What is 2048 resolution

[Mslater]

Ok I know what it is, but what exactly do you need to get the benefit ?

1 A TX that transmits 2048
2 An RX that understands the 2048 TX
3 Servos ?

Presumably a 2048TX and 2048RX is no good if you still got DS610's

Or is it ??

[coptermanic]

Good question!!!!!!

Where are the experts when we need one????!!!!

[GravityKills]

I measured a stick travel of 45 mm per channel on the DX7.
To fully appreciate it, you need fingers that are accurate to about 20 micrometers.

Just kiddin'. Although, a human hair is about 50 micrometers, hmmm...

A digital servo will probably use a smaller number of bits in its A/D converter. In this case, the increased resolution is truly lost.
An analog servo would in average become more accurate, even if the deadband is much wider. Well, under ideal circumstances.

The question is actually quite interesting, it's somewhat the same problem as to why we need dithering when mastering CDs. Geek stuff.
But to improve my flying, about as relevant as the gravitational pull of the moon

[oneandonlyian]

I measured a stick travel of 45 mm per channel on the DX7.To fully appreciate it, you need fingers that are accurate to about 20 micrometers.Just kiddin'. Although, a human hair is about 50 micrometers, hmmm...
A digital servo will probably use a smaller number of bits in its A/D converter. In this case, the increased resolution is truly lost.
An analog servo would in average become more accurate, even if the deadband is much wider. Well, under ideal circumstances.
The question is actually quite interesting, it's somewhat the same problem as to why we need dithering when mastering CDs. Geek stuff.
But to improve my flying, about as relevant as the gravitational pull of the moon

ERM Im even more confused lol.

[Mike W]

laymens terms .. .the speed/feel from stick input to servo movement ... in most eccpm heli's you lose most of it with swash mixes and lowering endpoints down n e way so would not see the full resolution n e thing is capable of ... theres other factors as well like servo speeds and the like that can be taken into the equation but my head hurts so wont bore you all with it lol...

I run one heli with a cyclock and one without and can tell the difference between one thats using alot more resolution than one that is not ....

[coolice]

Hey.

It is my understanding that any digital servo will see a benefit to increased resolution, at the end of the day we're not adding speed or torque, but just the amount of increment control steps a servo has from it's neutral to full left or right control.
An analogue servo will not be able to make full use of the increased resolution, but any servo will appear to be quicker even though it's not due to the faster data transmission rate from Tx to Rx.

I noticed a real marked difference going from my JR PCM10X to my DSX12, but I'd made two changes at the same time, the jump to 2.4 which is inherently quicker anyway and the increased resolution.

There is a much better and scientific explanation I am sure that I just do not know but out at the field I'm confident I can feel a difference between the old 35mhz and the now much loved 2.4ghz setup.
.

[Mslater]

Maybe it relies on the quality of the Servo's

At zero position on my 1024 set my ds610's only approximately return to the start position. They are always a little bit off.

would 2048 make this any better ?

[Steve]

Pete Christy (pchristy) wrote an excellent and very detailed forum post on just this topic some time ago (he is a bit of an electronics gem in this field), but I'm stuffed if I can find the post. Maybe drop him a PM with a link to this thread and ask him to comment on it again?

Found it here -> http://www.rcheliaddict.co.uk/main-d...its-bytes.html

[sacko]

Here it is

Bits & Bytes

There's been quite a few mentions of 1024 and 2048
resolution on here lately, and the general consensus
seems to be that the bigger the numbers, the better!

It ain't necessarily so, and I hope the following
explanation will clarify matters......

Any modern digital system (ie PCM or 2.4 GHz) breaks the
information it has to transmit down into binary numbers.
If you have an 8-bit system, it can represent 256
discrete steps. A 9-bit system can represent 512 steps.
A 10-bit system can represent 1024 steps, and an 11-bit
system will do 2048.

These "steps" are the smallest possible increment that
the system can resolve. You cannot make the system
respond to a change smaller than one of these "steps".

So what does this mean in practice?

Well, an 8-bit system with 256 discrete steps means that
the servo cannot possibly move by less than 1 part in
256 of its travel, or just under 1/2%. Similarly, a 9-
bit system will resolve 1 part in 512, or about 0.2%. By
the time you get to 11-bit systems you are talking about
a resolution of about .05%!

Now can anyone show me a servo that can resolve 0.05%?

No? Well, how about 0.2%? Perhaps if you were to pay a
small fortune, a brand new servo might just resolve that
amount, but it probably wouldn't manage it after a few
flights!

