Shop OBEX P1 Docs P2 Docs Learn Events
Submarine Servo Mixer — Parallax Forums

Submarine Servo Mixer

ArchiverArchiver Posts: 46,084
edited 2002-02-05 20:41 in General Discussion
Hello Paul,

As a first step, you might want to glance at a couple of sites offering mixers
for twin prop r/c boats.

http://www.astec.pwp.blueyonder.co.uk/

Their "amix" mixes rudder with twin props. They specify one mixer and two speed
controllers. All this is
transparent to the boat driver, who simply wiggles the rudder stick.

http://www.vantec.com/rdfr.htm

Same idea, but more sophisticated. This mixer works for robots, tanks, etc.,
and I notice they have a photo of a
minisub on their home page.

The vantec site in particular discusses the technical problems involved in some
detail.

For a long skinny military submarine with twin props mounted rather close
together, it is common to mix rudder and
differential prop steering. The two mixers noted in the links use the rudder as
the dominant channel, in the
sense that rudder deflection defines how much differential prop steering is
applied.

I would not know how to do this with a stamp, but I am a stamp novice at this
point. But let's say the stamp is
on the boat, and you are mixing rudder with differential prop steering. An
approach would be to 1) measure with
PULSIN the pulsewidth of the arriving PWM input for the rudder servo 2) go to a
lookup table and thus 3) locally
produce the PWM output you need, with PULSEOUT, to drive the motor speed
controllers. The tricky part is, you have
two motors and, thus, two controllers. For each indexed lookup, you need a
couple of values stored, one for each
motor.

A resource I have found helpful is Matt Gilliland' microcontroller cookbook. In
it you can find some ancillary
circuitry that might be useful. It shows, for example, how to set up circuits to
independently produce the PWM
pulse stream. These circuits can in turn be controlled by the stamp via a
digital pot or pots. This unloads the
PWM production chore from the stamp -- and gives it more time to think.

I was interested to see you have built a Dumas model. These are not easy!

For the racing sailboat I am mixing the mainsheet drum winch channel with a jib
trim servo channel.

The basic idea is to vary the slot between the jib and the main as a function of
boom position. From big boat
racing we know you are supposed to loosen up the jib sheet (open the slot) when
you go onto a reach. It makes the
boat go a lot faster.

But noboby really knows how much or precisely when to do to this -- its just a
body of sailing lore. In other
words, if there is a mixing formula, nobody knows what it is. Or (maybe more
likely) they are not telling. This is
racing, after all.

I want to do the mixing with the stamp because I want to set up the mixing
empirically, point by point. Notch the
main out by an increment -- then by trial and error try to discover the slot
setting that makes the boat go
fastest. Record the result for a lookup table. And repeat.

Commercial mixing gear for r/c will only give you five points to define a mixing
curve. I want 25 or so. Hence
the stamp.



md10bldr wrote:

