 | | | | | |
In lugnet.announce, Lee Magpili wrote:
> check out www.mindstorms.com/press
<snip>
>
> whaaa!!
> LeE
Wow! Very cool! I want two to start with!
Kev
| | | | | | | | | | | | | | |
Just out of interest, who'd going to sell their old Mindstorms stuff.
According to the Wired article, the new NXT series sets won't be
compatible with the old Mindstorms stuff which is a bit alarming. It
makes me wonder if it is worth holding onto the old stuff or trying to
recoup some value from it now.
Having said all this, the new NXT looks superb. Three, count 'em, three
motors with built in rotation sensors!! and and ultrasound sensor. Don't
even get me started on BlueTooth connectivity!
The wait until August 2006 is going to be very very long!
Allen.
Kevin L. Clague wrote:
> In lugnet.announce, Lee Magpili wrote:
>
>
> > check out www.mindstorms.com/press
> >
> >
>
> <snip>
>
>
>
> > whaaa!!
> > LeE
> >
> >
>
>
> Wow! Very cool! I want two to start with!
>
> Kev
>
>
>
| | | | | | | | | | | | | | | | | |
In lugnet.robotics, Allen Foster wrote:
> Just out of interest, who'd going to sell their old Mindstorms stuff.
>
> According to the Wired article, the new NXT series sets won't be
> compatible with the old Mindstorms stuff which is a bit alarming. It
> makes me wonder if it is worth holding onto the old stuff or trying to
> recoup some value from it now.
>
>
My main worry is that the programming interfaces for the new and the old systems
won't be able to co-exist on the same computer system, so if you want to use
your old RCX you have to program it on different PC from the one you use for the
new RCX. THAT would be a royal pain and force many people to use one or the
other.
Also, are these motors going to be "better" than the old ones (i.e. in terms of
workload, torque, etc)? Certainly they look bigger and harder to incorporate
into a structure...
| | | | | | | | | | | | | | | | | | | | | |
In lugnet.robotics, Gyl Midroni wrote:
> Also, are these motors going to be "better" than the old ones (i.e. in terms of
> workload, torque, etc)? Certainly they look bigger and harder to incorporate
> into a structure...
And where are the studs????????????? It looks 100% liftarms to me
| | | | | | | | | | | | | | | | | | | | | | | | | | |
Subject:
|
Re: mindstorms NXT
|
Newsgroups:
|
lugnet.robotics
|
Date:
|
Thu, 5 Jan 2006 15:39:44 GMT
|
Original-From:
|
Matthew Davidson <MATTHEW@BLANK.nomorespamORG>
|
Viewed:
|
7504 times
|
| |
|
I guess 'digital wire interface' implies the output from the sensor
itself is digital? The RCX inputs are analog and the A/D occurs
inside the RCX, does it not?
| | | | | | | | | | | | | | | | | | | | | | | | | | | | |
Subject:
|
Re: mindstorms NXT
|
Newsgroups:
|
lugnet.robotics
|
Date:
|
Thu, 5 Jan 2006 16:36:30 GMT
|
Original-From:
|
Ignacio Martinez Vazquez <IGNAMV@GMAIL.COMnospam>
|
Viewed:
|
7819 times
|
| |
|
Maybe they're just making it sound better, but it'd certainly be
better to have digital sensors... although homebrew will be harder.
However, I don't think I'll be able to ask my dad for NeXT after the
HuGE expense of the cnc machine...
On 1/5/06, Matthew Davidson <matthew@blank.org> wrote:
>
> I guess 'digital wire interface' implies the output from the sensor
> itself is digital? The RCX inputs are analog and the A/D occurs
> inside the RCX, does it not?
>
| | | | | | | | | | | | | | | | | | | | | | | |
I don't see why you wouldn't be able to install both progs on one comp. People
could use NCQ progs and stock Mindstorms on the same comp, after all.
