rrTV-PHOTO   New HD TV
HOME   rrTV-PHOTO   GALLERIES   MY GALLERY   HELP-FAQ
myHOME PM pmRR MEMBERS 622 ONLINE 26 EVENTS SEARCH REGISTER  START HERE
 
51 pages [ <<    <     5      6     ( 7 )     8      9     NEXT    >> ]80482 viewsPOST REPLY
CanoMod . Futaba-RC . A Main Hobbies

.
.
Radio - Servo - Gyro - Gov - Batt > TX/RX eCCPM Latency Test Results
 
 
Razmo
Key Veteran
Location: Chicago

Yes, would love to see the Futaba 6EX results!!

Raz
03-20-2007 Over year old.
 
 
Greg Faust
Senior Heliman
Location: Northern VA, USA

If you ever have the opportunity to test, I'd love to see how the Spektrum module for the 9303 and the AR9000 receiver stack up against the high dollar Futaba stuff.

<<<Nitromethanol Field Rep>>>
03-20-2007 Over year old.
HOMEPAGE  
 
 
AlexV
Senior Heliman
Location: Moscow, Russia

Still waiting for results of tests of Futaba/JR + Spectrum modules combo...

Alexander
03-20-2007 Over year old.
 
 
CRAZYKEV
Veteran
Location: Cincinnati,Oh - U.S.A.

I am very interested in seeing the 9303+module results.

Someone get JKos a module!
03-21-2007 Over year old.
 
 
JKos
Elite Veteran
Location: City of California in the state of Maryland

I started looking at the Futaba T6EX 2.4 GHz radio last night and, guess what, Futaba implemented a collective slew rate limit just like on the 9C Super and 7C. Very sad. It is quite a slow limit and I can very easily exceed it moving the stick with normal flight motions such as tic-tocs.

This makes latency measurements more difficult but I will prevail.

- John

Protos -- Logo 10
04-14-2007 Over year old.
 
 
AirWolfRC
rrProfessor
Location: 42½ N, 83½ W

Doesn't that also make tic-tocs a bit of a problem ?
04-14-2007 Over year old.
HOMEPAGE  
 
 
Yug
rrProfessor
Location: UK. Herts

It's giving me a problem Cant wait for my DX7 to turn up
04-14-2007 Over year old.
 
 
greg
Key Veteran
Location: Yorkville, IL

Maybe thats why I can't do tic-tocs worth a crap.....its my radios fault. I knew it couldn't have been little ol me.
04-15-2007 Over year old.
HOMEPAGE  
 
 
JKos
Elite Veteran
Location: City of California in the state of Maryland

OK guys, numbers posted (in the very first post) for the Futaba T6EX 2.4 GHz radio. Very fast on the low end but a huge spread to the long end. A fast low end number is nice, but when the latency has such a huge spread, it makes maneuver timing more difficult.

Also, a VERY important note about the T6EX. Futaba implemented a collective slew rate limiter in eCCPM mode just like on the 9C Super and 7C. I had to alter my test procedures to accommodate this "feature." I based the latency numbers on the first sign of change in the receivers output.

What the slew rate limiter does is basically slow down fast servos. Hook it up to a swash and compare collective motion between the T6EX and the DX7 on the same heli and you'll see what I mean.

I observed the same wide spread in latency with the 3-channel ground-based FASST system. On the T6EX, a PPM stream is used to transfer data to the SS deck. The transmitter front end is the fastest stick-to-PPM stream radio I've ever seen.

The transmitter front and SS deck run asynchronously with the transmitter front end running at about 22.5 ms frames and the SS system running at 16 ms frames.



- John

Protos -- Logo 10
04-15-2007 Over year old.
 
 
JKos
Elite Veteran
Location: City of California in the state of Maryland

If I did the math correctly, the slew rate limiter would appear to turn any servo faster than about 0.192 s/60 deg into a 0.192 s/60 deg servo. Here is the math flow...

- From the data, the maximum servo pulse width change rate allowed is 0.5 ms of pulse width change in 160 ms

- Using a Hitec servo specified as 40 deg of rotation per 0.4 ms change in pulse width, that's

- 50 degs of motion due to the 0.5 ms change in pulse width

- 50 degs in 160 ms yields a 60 deg servo rotation time of 192 ms

- Inverting that gives 0.192 s/60 deg.


- John

Protos -- Logo 10
04-15-2007 Over year old.
 
 
GimbalFan
Elite Veteran
Location: Copter County, Nv

Looks like they may have rushed that unit to market a bit too soon, eh? All due respect to Futaba, looks to be a park flyer unit at best.

Thanks again, John. We're eager to see some module latency comparisons too. Waiting patiently . . .

op-thwõp-thwõp-thwõp-thwõp-thwõp-thwõp-thwõp-thwõp-thwõp-thwõp-thwõp-thwõp-t
04-15-2007 Over year old.
 
