RunRyder - TX/RX eCCPM Output Test Results

rrTV-PHOTO

New HD TV

HOME rrTV-PHOTO GALLERIES MY GALLERY HELP-FAQ

myHOME PM pmRR MEMBERS 661 ONLINE 42 EVENTS SEARCH REGISTER START HERE

6 pages [ << < 1 ( 2 ) 3 4 NEXT >> ]

15983 views

POST REPLY

JR-Spektrum	.	Gyro Hobbies	.	E-flite
	.		.

Radio - Servo - Gyro - Gov - Batt > TX/RX eCCPM Output Test Results

JKos
Elite Veteran
Location: City of California in the state of Maryland

7) Futaba 9C Super with an R149 receiver
Of note is that Channel 6 leads Channels 1 and 2 by two full frames. Clearly the slow transition is software driven. Doing the latency tests will show how quickly it can actually transition.

Futaba 7C with an R149 receiver
Unlike the 9CS, the channels follow each other closely. Clearly the slow transition is software driven. Doing the latency tests will show how quickly it can actually transition.

- John

07-04-2005 Over year old.

JKos
Elite Veteran
Location: City of California in the state of Maryland

1) JR 9303 with 649S PCM receiver

- John

07-04-2005 Over year old.

JKos
Elite Veteran
Location: City of California in the state of Maryland

Shawn,
Yeah, the 9303 seems to be just fine. Funny thing is, it is the 9303 which started this whole investigation for me. The 9303 also came out decently well in latency as well.

The only thing I can think of is that my 9303 had the 140 deg eCCPM problem while the one I used for these last test does not. I.E., the one used for the test shown above has the latest firmware. I don't why the 140 deg eCCPM problem would have changed what I'm looking at here, but maybe that isn't the only thing they changed.

- John

07-04-2005 Over year old.

JKos
Elite Veteran
Location: City of California in the state of Maryland

Shawn,
> Why is it that Channel 3 has the strange jump, compared to the other
> two channels?

It most likely has to do with JR's PCM encoding scheme.

> What does this translate to in real-world terms...

Since the second frames follows in 20 milliseconds, it probably doesn't mean much at all since the servo will not have arrived at the "first commanded position" of the first frame before it receives the second and final position. If this behaviour lasted more than one frame, it might be an issue, but it never did in any of the collects.

- John

07-05-2005 Over year old.

Britflyer
Veteran
Location: Derby, England

Thankyou

Thankyou John for all your work on this issue, it is much appreciated by me and I am sure many others.

I have a 9X Ver2 on order, and it will be used on a CCPM model. It is reassuring to know that your earlier findings with the 9303 have been disproven.

Paul

-------------------------------------------------------

Climb-out.com
CSM
Hyperformance RC
Flightpower

07-05-2005 Over year old.

HOMEPAGE

AirWolfRC
Elite Veteran
Location: 42½ N, 83½ W

If you really want a solid answer, plot the TX pot positions against the coresponding servo output positions and see the latency. This would take everything except servo load into account.

If you don't get a 45º straight line on the scope for each channel, you have latency. The deviation from a straight line will tell you how much latency.

Wolfgang

07-05-2005 Over year old.

HOMEPAGE

JKos
Elite Veteran
Location: City of California in the state of Maryland

Paul,
Thanks.

AirWolfRC,
Are you referring to all the latency tests or specifically for the 9C Super and 7C? The latency I have measured and reported in the other thread for the latency is valid for what is listed there. If I can get resolution from Futaba with respect to what the 9C Super and 7C are doing, I will retest the radios.

I have purposely left off any additional latency from the servos themselves. There are way too many servos to delve into such tests. And then, as you noted, the results will change based on the load on the servos. The tests I am performing are purposely transmitter/receiver effects only.

Tonight I will post a plot of the 9C Super latency test response with the stick input shown and positioned correctly in time. I may be able to deduce a latency time range from the data.

- John

07-05-2005 Over year old.

AirWolfRC
Elite Veteran
Location: 42½ N, 83½ W

There are definitely a lot of different factors involved.

Let's say that there is a max of 5ms slew between channels. That means a typical digital servo will have traveled about 6º by the time the next servo starts to move. 6º servo travel in 5ms will have very little total effect on air frame position from desired position in the real world.

End to end measurements are needed to tell what is really going on.

Wolfgang

07-05-2005 Over year old.

HOMEPAGE

JKos
Elite Veteran
Location: City of California in the state of Maryland

Wolfgang,
Between the graphs shown in this thread and the latency figures given in the other thread, what you are saying has been presented.

The graphs in this thread which show the channel output structures give you the "slew between channels" as well as how the system is actually changing those channels. For example, note how all three channels have output pulses at the same time for the 14MZ, G3, and same group channels (4, 5, and 6 for example) and all three change at the same time. Compare that to how the 9C with an R149 behaves and you can clearly tell there will be a difference.

"End to end measurements" from stick to receiver output have been collected and presented.

I've thought of coding up a "virtual swashplate" which would graphically display in an animated way what the data I've collected would suggest that the swashplate will do. Now, it would be a rather simplified model behind the animation, but I can do it.

What you are really asking for is what I originally suggested in another thread and that was to take video of a swashplate moving with various radio equipment combinations. I decided that was way too time intensive for the amount of time I have to do these things.

- John

07-05-2005 Over year old.

JKos
Elite Veteran
Location: City of California in the state of Maryland

> 6º servo travel in 5ms will have very little total effect on air frame
> position from desired position in the real world.

Exactly that type of effect is what got me looking into this. My original troubles with channel 6 appearing to lag on the 9303 I had was causing the heli to deviate from my intended course during fast collective type maneuvers. Through swapping of channels and servos I was able to determine that the issue always followed channel 6. This was easily observed on the bench as well and by several other sets of eyes. (All of this is in another thread.)