-Stefan-
| | | | | | | | | | | | | | | | | | | | | | |
In lugnet.robotics, Stefan Garcia wrote:
> I don't see why you wouldn't be able to install both progs on one comp. People
> could use NCQ progs and stock Mindstorms on the same comp, after all.
>
> -Stefan-
It sounds simple, but Mindstorms has never been very friendly with things like
ports and interrupts. If both systems are installed and competing for the same
computer resources..... trouble will occur. If they over-write each others'
DLLs, etc.
I guess only a hands-on trial will tell, and usually people find workarounds
either way.
| | | | | | | | | | | | | | | | | | | |
In lugnet.robotics, Allen Foster wrote:
> Just out of interest, who'd going to sell their old Mindstorms stuff.
Not me. From the press release it looks like this new NXT is going to
outclass the RCX, but that doesn't mean I won't play with the "old" stuff a
whole lot as well - I've not come near the limits of it as yet (I'm fine tuning
my 3T robot as I type... well, actually waiting for the program to download to
the brick).
> According to the Wired article, the new NXT series sets won't be
> compatible with the old Mindstorms stuff which is a bit alarming.
If they both are LEGO, then I can at least stick them on the same robot
platform and use them for different things. "Compatible" for Wired may not be
the same as "compatible" for me!
> The wait until August 2006 is going to be very very long!
Then apply to be in the test group (I'm *certain* you'll not be the first...
or last!):
http://mindstorms.lego.com/community/default.aspx
--
Brian Davis
| | | | | | | | | | | | | | | | | | |
Subject:
|
Re: mindstorms NXT
|
Newsgroups:
|
lugnet.robotics
|
Date:
|
Thu, 5 Jan 2006 14:08:48 GMT
|
Original-From:
|
"Simon Bogaert" <simon.bogaert@advalvas.be>
|
Viewed:
|
7302 times
|
| |
|
If this new sensor interface indeed is digital and indeed has an "open"
protocol, there will soon enough be some eletronics engineer designing an
interface or a multiplexer for the old sensors... Lego won't be the one
providing that interface, I guess, because it would slow down sales of their
new sensors. I'm confident the same will happen for the "old" motors...
Regards, Simon Bogaert
----- Original Message -----
From: "Allen Foster"
> According to the Wired article, the new NXT series sets won't be
> compatible with the old Mindstorms stuff which is a bit alarming. It
> makes me wonder if it is worth holding onto the old stuff or trying to
> recoup some value from it now.
| | | | | | | | | | | | | | | | | |
In lugnet.robotics, Allen Foster wrote:
> Just out of interest, who'd going to sell their old Mindstorms stuff.
>
> According to the Wired article, the new NXT series sets won't be
> compatible with the old Mindstorms stuff which is a bit alarming. It
> makes me wonder if it is worth holding onto the old stuff or trying to
> recoup some value from it now.
Yeah it woulda been nice if the new one had IR as well and be able to control
the "legacy" bricks. I'm sure the old ones will find a myriad of uses doing
simpler jobs like moving cows heads on train layouts ;)
> Having said all this, the new NXT looks superb. Three, count 'em, three
> motors with built in rotation sensors!! and and ultrasound sensor. Don't
> even get me started on BlueTooth connectivity!
Well, I hope they are also planning a smaller motor, without RPM sensor. I'm
sure there will be many situations when I'm happy to sacrifice accuracy for
size.
ROSCO
| | | | | | | | | | | | | | | | | |
Subject:
|
Re: mindstorms NXT
|
Newsgroups:
|
lugnet.robotics
|
Date:
|
Thu, 5 Jan 2006 22:14:06 GMT
|
Original-From:
|
Matthew Davidson <matthew@SPAMCAKEblank.org>
|
Viewed:
|
7377 times
|
| |
|
On Jan 5, 2006, at 8:21 PM, Ross Crawford wrote:
> Yeah it woulda been nice if the new one had IR as well and be able
> to control
> the "legacy" bricks. I'm sure the old ones will find a myriad of
> uses doing
> simpler jobs like moving cows heads on train layouts ;)
Interesting scenario. Perhaps someone can work up a homebrew
bluetooth to IR repeater. Sound crazy? I'll believe anything after
what I've seen people accomplish with the RCX, extending it with
gameboys and such. I'm sure there are existing portable devices with
both bluetooth and IR that can be hacked.
