|
|
| |
Iain Hendry wrote:
> I figure I need to figure out what the error of each axis is (how far
> it is from the setpoint), and let the motor that moves the furthest
> "drive" the other two as slaves. Divide the error of the other 2
> motors by the error of the "drive" motor and use that with the
> HiTechnic PID block.
That sounds like a sensible solution.
Play well,
Jacob
--
Jacob's LEGO trains:
http://lego.sparre-andersen.dk/Transport/Tog/
|
|
| |
Hi all,
I'm only just now making the transition from RCX to NXT. The power of the new
servos is really something. For years I've "hard-coded" my own servo code to
couple a rotation sensor to a gearmotor to do setpoint-based positioning for
manipulators (pick and place robots). That is to say, I have a task running for
each motor on the RCX, watching it's current position, and a variable (a
setpoint) of where I want the motor to be. If there's a difference, the motor
moves and sets a flag when it's done.
I like to program these "drive" tasks so they run in the background, and my main
loop of code can just be a repetition of:
-Set the positions
-Tell it to go
-Wait for it to get there.
I want to do the same thing with NXT, but I want to go a step further. Let's
say I have a 3-axis cartesian robot. I know where I am (using the rotation
sensor block), and I can set a variable (a setpoint) for each of the 3 axes.
But now I want all 3 axes to move together so they all end up at the same
setpoint at the same time.
I figure I need to figure out what the error of each axis is (how far it is from
the setpoint), and let the motor that moves the furthest "drive" the other two
as slaves. Divide the error of the other 2 motors by the error of the "drive"
motor and use that with the HiTechnic PID block.
Does any of this make sense? Am I totally overcomplicating what I'm trying to
do? Surely I can't be the first person to want two or three motors to arrive at
a destination at the same time. :)
Thanks for any suggestions,
-Iain
|
|
| |
Hi I'm trying to read direct command sent to NXT via Bluetooth (i.e. via
SetOutputState) in a NXC program.
the goal is to remote control an NXT robot with a mobile phone, having the
program direct access to the data sent to the robot.
i.e. I want to know the direction and speed of both motor sent to the robot.
is this possible or maybe I have to use other type of message?
thank you
Cristian
|
|
| |
> you 4 have only 4 NXT
...I meant "you can have only 4 NXT" !!!
Philo
|
|
| |
> I don't know what the numeric limit is, but there is a limit of some sort, as I
> heard that Monster Chess ran into it and the central computer has to continually
> set up and tear down connections to the 32 different pieces, it cannot be in
> constant contact with all of them at once. So the limit has to be less than 32,
> for sure.
>
> I vaguely recall mention of 7? Not sure.
>
> Hopefully Steve H or another MC builder will comment
As far as I know the only limit is that you 4 have only 4 NXT connected together
at one time, one master and three slaves. This explains why the Monster Chess
that contains at least 32 NXT must constantly make and break connections. But
you can have a much higher number of NXT operating in the same room (eg. the
Monster Chess) without problems. There can be many masters talking at the same
time to 3 slaves each. One thing I heard (didn't experiment myself), the
discovery phase can take a VERY long time if there are many BT devices in the
same room. But if the NXT know each other this is not an issue either.
Philo
|
|
| |
In lugnet.robotics.nxt, Larry Pieniazek wrote:
> I vaguely recall mention of 7? Not sure.
Class is over and I'm still wondering why we could not fire up more than 5
bluetooth connections. Actually, these were NXT to NXT connections. The
RaceCar project at nxtprograms.com served as the basis for the RC Bluetooth
connections.
We tried quite a few things to get more than five students on at a time. We
know that when one disconnects, then another can grab the available connection.
We tried using another message box, making the BT connection down the hallway
and far from the classroom, etc.
We made sure firmware versions were the same. We mixed up the NXTs. We made
sure all the NXTs had at least 7.5v, etc. etc.
When I get home, I'm going to write a simple counter and display program and
have it run on as many pairs of NXT (I have 24). I'll report what I find.
|
|
| |
In lugnet.robotics.nxt, Edwin Pilobello wrote:
> I have a class of 10 kids each trying to fire up a remote control via bluetooth
> to their respective robots. Seems there is a limit of 5 connections to a room?
> Can anyone confirm?
I don't know what the numeric limit is, but there is a limit of some sort, as I
heard that Monster Chess ran into it and the central computer has to continually
set up and tear down connections to the 32 different pieces, it cannot be in
constant contact with all of them at once. So the limit has to be less than 32,
for sure.
I vaguely recall mention of 7? Not sure.
Hopefully Steve H or another MC builder will comment
|
|
| |
I have a class of 10 kids each trying to fire up a remote control via bluetooth
to their respective robots. Seems there is a limit of 5 connections to a room?
Can anyone confirm?
|
|
| |
OK, so I'm on my second ever NXT project, and I am wondering if this is even
possible. I have a program that currently has 3 WAIT Blocks, each set to
different lengths of time (4 sec, 2 sec, 1 sec) at different points. What I
want to do, to add more realism, is have that wait time be a random amount of
time between 1 and 5 seconds. The WAIT block, when set to time, seems to have
no method for input. Is there any other way to do this?
Thanks!
Chris Eyerly
|
|
| |
I have both 1.0 and 2.0 NXT sets, and I'd like to have all of the models from
both versions of the software to appear. I thought I could install both and I'd
get both sets of models, but I can't install the 1.0 software- it tells me that
I need at least 256 MB of memory, when I've got 4 GB. Is there some set of
files I can copy over from the disc that will give me the other models?
|
|
|
