near as knowledgeable about the hows and whys as the EE guys.

op-thwõp-thwõp-thwõp-thwõp-thwõp-thwõp-thwõp-thwõp-thwõp-thwõp-thwõp-thwõp-t
04-05-2008 04:47 AM
 
 
Four Stroker
Veteran
Location: Atlanta

These two things go together and add up. They are just hard to separate. Sometimes they are of the same magnitude. With 9202 servos on a 7C - radio much faster than servo. 9C with 9254 servos - servos much faster than radio. 7C is fast, 9252 moderately slow at 140 ms per 60 degrees at best. If you could move the stick up and down really fast you could get the same result with a 0 ms latency radio and a typical servo.
04-05-2008 05:32 PM
 
 
human213
Senior Heliman
Location: malibu

Amen four stroker

Like the pseudo religous claiming there are too many scantily clad
women in america, in public places...



1)Where are all these scantily clad women?

2)There will never be enough scantily clad women for my tastes...



thus, I do not think latency or 2.4 problems, limitations, and advances
can be discussed too much.





ps rexxi...you need to hear my new reference quads...
and, John, thanks again for all your work... and I may have to send you my fx40 when the new 2.4 mod
comes out for it...

Not everything that can be counted counts, and not everything that counts can be counted.
04-05-2008 05:46 PM
 
 
GimbalFan
Elite Veteran
Location: Copter County, Nv

Something about this 'attenuation of movement' effect isn't adding up.

If the only thing changed in the comparison is a 60 ms improvement in the tx's stick-to-servo speed, and the effect is observable using two different servos, this must indicate that what's causing the effect is somewhere between the stick and the servo.

op-thwõp-thwõp-thwõp-thwõp-thwõp-thwõp-thwõp-thwõp-thwõp-thwõp-thwõp-thwõp-t
04-05-2008 08:25 PM
 
 
JKos
Elite Veteran
Location: Town of California in the state of Maryland

New XPS numbers posted based on tx module firmware 2.3 and rx firmware 2.4.

GimbalFan,
Was the 7C in CCPM mode when you observed the stated behaviour?

- John

Protos -- Logo 14 -- Logo 10
04-09-2008 01:41 AM
 
 
GimbalFan
Elite Veteran
Location: Copter County, Nv

No, and the two servos were connected to the aileron and elevator channels.

op-thwõp-thwõp-thwõp-thwõp-thwõp-thwõp-thwõp-thwõp-thwõp-thwõp-thwõp-thwõp-t
04-09-2008 01:45 AM
 
 
JKos
Elite Veteran
Location: Town of California in the state of Maryland

Please note that there is currently a problem with the 2.4 GHZ sorted by range graph. I didn't realize Excel forced the bottom of the red line to match the blue tick if the blue tick value was less then the minimum red line value. The radios are in the correct order and the blue ticks (the range) are properly placed, but some of the red lines are longer than they should be. I'll get it fixed tonight.

- John

Protos -- Logo 14 -- Logo 10
04-09-2008 03:51 PM
 
 
XPS
Heliman
Location: Lake Havasu City, AZ USA

Thanks for posting the new numbers.

I did figure out a way to decrease the time across the board by not outputting all of the channels. With your JR9303, you are always getting 9 channels of data sent. Since most helis use only 6 channels, sending/receiving the extra 3 channels worth of data requires between 5.5ms and 7.9ms, depending on the channel position. This time savings would put us in the 28ms area for your MINLF. This would be configured on the transmitter side as a "max channels" setting in the programming menu.

Jim Drew, CEO/President - Xtreme Power Systems
04-12-2008 06:45 PM
HOMEPAGE  
 
 
RCHelicopterGuy
Veteran
Location: Michigan

Quote 
Since most helis use only 6 channels

I don't know what helicopters you fly, Jim, but if you use a governor and/or a gyro with remote gain, that's a minimum of 7 channels. With my 601 and GV-1 (both with remote adjustability), that's 8 channels.

Granted, it's hair splitting, but it seemed a good time to point it out. I don't know of anyone outside a newcomer who only uses 6 channels.
04-12-2008 06:56 PM
 
 
XPS
Heliman
Location: Lake Havasu City, AZ USA

Thanks for the feedback. My electric helis (TREX 600, TREX 450, EP100) only using 6 channels - Throttle, 3 swash controls, rudder, and gyro gain. The max channels would default to whatever the transmitter module used for the maximum number and could be reduced. Even saving 1 channel (in your case) would save about 2ms in latency minimum.

Jim Drew, CEO/President - Xtreme Power Systems
04-12-2008 08:14 PM
HOMEPAGE  
 
 
AlexV
Senior Heliman
Location: Moscow, Russia

To John
As I understand latency results of Futaba TM14+R6014 is much better then TM14+R617. Here what I read in futabas FAQ:
Q - Is there going to be a latency issue on eCCPM helicopters if you use an R607FS or 617FS receiver with my TM-14 module.

A - The R607FS is not capable of receiving its information in "packets" as is a 2048 receiver such as the R608FS or R6014FS. Therefore information is sent as it is in any other 72 MHz or 50 MHz receiver. "Channel grouping" only applies to the R608FS and R6014FS receivers

May be it would be interesting to test eccpm latency of TM14+R608...
Sounds like it can show the same results as R6014.

Alexander
04-15-2008 03:44 PM
 
 
48 pages [ <<    <     41      42     ( 43 )     44      45     NEXT    >> ]74522 viewsPOST REPLY
ZoomsHobbies . HeliHobby . Ron’s HeliProz South

.
.
Radio - Servo - Gyro - Gov - Batt > TX/RX eCCPM Latency Test Results
  UPDATE SCREEN   PRINT TOPIC Advertisers 

Subscribe to This Topic

Saturday, August 30 - 6:00 pm - Copyright © 2000 - 2008 runryder.com | email | link to rr | runryder needs cookie