> Well, I hope they are also planning a smaller motor, without RPM
> sensor. I'm
> sure there will be many situations when I'm happy to sacrifice
> accuracy for
> size.
Me too. Those motors look bulky. Sometimes a micromotor is just what
you need for the job.
| | | | | | | | | | | | | | | | | | | | |
In lugnet.robotics, Matthew Davidson matthew@blank.org wrote:
| |
On Jan 5, 2006, at 8:21 PM, Ross Crawford wrote:
| |
Yeah it woulda been nice if the new one had IR as well and be able
to control
the “legacy” bricks. I’m sure the old ones will find a myriad of
uses doing
simpler jobs like moving cows heads on train layouts ;)
|
Interesting scenario. Perhaps someone can work up a homebrew
bluetooth to IR repeater. Sound crazy? I’ll believe anything after
what I’ve seen people accomplish with the RCX, extending it with
gameboys and such. I’m sure there are existing portable devices with
both bluetooth and IR that can be hacked.
|
I thought of this yesterday. If I end up as one of the 100 people who get it
early, I’ll connect a LEGO USB tower to the NXT USB port and see if I can get
the NXT to control the tower. I assume I’ll need a USB type-A to type-B
converter. As for the actual control, I’m hoping it’s just a matter of hacking
the USB driver in the Mindstorms SDK.
It would be cool to control the NXT from a Gameboy Advance using Bluetooth. I’ll
maybe tackle that, too.
| | | | | | | | | | | | | | | | | | | | | | | |
In lugnet.robotics, Jordan Bradford wrote:
On Jan 5, 2006, at 8:21 PM, Ross Crawford wrote:
I'm hoping it's just a matter of hacking the USB driver in
the Mindstorms SDK.
Which SDK would that be?
| | | | | | | | | | | | | | | | | | | | | | | | | | |
In lugnet.robotics, Dick Swan wrote:
| |
In lugnet.robotics, Jordan Bradford wrote:
On Jan 5, 2006, at 8:21 PM, Ross Crawford wrote:
I’m hoping it’s just a matter of hacking the USB driver in
the Mindstorms SDK.
Which SDK would that be?
|
http://mindstorms.lego.com/eng/community/resources/default.asp
The Ghost communication stack and API would have to be implemented in the NXT.
That’s actually what I meant when I said “driver.” A custom driver for the tower
would probably have to be written for the tower to run from within the NXT.
I think.
:)
| | | | | | | | | | | | | | | | | | | | | | | | |
Subject:
|
Re: mindstorms NXT
|
Newsgroups:
|
lugnet.robotics
|
Date:
|
Fri, 6 Jan 2006 18:46:25 GMT
|
Original-From:
|
"Simon Bogaert" <simon.bogaert@advalvas.be>
|
Viewed:
|
7821 times
|
| |
|
If Lego obeys the USB standard and uses fullsize receptacles, it'll be a "B"
type connector on the NXT. Hence, you'd need a B-male to A-female cable,
which is "illegal" according to the USB standards. You might hack up your
own, but that's cheating... ;-) So there's nog "legal" way to hook up the
low speed IR tower (fixed A-male connector) to the NXT brick. Also, I
repeat my concerns that Lego won't be putting a host-capable USB solution in
the NXT brick. Until we have certainty about that, It's all just guesswork,
and I certainly wouldn't count on using the USB port as a means of out-bound
communications other than to a computer.