> So many good questions!
> Let me start with the general concept. I'm thinking along the lines
> of Jason - one of the mini sub/drones used when diving for the
> titanic. There are motors on each side, which would allow you to spin
> in place, or make slow tuns as you go forward. Similar to diffential
> steering on a robot. There are also motors within the body pointed
> straight up and down. I guess you could say that it moves a bit like
> a helicopter.
>
> So:
> To equate this to a two stick transmitter, rudder side to side would
> spin the motors in opposite directions, producing a spin in place.If
> the elevator stick is pushed forward, both The two side motors would
> drive, producing (hopefully) a stright line forward. The mixture
> proportions are to be determined. The throttle stick would be used to
> control vertical motion. Dominant channel? Hmmmm. Guess I would have
> to finalize the configuration (dimensions, placement, etc.) to
> determine that.
>
> As I don't know too much about how the reciever decodes the signal, I
> figured I would do the mixing between the reciever and servos.
>
> Although I plan to start with some simple formulas for mixing, I
> don't know that a linear response will do the job. Depending on the
> complexity of the formula, and the Stamp speed, maybe a lookup table
> would be faster.
>
> And to the last question regarding the availablity of mixing radio
> systems - I have a fair amout of radio control equipment laying
> around from previous hobby endevours (including a dumas Huscon 36), a
> new house with pool, and a shiny new Stamp. You get the picture. The
> Stamp has so much potential, constrained only by my ignorance. It's
> all about learning something new!
>
> By the way, what are you considering for your sailboat? Sail control
> based on wind speed/direction, or maybe "autopilot"??? could be even
> more intresting than a submarine.
>
> --- In basicstamps@y..., Michael Gianturco <michcg@m...> wrote:
> > Hello Paul,
> >
> > Do you want to mix props and rudder, for example? Or do you have in
> mind a submarine system that works like an
> > v-tail or x-tailed aircraft?
> >
> > What is the dominant channel? Rudder?
> >
> > I am struggling through an r/c mixing problem, for a racing
> sailboat, so perhaps we have some similar problems.
> > Are you planning to mix at the xmtr, or at the receiver? Do you
> have a formula for relating the two channels to
> > be mixed? Or do you plan to do it point by point?
> >
> > The submarine R/C group is a very helpful bunch, incidentally. They
> have a very active forum at:
> >
> > http://www.subcommittee.com
> >
> > Lastly, quite a few radios offer mixing as a built-in feature, and
> they are not too expensive, to judge from the
> > offerings on ebay. Maybe the commercial mixers won't do exactly
> what you want, which is a good reason to start
> > from scratch, but perhaps they would come pretty close.
> >
> > best, Michael
> >
> >
> >
> > md10bldr wrote:
> >
> > > I'd like to use my BS2E as a multi channel servo mixer. Is the
> BS2E
> > > fast enough? I'm thinking along the lines of a swiming pool
> > > submarine, so the speed and accuracy needed for an airplane
> probably
> > > wouldn't be necessary.
> > >
> > > Has anyone done this before? How do you establish what order to
> check
> > > the output of the receiver, and how do you ensure you don't miss
> too
> > > many pulses?
> > >
> > > To UNSUBSCRIBE, just send mail to:
> > > basicstamps-unsubscribe@y...
> > > from the same email address that you subscribed. Text in the
> Subject and Body of the message will be ignored.
> > >
> > >
> > > Your use of Yahoo! Groups is subject to
> http://docs.yahoo.com/info/terms/
>
> To UNSUBSCRIBE, just send mail to:
> basicstamps-unsubscribe@yahoogroups.com
> from the same email address that you subscribed. Text in the Subject and Body
of the message will be ignored.
>
>
> Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/

Comments

  • ArchiverArchiver Posts: 46,084
    edited 2002-02-02 14:41
    Afterthought. The Astec site is an incomprehensible jumble of racing boat news
    and graphics. To find the mixing
    circuit, go to the menu panel on the left and click on "Electronics". Then
    scroll down past the ESCs (electronic speed
    controllers) to find the mixer.



    Michael Gianturco wrote:

