Servo and pulse widths

[dam74]

Edders will have to take a look at them. Not sure how much I bought my Batron but I was looking best bang for buck. I really want to get into Denistron OLED displays to have a half decent graphics display (same sort of displays used in mobiles). However they're not cheap so I haven't bothered yet. Think the next thing I want to look into is making my own PCBs but the easiest method for making them seems to be with laser printer film transfers but I don't have a laser printer and I'm sure my Dad got rid of my old HP one that I had in Australia.

[pchristy]

Angelos makes some good points about the types of "creative accounting" used by some manufacturers of PCM equipment to try and improve data rates. Another method used is to transmit non-critical channels (such as gyro gain) at a much lower frame rate ie: every other or every third frame.

One of the problems with creative accounting however is that it requires much more decoding at the receiving end - where processing power is minimal due to power requirements. Further, as with any compression system, the decoding cannot start until the full data set has been received. This adds to the latency and explains why PCM sets always feel less connected than either PPM or 2.4 GHz systems regardless of the apparent frame rate put out by the receiver.

Just because the receiver is outputting a frame rate at 71 Hz, doesn't necessarily mean the pulses are actually being updated at that rate!

--
Pete

[Angelos]

pchristy,
I worked our how PCM1024 works some 6-7 years ago; down to the slightest detail. I even made my own decoder using a PIC. It is not data compression as used in computers. It is a simple remapping from one bit sequence to another in order to reduce the bandwidth requirements. Very easy to do with lookup takes. The protocol is very cleverly engineered. Each radio frame is broken down to sub-frames for 2 servos each. Two servo pulses get generated as soon as the sub-frame is received and decoded. This is why the PCM1024 receivers generate the servo pulses in pairs.

EDIT: The radio frame rate is 35.5Hz, however each sub-frame carries one absolute and one relative servo position. With the correct multiplexing all servos get updated at 71Hz.

-Angelos

[pchristy]

Neat! And well done on "reverse engineering" how they did it! Not an easy task!

I still maintain that the latency on even the most modern PCM systems can't match either PPM or 2.4 GHz though!

The easiest way to see it is when the set is on the bench. Simply slam a stick hard over and watch the servo closely. On all PCM systems there is a small but perceptible lag before the servo moves (on older systems, the lag is really noticeable!). On PPM and 2.4 GHz, the lag is imperceptible. However, as I said earlier, PPM does no error checking, so is much less secure!

How much difference this makes to an average pilot, I'm not sure! My reactions are a lot slower than they were 30 years ago, and the kinds of model I fly tend not to react that quickly anyway! My son maintains he can tell the difference, but me....! Hmmmm!!!

8/

--
Pete

[dam74]

I bought this

TextStar 'Stick-On' Serial LCD Display with 'GraphSmart' | Cat's Whisker Technologies


Edders

Looks very similar to the display they're using for this:

Midland Helicopters Ltd

[Edders]

it is the same one ;-)


Edders