And what about linkages? How many of us can build a
linkage with less than 0.2% slop in it? When was the
last time you got hold of your paddles and gave them a
waggle (if you'll pardon the expression!)? Most of us
are lucky to get it down below 5%!!!

So in short, there is absolutely no point in going for
these ludicrous resolutions - or is there?

Well, internally - within the system - there is.

Everytime you mix a couple of functions together, you
also add their potential errors together. For example,
if you mix two channels equally, and each channel has a
resolution of 1%, the mixed output will only have a
resolution of 2%. When you are doing complex mixing - as
with eCCPM - the resolution of the final mix will be
lowered dramatically. So yes, you want to do all the
arithmetic at a high resolution, so that the final
output still falls within acceptable limits.

But there is absolutely NO POINT in *transmitting* that
level of resolution, because 1) the servos just aren't
that accurate and 2) neither are the linkages attached
to them!

So is there any downside to transmitting high resolution
data? Well certainly on 35 MHz, yes!

A 35 MHz channel is only 10 KHz wide. The more data you
try and stuff down it, the SLOWER you have to transmit
it! PCM already suffers from quite a high level of
latency due to the fact that once the data is received
it has to be checked for corruption, and then translated
into something the servos understand (PPM). This is why
on all PCM systems, there is a tiny but observable lag
between moving the stick and the servo moving. And the
more bits you use, the longer that lag gets!

The problem is much less significant at 2.4 GHZ, where
you have over 10 times the bandwidth available. But it
is still a total waste of effort because neither the
servos nor the linkages can possibly resolve that level
of accuracy!

So don't get too hung up on the numbers game! It doesn't
mean quite what you may think!

That is not to say that 2048 systems may not have other
benefits that the manufacturers have added, such as
additional data integrity checks. But it won't help to
improve the resolution of the controls in practice, and
it may even slow things down!

[GravityKills]

ERM Im even more confused lol.

*grin*


>>At zero position on my 1024 set my ds610's only approximately return to the start position. They are always a little bit off.

>>would 2048 make this any better ?
in practice, not. It's called "deadband": Once the servo gets within the deadband, it says "close enough" and goes to sleep.

[Hangar-7]

When we moved from 512 to 1024, I noticed a big difference (he he, I just started flying then) but as I progressed to digital servos etc - I could not tell much difference. My conclusion is that over the years, my anticipations have improved so a fraction faster does not mean much to me.

Same is probably true with 2048 - if you've just started you'd see a difference but to an already experienced person it may not mean much ....

[moyesboy]

Human perception is a complex thing, and although the numbers and theory suggest that you should not be able to tell the difference, anything that improves the response time of the control system seems to translate into a more connected feel to the controls.
However it might be a lot more noticable if you replace all the bearings and ball links in your heli than invest in higher resolution radio gear!

[Mslater]

Having studied the transmision protocols I've found the following.

on 1024 there is approx 1% resolution loss for each channel transmitted simultaneously.

on 2048 there is approx 0.5% resolution loss for each channel.

Here's where the improvement is. If you are using 6 channels (as most of us do) then 1024 system has 6% resolution loss.

That is more than the average 5% you may loose to link slop


on 2048 there is a 3% loss that is less than 5% link slop

So the benefit of a 2048 system over 1024 is a 1% resolution gain on 6 channels (yes 1% not 2% you still have 5% in the links remember)


On a 7 channel system this is 2% resolution gain
On 8 channel its 3% and so on.

So those of you who think you see an improvement are correct.

In summary..

1. If you have really high quality links then 2048 could make a big difference.
2. If you are flybarless 2048 will make a big difference.
3. 2048 slightly improves an already good heli
4. If you are running 8 Channels or more there will be a small difference
4. Makes no difference on an average heli.


So the answer for me on a standard 7Ch Trex 600 is there could might just possibly be a small improvement. Having said that most 2048Tx have other benefits - ball raced etc.

[oneandonlyian]

Hmm sound like a huge expense with minimal gain if you ask me.

[coolice]

Hey.

I think the opposite is true, it is the experienced fliers who spent many years on the older PCM radio which will most likely feel a difference. Also those fliers who are not true stick bangers, the latter just need to have a feeling of speed to go from one manouver to the next without much finer stuff inbetween.

For me I really noticed a change going from my JR 10X to my DSX12, but for me this was a big change as I'd also left 35mhz and so would have had the latency decrease of going to 2.4 as well as any advantage of 2048.
To me the models response to stick input was noticably quicker and more precise, using exactly the same model avionics I did on the former 35mhz.

I'm confident in myself and my flying to know it wasn't just me trying to justify my expense in buying the DSX12 ;-)
.