    > Hello Paul,
    >
    > As a first step, you might want to glance at a couple of sites offering mixers
    for twin prop r/c boats.
    >
    > http://www.astec.pwp.blueyonder.co.uk/
    >
    > Their "amix" mixes rudder with twin props. They specify one mixer and two
    speed controllers. All this is
    > transparent to the boat driver, who simply wiggles the rudder stick.
    >
    > http://www.vantec.com/rdfr.htm
    >
    > Same idea, but more sophisticated. This mixer works for robots, tanks, etc.,
    and I notice they have a photo of a
    > minisub on their home page.
    >
    > The vantec site in particular discusses the technical problems involved in
    some detail.
    >
    > For a long skinny military submarine with twin props mounted rather close
    together, it is common to mix rudder and
    > differential prop steering. The two mixers noted in the links use the rudder
    as the dominant channel, in the
    > sense that rudder deflection defines how much differential prop steering is
    applied.
    >
    > I would not know how to do this with a stamp, but I am a stamp novice at this
    point. But let's say the stamp is
    > on the boat, and you are mixing rudder with differential prop steering. An
    approach would be to 1) measure with
    > PULSIN the pulsewidth of the arriving PWM input for the rudder servo 2) go to
    a lookup table and thus 3) locally
    > produce the PWM output you need, with PULSEOUT, to drive the motor speed
    controllers. The tricky part is, you have
    > two motors and, thus, two controllers. For each indexed lookup, you need a
    couple of values stored, one for each
    > motor.
    >
    > A resource I have found helpful is Matt Gilliland' microcontroller cookbook.
    In it you can find some ancillary
    > circuitry that might be useful. It shows, for example, how to set up circuits
    to independently produce the PWM
    > pulse stream. These circuits can in turn be controlled by the stamp via a
    digital pot or pots. This unloads the
    > PWM production chore from the stamp -- and gives it more time to think.
    >
    > I was interested to see you have built a Dumas model. These are not easy!
    >
    > For the racing sailboat I am mixing the mainsheet drum winch channel with a
    jib trim servo channel.
    >
    > The basic idea is to vary the slot between the jib and the main as a function
    of boom position. From big boat
    > racing we know you are supposed to loosen up the jib sheet (open the slot)
    when you go onto a reach. It makes the
    > boat go a lot faster.
    >
    > But noboby really knows how much or precisely when to do to this -- its just a
    body of sailing lore. In other
    > words, if there is a mixing formula, nobody knows what it is. Or (maybe more
    likely) they are not telling. This is
    > racing, after all.
    >
    > I want to do the mixing with the stamp because I want to set up the mixing
    empirically, point by point. Notch the
    > main out by an increment -- then by trial and error try to discover the slot
    setting that makes the boat go
    > fastest. Record the result for a lookup table. And repeat.
    >
    > Commercial mixing gear for r/c will only give you five points to define a
    mixing curve. I want 25 or so. Hence
    > the stamp.
    >
    > md10bldr wrote:
    >
    > > So many good questions!
    > > Let me start with the general concept. I'm thinking along the lines
    > > of Jason - one of the mini sub/drones used when diving for the
    > > titanic. There are motors on each side, which would allow you to spin
    > > in place, or make slow tuns as you go forward. Similar to diffential
    > > steering on a robot. There are also motors within the body pointed
    > > straight up and down. I guess you could say that it moves a bit like
    > > a helicopter.
    > >
    > > So:
    > > To equate this to a two stick transmitter, rudder side to side would
    > > spin the motors in opposite directions, producing a spin in place.If
    > > the elevator stick is pushed forward, both The two side motors would
    > > drive, producing (hopefully) a stright line forward. The mixture
    > > proportions are to be determined. The throttle stick would be used to
    > > control vertical motion. Dominant channel? Hmmmm. Guess I would have
    > > to finalize the configuration (dimensions, placement, etc.) to
    > > determine that.
    > >
    > > As I don't know too much about how the reciever decodes the signal, I
    > > figured I would do the mixing between the reciever and servos.
    > >
    > > Although I plan to start with some simple formulas for mixing, I
    > > don't know that a linear response will do the job. Depending on the
    > > complexity of the formula, and the Stamp speed, maybe a lookup table
    > > would be faster.
    > >
    > > And to the last question regarding the availablity of mixing radio
    > > systems - I have a fair amout of radio control equipment laying
    > > around from previous hobby endevours (including a dumas Huscon 36), a
    > > new house with pool, and a shiny new Stamp. You get the picture. The
    > > Stamp has so much potential, constrained only by my ignorance. It's
    > > all about learning something new!
    > >
    > > By the way, what are you considering for your sailboat? Sail control
    > > based on wind speed/direction, or maybe "autopilot"??? could be even
    > > more intresting than a submarine.
    > >
    > > --- In basicstamps@y..., Michael Gianturco <michcg@m...> wrote:
    > > > Hello Paul,
    > > >
    > > > Do you want to mix props and rudder, for example? Or do you have in
    > > mind a submarine system that works like an
    > > > v-tail or x-tailed aircraft?
    > > >
    > > > What is the dominant channel? Rudder?
    > > >
    > > > I am struggling through an r/c mixing problem, for a racing
    > > sailboat, so perhaps we have some similar problems.
    > > > Are you planning to mix at the xmtr, or at the receiver? Do you
    > > have a formula for relating the two channels to
    > > > be mixed? Or do you plan to do it point by point?
    > > >
    > > > The submarine R/C group is a very helpful bunch, incidentally. They
    > > have a very active forum at:
    > > >
    > > > http://www.subcommittee.com
    > > >
    > > > Lastly, quite a few radios offer mixing as a built-in feature, and
    > > they are not too expensive, to judge from the
    > > > offerings on ebay. Maybe the commercial mixers won't do exactly
    > > what you want, which is a good reason to start
    > > > from scratch, but perhaps they would come pretty close.
    > > >
    > > > best, Michael
    > > >
    > > >
    > > >
    > > > md10bldr wrote:
    > > >
    > > > > I'd like to use my BS2E as a multi channel servo mixer. Is the
    > > BS2E
    > > > > fast enough? I'm thinking along the lines of a swiming pool
    > > > > submarine, so the speed and accuracy needed for an airplane
    > > probably
    > > > > wouldn't be necessary.
    > > > >
    > > > > Has anyone done this before? How do you establish what order to
    > > check
    > > > > the output of the receiver, and how do you ensure you don't miss
    > > too
    > > > > many pulses?
    > > > >
    > > > > To UNSUBSCRIBE, just send mail to:
    > > > > basicstamps-unsubscribe@y...
    > > > > from the same email address that you subscribed. Text in the
    > > Subject and Body of the message will be ignored.
    > > > >
    > > > >
    > > > > Your use of Yahoo! Groups is subject to
    > > http://docs.yahoo.com/info/terms/
    > >
    > > To UNSUBSCRIBE, just send mail to:
    > > basicstamps-unsubscribe@yahoogroups.com
    > > from the same email address that you subscribed. Text in the Subject and
    Body of the message will be ignored.
    > >
    > >
    > > Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/
    >
    > To UNSUBSCRIBE, just send mail to:
    > basicstamps-unsubscribe@yahoogroups.com
    > from the same email address that you subscribed. Text in the Subject and Body
    of the message will be ignored.
    >
    >
    > Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/
  • ArchiverArchiver Posts: 46,084
    edited 2002-02-05 20:41
    Michael -
    Your sailboat project sounds very interesting. What inputs do you
    plan on using? I'm thinking windspeed, wind direction and boat speed.
    I would think that with the relatively slow speed of a sailboat, the
    stamp should be able to perform the calculations realtime as opposed
    to a table. Have you given any thought to having the stamp control
    both the main and the jib to attain the best speed for a given
    heading?