It would be very interesting indeed to be able to go from the stick all the way to "real effect on the heli." The number of variables to control is extremely large, however.

In the end, what my tests seem to show is that there are clear differences in the operation of various radios.

- John

07-05-2005 Over year old.

JKos
Elite Veteran
Location: City of California in the state of Maryland

I would like to thank Stan Johnson of Hobbies 'N Stuff (home of Callie Graphics) for allowing me to borrow the Futaba 9C Super and the 7C for testing purposes.

I would also like to thank John Pompa for allowing me to use his JR 9303 and Tom Medland for allowing me to use his Futaba 9C.

- John

07-05-2005 Over year old.

AirWolfRC
Elite Veteran
Location: 42½ N, 83½ W

Those test results are strange indeed considering that when I called the tech's at Horizon about the 9303, they claimed that the pulses are a straight sequential and there are no half frames or any other such thing going on with the 9303.

Wolfgang

07-05-2005 Over year old.

HOMEPAGE

JKos
Elite Veteran
Location: City of California in the state of Maryland

Wolfgang,
I should add the graph showing the pulse alignment for the 9303. The outputs on the JR PCM receiver are sequential right down the line from 1 to 9. So, they were correct about that.

> and there are no half frames

Are you referring to differential encoding or something to that effect? Yes, it certainly appears that there is something occurring besides straight up absolute positions being transmitted every frame. If you look at the graphs above, it appears that channels 2 and 6 have absolute position updates in the same frame and channel 3 has a delta update. Then the opposite for the next frame. Just a guess based on the observed data.

- John

07-05-2005 Over year old.

AirWolfRC
Elite Veteran
Location: 42½ N, 83½ W

I guess I'm not seeing what you refer to. I am not seeing each channel in it's own time space within each frame. I see all channels occupying the entire frame with the amplitude as it's width.

With only 9 channels, it shouldn't be necessary to time share frame time space with the less significent channels. At up to 2ms per channel times 9 chalnnels = 18ms. 50cps frame rate is still 20ms leaving a minimum of 2ms reset time in pwm mode. I wonder if JR streched the 20ms frame window a bit for the 9303?

Wolfgang

07-06-2005 Over year old.

HOMEPAGE

JKos
Elite Veteran
Location: City of California in the state of Maryland

Wolfgang,
This is the type of graph I'm speaking of for showing the channel output scheme. This one is for Airtronics but there are others posted throughout this thread.

I'll post this type of graph for the 9303 but it isn't very exciting... The channel 2 pulse is followed by the channel 3 pulse. Then channel 4 followed by channel 5. Finally, the channel 6 pulse occurs.

- John

EDIT: New 9303 plot posted in 9303 post above.

07-06-2005 Over year old.

seattle_helo
Key Veteran
Location: Washington State, USA

John, thank you for all of the definitive and useful data you've posted here. This kind of stuff is the real deal and it's awesome. Thanks again.

07-06-2005 Over year old.

HOMEPAGE

JKos
Elite Veteran
Location: City of California in the state of Maryland

Thanks seattle_helo!

Wolfgang,
I have a new type of plot which is more of what you were originally asking about. It combines the two types of plots I've been posting above. The x-axis is time and the y-axis is pulse width. So, they show both the respective channel timing as well as the change behavior. They do indicate a slightly different result for some of the data collects versus the frame-based plots. I will post them tonight.

I also have the beginnings of a virtual swash routine. I was thinking the output could be a plot showing the deviation in elevator and aileron as a function of time. I have to make several simplifying assumptions, but the results might be interesting none-the-less.

- John

07-06-2005 Over year old.

AirWolfRC
Elite Veteran
Location: 42½ N, 83½ W

Quote
The x-axis is time and the y-axis is pulse width

That's where I don't follow you. Pulse width IS time. You're plotting time against time ?

The nature of the system will never allow channels to update simultaneoulsy since data is sent sequentially.

If you want to get rid of slew, you need a system that allows parallel in (control sticks) and trot the data out of a shift register, transmit it sequentially and have the RX trot the data into a shift register for parallel out to the servos. (this may give electrical supply problems when all servos kick at the same time) Then you only have the latency of the shift register out and in to contend with and your control inputs/servo outputs will stil be simultaneous.

The above does not necessarily apply to pcm. Each manufacturer does their own (secret) thing here.

Wolfgang

07-06-2005 Over year old.

HOMEPAGE

AirWolfRC
Elite Veteran
Location: 42½ N, 83½ W

John, I see some strange pulse coincidences on the three trace plots where channels 1 and 2 have coincident pulses and channel 6 is down the line somewhere. My question is why are 1 and 2 coincident?

The other graphs you show with pulse width versus frames are showing exactly what? Is that data the outputs from the RX? If so, where are the stick inputs?

Wolfgang

07-07-2005 Over year old.

HOMEPAGE

AirWolfRC
Elite Veteran
Location: 42½ N, 83½ W

Another question, which of these test are done using e-CCPM and which are with straight pitch, roll, collective on their own channel? I would expect the dedicated channel operation has the least problems of latency and slew.

Wolfgang

07-07-2005 Over year old.

HOMEPAGE

6 pages [ << < 1 ( 2 ) 3 4 NEXT >> ]

15983 views

POST REPLY

Next D	.	Fast Lad Performance	.	Ace Hobby
	.		.

Radio - Servo - Gyro - Gov - Batt > TX/RX eCCPM Output Test Results

UPDATE SCREEN PRINT TOPIC

Advertisers

Subscribe to This Topic