🚀 go-pugleaf

Doug Braun (dbraun@cad636.intel.com) wrote:
: 
: --
: 
: I have a 4-channel Futaba FM system, and I was thinking of getting
: a plane (Sig Samuri glider) that needs elevon mixing.
: 
: Does anyone make a little black box that goes in the plane
: to do the mixing electronically?  Otherwise, has anyone ever
: heard of a modifiecation to the transmitter to hard-wire
: it to do the mixing?  How about a mod to add more channels?
: I would assume the chip in the transmitter that does all the work
: has the capability of doing more than 4 channels, and you
: could wire in pots or switches to some unused inputs to
: get more channels.

As has been posted...

    You could use the ACE Christy electronic mixer.
    You could use a mechanical mixer (sliding tray or DuBro).
    You could use the RCD "MOM" (Master Onboard Mixer).

As to mixing... it is probably possible.  The biggest difficulty
would be creating two "inverse" signals to drive the alternate
ailerons.  Once this is accomplished, then mixing the elevator
signal onto these signals is probably fairly simple.  To provide
fuel for thought... I'll repost something I wrote up a while back
on adding channels to a Tx.  The main relevance to this discussion
is basic control stick circuitry.

---------------------------------------------------------------------------
  
Due to some limited demand... I present a little background info for
those interested in adding more channels to their transmitter.  Note
that to use these additional channels will probably require a new
receiver, but with pagers and 1991 specs most of you will probably be
purchasing a new receiver (at least) anyways.

Please excuse the resolution in the pictures, but even in low-res I
think a picture is worth a thousand words.  Also note that some of the
specifics I list here apply to a Futaba 4FG Tx... yours may be slightly
or massively different.  The principles will be the same though.

The heart of the Tx is the encoder chip.  This may be the only IC you
will find inside your Tx (if you take the back cover off).  One of the
more popular encoders is the Signetics NE5044 7 channel RC PPM encoder
(PPM = Proportional Pulse Modulation).  For those so inclined, this chip
may be found in the Signetics Linear Data Manual, Volume 2 Industrial
(as well as it's companion *decoder* NE5045 and a servo controller IC
NE544).  The encoder has 7 inputs which "measure" a voltage for each
channel.  The encoder also has a builtin voltage regulator which 
converts the ~10v battery power into a ~5.22v reference voltage.
The pin-out of the NE5044 follows:

	input 1		1	16		Vcc (Battery positive input)
	input 2		2	15		Vreg (Regulated ~5v output)
	input 3		3	14		mpx Capacitor
	input 4		4	13		Set Res.
	input 5		5	12		Range
	input 6		6	11		Serial Output (PPM signal to RF module)
	input 7		7	10		Output pulse R C oscillator
	Ground		8	 9		Frame Rate R C oscillator

Unused inputs are typically grounded, disabling pulse signal output
for that channel.  The operational range of the inputs is limited to the
range of 1.5v to 3.5v, however normal operating range is in the 1.9v to
3.1v range.  Note the implications of say 80 degrees of servo motion from
a voltage range of 1.2v... 0.015v per degree of servo movement.  In other
words these things are pretty sensitive/precise.  The inputs are also 
typically bypassed to ground with a capacitor to reduce circuit noise.  
Note that these inputs consume very little current (something like 200na 
max) so we can think of these as "open" connections.

What this means is that to add channels all we need to do is generate
a voltage signal that can be controlled in the range of ~1.9v to ~3.1v.
There are several ways to do this, and shown/discussed below is how it
is often done for general purpose, trim-able and reversable controls.
Note that this is one channel and that this circuit is repeated for
each of the channels.


             TYPICAL CONTROL CIRCUIT

	Vreg ---------------------------\
                                    |
  Control         Trim              |
     /--------------+-----------o   o   o        <<<=== DPDT switch used
     |              |           |       |
     |              |            \---\ /            for servo reversing.
     /              /                 \             Swaps Vreg and Ground
     \              \            /---/ \            connections
     /              /           |       |
  5k \<---o  9.26k  \<---o      o   o   o
     /    |         /    |      |   |
     \    |         \    |      |   |
     /    |         /    |      |   \--------------- Ground
     \    |         \    |      |
     |    |         |    |      |
     \--------------+-----------/
          |              |
          /              /            <<<=== Control/Trim mixing... relative
          \              \               values control relative sensitivity.
      39k /         150k /               Values should also be large relative
          \              \               to variable pots so as to minimize
          |              |               "non-linearity" (interactions).
          \-------+------/
                  |              10k
                  \------------v^v^v^----- Encoder channel input


Note that due to how the joy stick/trim lever mechanics are set up, little 
of the variable pot range is actually used, the control 5k pot ranges from 
2k to 3k while the trim pot ranges from 3.23k to 6.16k.