 
JKos
Elite Veteran
Location: City of California in the state of Maryland

> Looks like they may have rushed that unit to market a bit too soon,
> eh? All due respect to Futaba, looks to be a park flyer unit at
> best.

I don't know what Futaba is up to with their data transfer methods. If you look at their PCM1024 protocol, it could be used better than what we get from any of the radios. And now with the FASST part of this radio running at a higher frame rate than the tx front end and obviously capable of very fast transfer, there is no reason for changes in the PPM stream to be spewed randomly across 3 frames at the receivers output. Maybe they are only sending two channels per 16 ms frame?

- John

Protos -- Logo 10
04-15-2007 Over year old.
 
 
Harris
Senior Heliman
Location: A Greek in Cyprus

How about non eCCPM mode? Is there a slew rate limitation there too? Perhaps it is a matter of the CPU not been fast enough to encode eCCPM in real time.
04-15-2007 Over year old.
HOMEPAGE  
 
 
Yug
rrProfessor
Location: UK. Herts

This issue is one of the things that pushed me towards the DX7. I discovered that with both mCCPM and eCCPM and with very fast servos, my 9C system make the servos move very roughly under certain types of stick inputs. It later transpired that under these input conditions, the 9C/149 was interpolating every other sample and getting it wrong
04-15-2007 Over year old.
 
 
GimbalFan
Elite Veteran
Location: Copter County, Nv

EDIT -- pic removed. John has updated his data at the start of this thread.

op-thwõp-thwõp-thwõp-thwõp-thwõp-thwõp-thwõp-thwõp-thwõp-thwõp-thwõp-thwõp-t
04-15-2007 Over year old.
 
 
TMoore
rrProfessor
Location: Cookeville, TN

Wow, all these shiny new radios and my 11 year old Stylus is still Champeen!

He, he, he, and they said it couldn't be done.

TM

"If you can do it, it ain’t bragging." - Will Rogers
04-15-2007 Over year old.
HOMEPAGE  
 
 
AirWolfRC
rrProfessor
Location: 42½ N, 83½ W

Quote 
f I did the math correctly, the slew rate limiter would appear to turn any servo faster than about 0.192 s/60 deg into a 0.192 s/60 deg servo. Here is the math flow...

- From the data, the maximum servo pulse width change rate allowed is 0.5 ms of pulse width change in 160 ms

- Using a Hitec servo specified as 40 deg of rotation per 0.4 ms change in pulse width, that's

- 50 degs of motion due to the 0.5 ms change in pulse width

- 50 degs in 160 ms yields a 60 deg servo rotation time of 192 ms

- Inverting that gives 0.192 s/60 deg.
Let me try a slightly different approach.
You state that Futaba limits the slew rate to 0.5ms signal / 160ms
Most all servos give 90º for a 1.0ms to 2.0 ms input signal change even though their speed is spec'ed for only 60º
So a 0.5ms input signal change is good for 45º position change which is limited (by Futaba) to a travel time of 160ms. To get 60º, multiply that 160ms by 1-1/3 = 213.33ms for 60º travel for a number that can be compared (apples to apples) to other servos. (plus whatever lag time there is in the particular servo in use)

The question comes up, why bother with faster servos with Futaba ?

I have not verified your measurements but using your numbers, the above applies. This slew rate limiting does not apply to all Futaba systems in all modes.
04-15-2007 Over year old.
HOMEPAGE  
 
 
AirWolfRC
rrProfessor
Location: 42½ N, 83½ W

It would be interesting to include position resolution in your chart also.
04-15-2007 Over year old.
HOMEPAGE  
 
 
JKos
Elite Veteran
Location: City of California in the state of Maryland

> How about non eCCPM mode? Is there a slew rate limitation there too?

I just tested it and no, there is no slew rate limiter on collective in normal swash mode. Added latency data for the collective channel in normal swash mode.

> It would be interesting to include position resolution in your
> chart also.

Unfortunately, it is much too late for me to collect that data for most of radios. We know the answers for all the PCM, DSM, and DSM2 systems.

- John

Protos -- Logo 10
04-15-2007 Over year old.
 
 
AirWolfRC
rrProfessor
Location: 42½ N, 83½ W

In particular, I was thinking of the DX6 system which comes in at 256, even though the DSM RF deck can handle more.
04-15-2007 Over year old.
HOMEPAGE  
 
 
51 pages [ <<    <     5      6     ( 7 )     8      9     NEXT    >> ]80482 viewsPOST REPLY
Boca Bearings . Modefo's RC Helicopters . XHELI.COM

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

Subscribe to This Topic

Thursday, October 16 - 5:06 am - Copyright © 2000 - 2008 runryder.com | email | link to rr | runryder needs cookie