    Don't know if you've seen this, but there is some useful information
    on wind speed/direction in the sensor interface section.

    http://www.emesystems.com/BS2index.htm

    For boat speed, I'd imagine a short trailing wire with some type of
    prop on it would do the job - although I don't know if the rules you
    are racing(?) under will allow that.

    Best of luck, and if you fell like bouncing some ideas back and
    forth, feel free to e-mail me direct.
    PG


    --- In basicstamps@y..., Michael Gianturco <michcg@m...> wrote:
    > Hello Paul,
    >
    > As a first step, you might want to glance at a couple of sites
    offering mixers for twin prop r/c boats.
    >
    > http://www.astec.pwp.blueyonder.co.uk/
    >
    > Their "amix" mixes rudder with twin props. They specify one mixer
    and two speed controllers. All this is
    > transparent to the boat driver, who simply wiggles the rudder stick.
    >
    > http://www.vantec.com/rdfr.htm
    >
    > Same idea, but more sophisticated. This mixer works for robots,
    tanks, etc., and I notice they have a photo of a
    > minisub on their home page.
    >
    > The vantec site in particular discusses the technical problems
    involved in some detail.
    >
    > For a long skinny military submarine with twin props mounted rather
    close together, it is common to mix rudder and
    > differential prop steering. The two mixers noted in the links use
    the rudder as the dominant channel, in the
    > sense that rudder deflection defines how much differential prop
    steering is applied.
    >
    > I would not know how to do this with a stamp, but I am a stamp
    novice at this point. But let's say the stamp is
    > on the boat, and you are mixing rudder with differential prop
    steering. An approach would be to 1) measure with
    > PULSIN the pulsewidth of the arriving PWM input for the rudder
    servo 2) go to a lookup table and thus 3) locally
    > produce the PWM output you need, with PULSEOUT, to drive the motor
    speed controllers. The tricky part is, you have
    > two motors and, thus, two controllers. For each indexed lookup, you
    need a couple of values stored, one for each
    > motor.
    >
    > A resource I have found helpful is Matt Gilliland' microcontroller
    cookbook. In it you can find some ancillary
    > circuitry that might be useful. It shows, for example, how to set
    up circuits to independently produce the PWM
    > pulse stream. These circuits can in turn be controlled by the stamp
    via a digital pot or pots. This unloads the
    > PWM production chore from the stamp -- and gives it more time to
    think.
    >
    > I was interested to see you have built a Dumas model. These are
    not easy!
    >
    > For the racing sailboat I am mixing the mainsheet drum winch
    channel with a jib trim servo channel.
    >
    > The basic idea is to vary the slot between the jib and the main as
    a function of boom position. From big boat
    > racing we know you are supposed to loosen up the jib sheet (open
    the slot) when you go onto a reach. It makes the
    > boat go a lot faster.
    >
    > But noboby really knows how much or precisely when to do to this --
    its just a body of sailing lore. In other
    > words, if there is a mixing formula, nobody knows what it is. Or
    (maybe more likely) they are not telling. This is
    > racing, after all.
    >
    > I want to do the mixing with the stamp because I want to set up the
    mixing empirically, point by point. Notch the
    > main out by an increment -- then by trial and error try to discover
    the slot setting that makes the boat go
    > fastest. Record the result for a lookup table. And repeat.
    >
    > Commercial mixing gear for r/c will only give you five points to
    define a mixing curve. I want 25 or so. Hence
    > the stamp.
    >
    >
    >
    > md10bldr wrote:
    >
    > > So many good questions!
    > > Let me start with the general concept. I'm thinking along the
    lines
    > > of Jason - one of the mini sub/drones used when diving for the
    > > titanic. There are motors on each side, which would allow you to
    spin
    > > in place, or make slow tuns as you go forward. Similar to
    diffential