Some of you may recognize this as a combination of voltage dividers
(if you don't, now you know).  The adjustable pots for control and trim
each divide the voltage symetrically, ie: in the centered position the
output is 1/2 the input.  This allows for servo reversing by simply 
switching the positive voltage and ground inputs (which is what that
criss-cross on the switches is about in the diagram above).  These 
divided voltages are then again divided and combined by the mixing 
portion of the circuit.  The mixer adjusts relative sensitivity of 
the control and trim.  In the example above, control is scaled as 
155/194 (=.80) while trim is scaled as 41/191 (=.22).  These are 
approximate because the ground side/portion of the pots affect the 
action of each other (which is why the mixer resistor values need to 
be large... to minimize this interaction).

If your Tx is like mine it may actually already be partially
wired for additional channels.  The Futaba 4FG for example uses the
same PC board as the 6FG but without some of the components (like the
reversing switches, mixer resistors, etc.).  In this case you could
probably get factory replacement control pots and reversing switches
and have a "factory" upgrade.

On the other hand these parts may not be easy (or cheap) to get.  So
what then?  Well it depends on what kind of channel is desired.  Let's
start with a full range proportional channel.  Here we can use a generic
small variable resistor of some arbitrary size (like 10-50k) and create
a circuit like this:

                   10-50k Control Pot
     +5v -------------/\/\/\/\/\/-------------|> Ground (really)
                           ^
                           |
                           |
                           /
                           \ 2.5x k
                           /
                           \
                           |
                x k        |         x k
     +5v ----/\/\/\/\------o------/\/\/\/\----|> Ground
                           |
                           |
                           /
                           \  10k "magic" resistor
                           /
                           \
                           |
                           +---| |---|> Ground (noise filter capacitor)
                           |
                           V To encoder channel input

Note that if the +5v power and ground on the control pot are "crossable"
like in the previous circuit, then this will be a full range reversable
control channel.

We could also substitute a switch for the control pot if we only need
a two or three position control:

                DPDT switch (center "off" optional)
     +5v ----------o       o       o----------|> Ground (really)
                           ^
                           |
                           |
                           /
                           \ 2.5x k
                           /
                           \
                           |
                x k        |         x k
     +5v ----/\/\/\/\------o------/\/\/\/\----|> Ground
                           |
                           |
                           /
                           \  10k "magic" resistor
                           /
                           \
                           |
                           +---| |---|> Ground (noise filter capacitor)

Again if your radio is like mine, the "circuitry" for the secondary
divider resistors may already be there just without the resistors.

An important point on selection of resistor values for all of this is
to remember this is a battery powered device, thus use larger values
of resistors to minimize current and prolong battery life.  In my
case I used 50k control pots with 17.8k divider resistors (the "x" k
resisters above) and an 42.2/46.4k connector resistor (the "2.5x").

There is some tolerance in the actual value of the connector resistor.
Basically this resistor adjust the range of the control, thus you can
experiment to find the right value for the desired range (and thus
servo travel).  Or you could make it adjustable just for fun.  I 
suggest that you stick with a range slightly less than the range on
existing channels max. trim/control to min. trim/control.

As a final informational item, the following are the voltage values
I measured for a "typical" channel on my Futaba 4FG (all values in volts):

            Min Trim    Mid Trim    Max Trim    Delta Trim
Min Ctl      2.086       2.254       2.410       0.324
Mid Ctl      2.434       2.598       2.753       0.319
Max Ctl      2.805       2.954       3.122       0.317
Delta Ctl    0.719       0.700       0.712      MAX DELTA: 1.036

Well that's about it... next time: dual rates where there weren't
any!

Oh, and as always feel free to add enhancements, clarifications, or
to just ask questions (sorry about the ugly disclaimer but I don't
want anybody crying: you broke my radio! ): ).

Steve Jahr
Folsom/Roseville Ca.
------------------------------------------------------------------------
Disclaimer: Author assumes no responsibility for the use or reliability
of any of the preceeding information.  Use all info. etc. at your own risk!
------------------------------------------------------------------------
Steve Jahr
jahr@hprnd.rose.hp.com