coloradoz

My Feedback: (38)

OK Guys, here's one for you all, and it doesn't fall into the Chevy vs Ford catagory. Is there any real perceptable 'flying' difference between 8 bit (256 positions) vs 9 bit - (512 positions) vs 10 bit - (1024 positions) for the TX encoder?

On a real calm day with my plane just off the deck, I can tell the diff between 256 and 1024, but anything adjacent like 256 vs 512 or 512 vs 1024, I can't tell a diff. It seems like the errors introduced into the overall system are dominated by wind, pilot inputs/errors, hystersis between the commanded servo position and the TX/Pilot, and in general everything except the encoded servo position. The "old" radios, prior to the advent of the 'computer' radios offered a continum of commanded servo positions as opposed to the discretized digital systems. Did we loose something along the way or are the 'new' systems overdesiged?

I've been wondering about this for a while. Does anyone out there got some experience that can be related?

Lynx

Depends on the flying charactoristics of the aircraft in question more than anything else. 8bit probably isn't enough for a good smooth flight on say a 3D hellicopter or a high speed airplane, but 8bit is more than adequate for slow flyers or other sport planes. Considering the amount of movement you actually get at the control surface per bit is so rediculously small it has to be moving a lot of air (at high speed) or be highly unstable (like a helli) to really need more than 8 bits of resolution, though I'd absolutly go with 10bit if it were available simply because it is better even if it's not strictly speaking needed. Refresh rate would be more important if you ask me but there are very very few high refresh rate systems out there. On one note, once you go past 8 bits you're stuck with 16. Because that's how computers work. You'd have to have specially desiged custom micro processors to actually have a 10 bit system. I can't say this for sure, but most computer radio's should deal with the numbers internally as 16 bit and then translate them to 10 bit for xmitting, mainly because the bandwidth on RC systems is so limited.

coloradoz

My Feedback: (38)

Lynx, there is a big difference between the internal word representation of the uP and the A/D conversion accuracy - there are many processors out there that provide up to 10 channels of 8 to 12 bits of A/D accuracy (nothing to do with the internal word length representation of the uP). The resolution at the servo is also determined by the programmed throw, ie. is it 60 degrees/256 or 30 degrees/256, etc. I agree with your comments regarding refresh rate, but do you have direct experience with testing this out? I fly the old high speed pattern ships, and like I mentioned I can tell the difference between 256 and 1024 - but it is not problematic.

mr.rc-cam

I have a park flyer sport model (Electrajet) that uses a Plantraco DSP4 Rx. Its decoder has 5uS step resolution, which works out to about 7.5-bits of servo resolution. My intermediate level piloting skills cannot tell the difference between that and 10-bit step resolution. If I were into 3D then I would expect to feel a bit of difference.

RC-CAM

AirBearMA

My Feedback: (1)

I would think it is more a matter of use. If your control surfaces would normally require a long throw for normal flight and you do faster agressive aerobatics then you would probably notice the difference between an 8 bit vs 10 bit conversion. If you are more into slow flying type of simple pattern flying and your control surfaces have relatively short throws, you probably wouldn't notice the difference.