    > > steering on a robot. There are also motors within the body pointed
    > > straight up and down. I guess you could say that it moves a bit
    like
    > > a helicopter.
    > >
    > > So:
    > > To equate this to a two stick transmitter, rudder side to side
    would
    > > spin the motors in opposite directions, producing a spin in
    place.If
    > > the elevator stick is pushed forward, both The two side motors
    would
    > > drive, producing (hopefully) a stright line forward. The mixture
    > > proportions are to be determined. The throttle stick would be
    used to
    > > control vertical motion. Dominant channel? Hmmmm. Guess I would
    have
    > > to finalize the configuration (dimensions, placement, etc.) to
    > > determine that.
    > >
    > > As I don't know too much about how the reciever decodes the
    signal, I
    > > figured I would do the mixing between the reciever and servos.
    > >
    > > Although I plan to start with some simple formulas for mixing, I
    > > don't know that a linear response will do the job. Depending on
    the
    > > complexity of the formula, and the Stamp speed, maybe a lookup
    table
    > > would be faster.
    > >
    > > And to the last question regarding the availablity of mixing radio
    > > systems - I have a fair amout of radio control equipment laying
    > > around from previous hobby endevours (including a dumas Huscon
    36), a
    > > new house with pool, and a shiny new Stamp. You get the picture.
    The
    > > Stamp has so much potential, constrained only by my ignorance.
    It's
    > > all about learning something new!
    > >
    > > By the way, what are you considering for your sailboat? Sail
    control
    > > based on wind speed/direction, or maybe "autopilot"??? could be
    even
    > > more intresting than a submarine.
    > >
    > > --- In basicstamps@y..., Michael Gianturco <michcg@m...> wrote:
    > > > Hello Paul,
    > > >
    > > > Do you want to mix props and rudder, for example? Or do you
    have in
    > > mind a submarine system that works like an
    > > > v-tail or x-tailed aircraft?
    > > >
    > > > What is the dominant channel? Rudder?
    > > >
    > > > I am struggling through an r/c mixing problem, for a racing
    > > sailboat, so perhaps we have some similar problems.
    > > > Are you planning to mix at the xmtr, or at the receiver? Do you
    > > have a formula for relating the two channels to
    > > > be mixed? Or do you plan to do it point by point?
    > > >
    > > > The submarine R/C group is a very helpful bunch, incidentally.
    They
    > > have a very active forum at:
    > > >
    > > > http://www.subcommittee.com
    > > >
    > > > Lastly, quite a few radios offer mixing as a built-in feature,
    and
    > > they are not too expensive, to judge from the
    > > > offerings on ebay. Maybe the commercial mixers won't do exactly
    > > what you want, which is a good reason to start
    > > > from scratch, but perhaps they would come pretty close.
    > > >
    > > > best, Michael
    > > >
    > > >
    > > >
    > > > md10bldr wrote:
    > > >
    > > > > I'd like to use my BS2E as a multi channel servo mixer. Is the
    > > BS2E
    > > > > fast enough? I'm thinking along the lines of a swiming pool
    > > > > submarine, so the speed and accuracy needed for an airplane
    > > probably
    > > > > wouldn't be necessary.
    > > > >
    > > > > Has anyone done this before? How do you establish what order
    to
    > > check
    > > > > the output of the receiver, and how do you ensure you don't
    miss
    > > too
    > > > > many pulses?
    > > > >
    > > > > To UNSUBSCRIBE, just send mail to:
    > > > > basicstamps-unsubscribe@y...
    > > > > from the same email address that you subscribed. Text in the
    > > Subject and Body of the message will be ignored.
    > > > >
    > > > >
    > > > > Your use of Yahoo! Groups is subject to
    > > http://docs.yahoo.com/info/terms/
    > >
    > > To UNSUBSCRIBE, just send mail to:
    > > basicstamps-unsubscribe@y...
    > > from the same email address that you subscribed. Text in the
    Subject and Body of the message will be ignored.
    > >
    > >
    > > Your use of Yahoo! Groups is subject to
    http://docs.yahoo.com/info/terms/
Sign In or Register to comment.