Regards, Simon
----- Original Message -----
From: "Jordan Bradford"
> I thought of this yesterday. If I end up as one of the 100 people who get
> it
> early, I'll connect a LEGO USB tower to the NXT USB port and see if I can
> get
> the NXT to control the tower. I assume I'll need a USB type-A to type-B
> converter. As for the actual control, I'm hoping it's just a matter of
> hacking
> the USB driver in the Mindstorms SDK.
| | | | | | | | | | | | | | | | | | | | | | | |
In lugnet.robotics, Simon Bogaert wrote:
| |
If Lego obeys the USB standard and uses fullsize receptacles, it’ll be a “B”
type connector on the NXT. Hence, you’d need a B-male to A-female cable,
which is “illegal” according to the USB standards.
|
Indeed, see this image, also clearly showing motor connectors :
(from the article now available here http://www.nxtbot.com/)
| |
Also, I
repeat my concerns that Lego won’t be putting a host-capable USB solution in
the NXT brick.
|
An USB On The Go (host+slave) chip is not so expansive. But a host software
stack is much more complex, and would expose LEGO to a new heap of
incompatibilities with exotic device. Anyway, according to USB rules, a type B
connector is slave only, period.
Philo
| | | | | | | | | | | | | | | | | | | | | | | | | | |
In lugnet.robotics, Philippe Hurbain wrote:
| |
B connector is slave only, period.
|
Just as lego says:
What is the usability of the USB port in the NXT brick?
The USB port is a “slave” which means that it can only communicate with USB
“master” ports, e.g. computers. The usage of the USB port is for fast download
from PC to the NXT brick via USB cable and does not require any Bluetooth
capabilities in the computer.
http://www.lego.com/eng/service/faqs.asp?section=ConsumerService-FAQ-Products&catid=E8D0CD47-16B8-4B2F-900C-8FC40C163598&faqid=17264#17264
| | | | | | | | | | | | | | | | | | | | | | | | |
In lugnet.robotics, Simon Bogaert wrote:
| |
If Lego obeys the USB standard and uses fullsize receptacles, it’ll be a “B”
type connector on the NXT. Hence, you’d need a B-male to A-female cable,
which is “illegal” according to the USB standards. You might hack up your
own, but that’s cheating... ;-)
|
You mean like one of these? ;)
http://www.pacificcable.com/Picture_Page.asp?DataNAme=USBAFBM
| |
So there’s nog “legal” way to hook up the
low speed IR tower (fixed A-male connector) to the NXT brick. Also, I
repeat my concerns that Lego won’t be putting a host-capable USB solution in
the NXT brick. Until we have certainty about that, It’s all just guesswork,
and I certainly wouldn’t count on using the USB port as a means of out-bound
communications other than to a computer.
Regards, Simon
|
I’ve read the replies to this post, but I’ll respond here. As far as the cable
is concerned, the plug type is irrelevant; it’s just a standard. The real
determination of “master” and “slave” happens in the hardware/firmware. Could a
slave port be hacked into a master port in the firmware? We’ll see.
I suppose it would be easier, though, to set up a Bluetooth to computer to IR
tower communication protocol rather than try to run the tower directly from the
NXT.
Heh, this is an ivory tower academic project for me -- I’ve never built a robot
that needed more than one RCX! It’s like the time I briefly worked on a
communication protocol using the motor outputs and PWM as a code. :D
| | | | | | | | | | | | | | | | | | | | | | |
Subject:
|
hitechnic compass
|
Newsgroups:
|
lugnet.robotics
|
Date:
|
Sun, 8 Jan 2006 08:24:57 GMT
|
Original-From:
|
dan miller <danbmil99@yahoo!StopSpammers!.com>
|
Viewed:
|
8014 times
|
| |
|
I know everyone is psyched about the NXT, but on another topic--
does anyone have experience with the new compass sensor from Hitechnic? In
particular, I'm wondering what chip they use, and how sensitive it is to
non-horizontal placement. Also, can you use raw readings to get better than
3.6-degree accuracy?
any info appreciated
__________________________________________
Yahoo! DSL – Something to write home about.
