Brickley Roscowicz Engineering

Brickley's
Words

HomeBrickley's WordsBrickley's Gallery

Archive for the 'Development' Category

Parkes Radio Telescope

July 18th, 2020 by RoscoHead
Photo byAdrian Tritschler

This story begins in 2011, the year of the radio telescope’s 50th anniversary. That was when I decided to build the model, hopefully having it ready for the anniversary on 31st October that year.

Alas, circumstances meant that didn’t happen, but in 2012 work began. I had to choose a scale, but I also knew I’d be limited by how I chose to build the main components, so this was really a balancing act. I built a “circle” using 12L axles & #3 axle joiners. That resulted in a dish about 0.5m in diameter, but I couldn’t come up with a way to make a decent looking round structure at that scale.

The dish scale

So I ended up using 2 x 12L axles between each joiner, approximately doubling the diameter to just over 1m, resulting in a scale of about 1:60. The bonus was, that is in my “acceptable minifig scale” range!

But that also meant it was going to be big, so I knew my big MOC chops would be tested. Hence the protruding tongue.

Of course the other main thing that determined the scale was how I chose to build the round structure. This was actually determined y my choice of window – in order to get the right number of windows, the structure ended up being almost exactly the right size for my chosen scale!

Unfortunately, to look good I had to put the shutter holders on the inside, and because the inside is concave, they had to be filed a little to fit. If you consider that heresy, maybe you should be reading a different blog…

So onto construction of the dish itself, and it was no simple matter to get the curved shape, keeping it strong & light at the same time. I used my old favourite technique: trial & error, and I was pretty happy with the result. Each pair of struts joined by 2 12L axles is a single unit, and there are 16 units to make the full circle.

Three strands of non-stretch braided fishing line are used to hold them together – there was no way clutch power would be able to do that.

Holding them together at the centre was not trivial either – I ended up using hinge plates between each strut pair, and a combination of beams to provide a strong mount point. I wanted to be able to tilt this without needing any other support, so this had to be strong. This mount went through several iterations before arriving at the final version.

Note that only 8 of the strut pairs are actually attached directly to the centre mount – every other pair is only held on by those hinge plates (top & bottom), and the axles at the rim. So I could have halved the number of struts and still had enough support, but I don’t think it would have looked as good.

On to the supporting structure. The real telescope is supported by a main central column, but I already knew I wasn’t going to be able to balance the dish, so that wasn’t an option. I ended up with 4 simple frames, one on each side, with some bracing. Then I re-used the slew ring from my CC-2800 crane, employing the droid wheel and a race of 2×2 technic wheels. This provided plenty of support for the rotating platform.

On the platform I built 2 well braced A-frames to mount the the tilting mechanism. Under one A-frame is the asimuth (slew) gearing, the elevation (tilt) gearing is under the other. The 2 NXT motors are mounted centrally, in opposing directions, topped of with the dish access door and central support column.

Next came the tilting mount. I had to make this super strong, because not only did it have to support the weight of the dish, but also some fairly decent shear forces when it was tilted. I also wanted to make it as heavy as possible, to reduce strain on the tilt mechanism.

Strength is provided by internal Technic beams, and weight is provided by 13 2x6x2 weight bricks. It is still nowhere near heavy enough to balance the dish.

To attach the dish to the mount, the beams on the bottom locate on the blue axle-pins you can see in the photo above, with the mount frames passing through the centre of the dish. Then the 2 black “hands” (also seen above) rotate to grab the top beams.

A centre block is then inserted to hold the “hands” in place. This block also carries a circle of paper to cover the centre section of the dish.

We’re getting close now, but I had to come up with a 6-sided box for the focus cabin. Inside there’s a bunch of Technic to get the shape right, unfortunately adding fairly significantly to the weight. But I am happy how it turned out. I was also able to add the small dish on top.

It is mounted on 3 legs, one of which has a lift for easy access to the focus cabin. I didn’t even try to make it movable, but it does at least tilt as the dish moves.

Building a shed to hide the NXT was a welcome (i short) relief. The cables run through a conduit that could be described as a path, then up the centre to the motors. This means it cannot turn continuously, but in reality it would never need to.

Why red? Simply because I have more 1xN bricks in red than any other colour. And it matched the doors.

Lastly, I decided to add some interior detail. I’ve no idea why, you can’t see it from outside anyway. There’s a few figs, some computer screens, and lots of knobs & dials.

There’s even a model of the dish on the ground floor!

And we can see proof that figs work late into the night…

And finally we get to the hardest part – the programming! It is programmed in NXC, using the excellent Bricx Command Centre, by Mark Overmars & John Hansen. It also uses the Bluetooth library “BTLib” by Daniele Benedettelli.

It consists of 2 main programs, “master” that runs on the NXT in the red shed and actually controls the dish, and “remote” that runs on a 2nd NXT and can be used to control “master” remotely via Bluetooth.

The programs share UI code, and the current position of the dish is displayed on both a couple of times per second.

In order to follow the path of astronomical bodies, I had to decide whether to calculate positions on-the-fly, or use lookup tables. I decided on the latter, as the calculations are not trivial, and that was not the main focus of the MOC anyway.

So I created a spreadsheet in which I can enter the details of the object to follow, and spit out data I can put into a files that are compiled then downloaded (or is that uploaded???) to the NXT.

Anyway, I zipped up a bunch of stuff I figured you programming types might be interested in, you can download it here.

And finally, a short video of the dish following a brand new, as yet unknown constellation.

And below is a gallery of images, feel free to browse.


New version of txt2dat

May 16th, 2009 by RoscoHead

Well it’s been a while, with no bug reports! I guess that means it’s working great for everyone 🙂

So I decided to implement a new feature that someone once asked for – pair kerning. Pair kerning adjusts the space between some character pairs to make them look better, for example L & W, which generally look too wide apart if the default spacing is used.

The kerning information is stored in the font file, and txt2dat now uses it to make the text look better. Note that it only applies horizontal kerning because

  • txt2dat doesn’t support multiple lines
  • FreeType doesn’t support vertical kerning

Anyway, check it out. The first example is without kerning, the second is with it.

Oh, and you can download txt2dat here.

LW without kerning

LW without kerning

LW with kerning

LW with kerning

New version of txt2dat

July 14th, 2007 by RoscoHead

Version 0.2.1 of my text to LDraw converter has been released. The changes in this version are:

  • Upgraded to Freetype 2.3.5
  • Removed need for common.c
  • Include Triangle source with txt2dat
  • Use autoconf for build configuration
  • Allow compilation without default font

Download it here.

rsoplay – Play RSO file on PC

April 24th, 2007 by RoscoHead

This is a small program to play NXT sound files (.RSO) on your PC. Not sure why anyone would really want to, but now you can anyway. Now Mac and linux can be just as useless as Windows!

Download it here.

txt2dat now available for Linux and Mac

March 14th, 2007 by RoscoHead

Thanks to Jim DeVona, the latest version of my txt2dat program (0.2.0) is now available for both Linux and Mac. Jim has tested these binaries on the machines he has available, but if you run into problems, please contact Jim or myself and we’ll see what we can do.

Find them here.


LEGO®,Technic®, Mindstorms® are trademarks of The LEGO Group of companies, which does not sponsor, authorise or endorse this site. Please visit the official LEGO site here.

Send mail to webadmin@br-eng.info with questions or comments about this web site.
Copyright © 2002-2007 R. Crawford

Brickley's Words is proudly powered by WordPress
Entries (RSS) and Comments (RSS).