Testing/Review of the Spektrum module for the 14MZ

[JKos]

Finless,
> tell me why the 9000 has 2048 res since the PPM signal out the buddy box port is
> really NOT going to change?

The PPM stream out of the buddy box port is at least 2048 resolution, maybe higher. When the DM8 connects with an AR9000, it sends the pulse widths of the PPM stream in 2048 resolution. When it talks to an AR7000, the AR7000 can only accept 1024 resolution. So, the transmitter isn't changing what it does, it's just a difference in how the connected system can use the provided data.

> it is well documented having ail/elv/pitch on 4,5,6 is best.

This is due to how G3 works. Half the channels have simultaneous pulses, then the other half of the channels have simultaneous pulses. The output pulse groups are seperated by half the frame period. Keeping the three channels used for the swash in the same group is what is key, actually. I found nothing particularly special about 4, 5, and 6 when testing G3. As long as all three are kept in the same group it really doesn't matter much which of the channels in that group are used.

> Your saying the latency is in the RX remapping and not coming from the PPM train is
> part of the problem?

There is a more to the story than is immediately obvious. It has to do with a certain characteristic of Futaba PPM streams.

> Something is very different on how PCM 2048 works and how the PPM train is put out the trainer port.

Bingo. It appears PPM operation was not a high priority during 14MZ development. For some reason the latency has a spread of about 35 ms. That is just "not right" as the spread shoud be one frame period (as it is on every other radio I've tested in PPM mode). I didn't believe it at first, but that's what it is.

Don't bother borrowing the system again just as MrMel said. It still goes through the wobbles at times but it just makes it good when it is not in that state.

- John

[Finless]

Thanks for explaining John. Makes sense.

Yes I understood about the 14mz and "groupings" 456 was default so why change it.....

Couple of questions.

1) Will the DX7 be 2048 with a AR9000? I did feel a loss when I went grom G3 to DX7 with AR7000. If yes I mat have to go get a AR9000

2) So your aying if I set up for channels 1,4, 6 I wont feel the paddles STEP similar to a 9C? I understand the wobble but I wont feel a step? Or do I need to try the AR9000 to get rid of that too?

3) I have "heard" that to get the full 2048 with FASST I will need the 14 channel FASST RX? Is that confirmed or do you know yet?

Bob

[DominicD]

Originally Posted by JKos

Finless,
> tell me why the 9000 has 2048 res since the PPM signal out the buddy box port is
> really NOT going to change?

The PPM stream out of the buddy box port is at least 2048 resolution, maybe higher. When the DM8 connects with an AR9000, it sends the pulse widths of the PPM stream in 2048 resolution. When it talks to an AR7000, the AR7000 can only accept 1024 resolution. So, the transmitter isn't changing what it does, it's just a difference in how the connected system can use the provided data.
d.

- John

The PPM signal from the buddy box is just a time coded system. The module can be triggered at any clock rate it wants to to obtain any resolution it wants. Of course you cant generate NEW data but if the clock rate is faster then the PPM resolution then the LSB will just be garbage from jitter.

[JKos]

HFG,
True enough. What I'm saying is that the buddy box output (PPM output) of the radio actually is at least 2048 equivalent resolution.

Finless,
> Will the DX7 be 2048 with a AR9000?

No.

> So your aying if I set up for channels 1,4, 6 I wont feel the paddles STEP similar to a 9C?

It seemed good to me during the good times but the wobble stage is still unacceptable.

> I have "heard" that to get the full 2048 with FASST I will need the 14 channel FASST RX?
> Is that confirmed or do you know yet?

I don't have any info on that. Haven't they been advertising 16,384 resolution for the existing FASST system?

- John

[Funky]

Great vid Bob. I would love to see the same tests on a 9303 w/ Spectrum module. I have been considering buying the module to convert my 9303. Does anybody you fly with run the 9303/spectrum? Is it only the 14MZ or is it all Futabas that have this problem when running a Spectrum module?

[Finless]

I don't have any info on that. Haven't they been advertising 16,384 resolution for the existing FASST system?

I havent been paying attention but I "heard" that to get the full speed out of the 14MZ that you must use the 14 channel FASST RX. I hope that is not true....

Funky,
I dont know but based on JKos numbers for the 9303 and Spektrum, the latency is pretty high (54 as I recall). That is puzzling since the PPM latency on a 9303 is pretty good at 24. This makes me assume the Spektrum module is inducing double the latency!

John is WAY better than I at explaining this stuff though so hopefully he will respond to your question and I dont know if it has the same "sync" issue I saw ont he 14mz with module since I dont have a 9303 and module to test with.

Bob

[JKos]

Funky,
The 9303 with a Spektrum module does great for eCCPM.

The latency of the 9303/Spektrum combo is similar to the 10X in PCM mode number wise. But the better data transfer behaviour should make the 9303/Spektrum combo feel better than a similar measuring PCM setup.

Finless,
> I havent been paying attention

Me neither. I figure it will be whatever it is when it is finally released.

- John

[Funky]

Thanks Bob and Jkos

John, are your tests posted somewhere?

[JKos]

Funky,
http://www.runryder.com/helicopter/t172571p1/
http://www.runryder.com/helicopter/t171302p1/

- John

[DominicD]

How about the 9303 with the XPS module ?

[MrMel]

Originally Posted by HFG

How about the 9303 with the XPS module ?

Head over to RR at the link posted and you will see that combo too.

[JKos]

HFG,
It's in the graph and table.

- John

[Finless]

John, if you would like to post your numbers here as well I can create a sticky post in this forum you can edit. If so let me know the title of the topic you would like.

Bob

[DominicD]

Oops, reading is overrated. Having only ever flown my stock 9303 i wonder how critical latency really is ?

[MrMel]

Thing is, if you only flew it "bad" you will be Ok, but as soon as you try it "good" there is no going back, ever... its that much of a difference.
Then again, a badly setup heli wont fly good either, ive seen a good portion of those too.

[Trooper Sam]

Bob, you took a quick look at the 14MZ/XPS combo before its release when you visited the AMA Convention this year. As I recall, it didn't exhibit behavior like this during your brief examination.

Do you think that perhaps they dialed it out, or could it be a better performer in this arena?

[Finless]

Sam, until I test it myself I cannot really say yes or no! You know me Sam and I tell it like it is but until I have one (if I ever will) I cant say how XPS works or not. Sorry I just cant afford to show you guys everything I would like to show. Sorry I cant answer the question

Bob

[DominicD]

Ok folks thats it, time to kick in some money so we can hire finless as a fulltime tester.

[R-4-L]

I wonder if this whole thing also applies to the 12Z :?

[Flyingagain]

So since this apears to be a performance issue will spectrum be refunding/replacing this unit with one that should live up to par?

I am sitting here with a unit I purchased at IRCHA for $250(figured it wasn't a bad price to play with), but havent installed yet and debating whether to leave it in the package or test it for myself, but the more I read, the more it seems to be a problem.

Thoughts?