Just $16.99/mo. or less.
dsl.yahoo.com
| | | | | | | | | | | | | | | | | | | | | | |
On Sun, January 8, 2006 3:24 am, dan miller wrote:
> I know everyone is psyched about the NXT, but on another topic--
>
> does anyone have experience with the new compass sensor from Hitechnic? In
> particular, I'm wondering what chip they use, and how sensitive it is to
> non-horizontal placement. Also, can you use raw readings to get better than
> 3.6-degree accuracy?
I got one a little while ago. It's pretty cool. With that, and one rotation
sensor, I was able to make a robot that could navigate a pre-defined course around
my house.
I'm not sure what chip is used, but I doubt you'll get better accuracy with raw
values.
I was just checking out the HiTechnic web site <http://www.hitechnic.com>. They
must have just redone it, because it looks like they're planning on making compasses
for the NXT, too.
I think he knows something... :)
Steve
| | | | | | | | | | | | | | | | | | |
In lugnet.robotics, Ross Crawford wrote:
>
> Well, I hope they are also planning a smaller motor, without RPM sensor. I'm
> sure there will be many situations when I'm happy to sacrifice accuracy for
> size.
>
> ROSCO
Actually, the RCX outputs 0's and 1's to power the motors as it is right now.
It uses Pulse Width Modulation (PWM). Basically for 1/3 power 1/3 of the pulse
is 1 and the other 2/3 is 0. For full power, the signal is all 1. This creates
a non-linear input/output curve, but works.
http://www.restena.lu/convict/Jeunes/ultimate_stuff/speed.htm
So all all you have to do is crimp your lego wires into a RJ11 and connect it to
the NXT and send it a digital PWM signal. Voila! Your old motor works again.
Of course, someone will have to build a block in "LabVIEW" to send the PWM
signal to the motor. Easy as pie if the software allows this.
Matthew C. Ruschmann
http://superpositioned.com
| | | | | | | | | | | | | | | | | |
In lugnet.robotics, Matthew Ruschmann wrote:
> Actually, the RCX outputs 0's and 1's to power the motors as it is right now.
> So all all you have to do is crimp your lego wires into a RJ11 and connect it to
> the NXT and send it a digital PWM signal. Voila! Your old motor works again.
I doubt that will do what you want. The RCX motor ports are equipped with motor
drivers - ICs that can provide rather a lot of current, switch direction, and
even short circuit the port. The digital wires would be I/O, not meant for
noticeable current, and only 1 or 0. So you'd wind up with an NXT shutting
itself off as the signal lines draw too much current.
No, what will be needed in the adapter will be such a motor driver, speaking the
appropriate serial protocol. Probably that means the device will contain a PIC
or PLC along with one or more motor drivers, and will be inefficient to build
for less than two motors per port. Personally I'm hoping for such a device with
separate power input.
| | | | | | | | | | | | | | | | | | | | | |
| |
|
In lugnet.robotics, Yann Vernier wrote:
> In lugnet.robotics, Matthew Ruschmann wrote:
> > Actually, the RCX outputs 0's and 1's to power the motors as it is right now.
> > So all all you have to do is crimp your lego wires into a RJ11 and connect it to
> > the NXT and send it a digital PWM signal. Voila! Your old motor works again.
>
> I doubt that will do what you want. The RCX motor ports are equipped with motor
> drivers - ICs that can provide rather a lot of current, switch direction, and
> even short circuit the port. The digital wires would be I/O, not meant for
> noticeable current, and only 1 or 0. So you'd wind up with an NXT shutting
> itself off as the signal lines draw too much current.
> No, what will be needed in the adapter will be such a motor driver, speaking the
> appropriate serial protocol. Probably that means the device will contain a PIC
> or PLC along with one or more motor drivers, and will be inefficient to build
> for less than two motors per port. Personally I'm hoping for such a device with
> separate power input.
I see no other wires into the motors, which means all the current is coming from
that one connection. Whether or not there is separate supply and control lines
in that connector is debatable until we get our hands on one - I would prefer
not, because that may make it harder to use the motor for non-NXT applications.
ROSCO
| | | | | | | | | | | | | | | | | | | | | | | |
| |
|
In lugnet.robotics, Ross Crawford wrote:
> In lugnet.robotics, Yann Vernier wrote:
> > In lugnet.robotics, Matthew Ruschmann wrote:
> > > Actually, the RCX outputs 0's and 1's to power the motors as it is right now.
> > > So all all you have to do is crimp your lego wires into a RJ11 and connect it to
> > > the NXT and send it a digital PWM signal. Voila! Your old motor works again.
> >
> > I doubt that will do what you want. The RCX motor ports are equipped with motor
> > drivers - ICs that can provide rather a lot of current, switch direction, and
> > even short circuit the port. The digital wires would be I/O, not meant for
> > noticeable current, and only 1 or 0. So you'd wind up with an NXT shutting
> > itself off as the signal lines draw too much current.
> > No, what will be needed in the adapter will be such a motor driver, speaking the
> > appropriate serial protocol. Probably that means the device will contain a PIC
> > or PLC along with one or more motor drivers, and will be inefficient to build
> > for less than two motors per port. Personally I'm hoping for such a device with
> > separate power input.
>
> I see no other wires into the motors, which means all the current is coming from
> that one connection. Whether or not there is separate supply and control lines
> in that connector is debatable until we get our hands on one - I would prefer
> not, because that may make it harder to use the motor for non-NXT applications.
>
> ROSCO
From the LEGO NXT FAQ...
"An electric motor whose speed or position is controlled by a closed loop
feedback circuit that can sense its position. The speed of the motor is measured
by a tachometer. The tachometer produces a voltage that is proportional to the
speed. This voltage is compared to a reference point and the difference, or
error, is used to adjust the speed of the motor, either up or down."
http://www.lego.com/eng/service/faqs.asp?section=ConsumerService-FAQ-Products&catid=E8D0CD47-16B8-4B2F-900C-8FC40C163598&faqid=17251#17251
So the motors operate in a variable voltage/current mode, where a target RPM in
the objective. Two wires for the tach (as a sensor), two for the motor current,
two for something else.
Ray
| | | | | | | | | | | | | | | | | | | |
| |
|
In lugnet.robotics, Matthew Ruschmann wrote:
> Actually, the RCX outputs 0's and 1's to power the motors as it is right now.
> It uses Pulse Width Modulation (PWM). Basically for 1/3 power 1/3 of the pulse
> is 1 and the other 2/3 is 0. For full power, the signal is all 1. This creates
> a non-linear input/output curve, but works.
>
> http://www.restena.lu/convict/Jeunes/ultimate_stuff/speed.htm
>
> So all all you have to do is crimp your lego wires into a RJ11 and connect it to
> the NXT and send it a digital PWM signal. Voila! Your old motor works again.
>
> Of course, someone will have to build a block in "LabVIEW" to send the PWM
> signal to the motor. Easy as pie if the software allows this.
I think that saying the RCX outputs 0's and 1's to the motor may be a bit
misleading with respect to considerations of digital signalling.
It is true that the CPU inside the RCX creates a logic level (1's and 0's) PWM
signal intended for the motors. But there is a key element between the RCX's CPU
and the motor outputs - an H bridge chip. This converts the switching logic
levels to switching power outputs. These switching power outputs are what makes
the motor go. Regular logic outputs are typically limited to a few tens of
milliamps at most, so a "digital" (ie; logic level) output would not be adequate
to drive the motor. But the H bridge outputs can handle upto about 800mA at
above logic level voltages, 9v for the RCX which is perfect for small DC motors.
JB
| | | | | | |
