Amateur satellite antenna controller

[geoff07]

Before I get too carried away with my design for a controller for my antenna for accessing OSCAR satellites I wondered if anyone out there might have had a go at this?
There are many possible features but the basic idea is to replace a Kansas City Tracker (a very good but obsolete VLB-bus PC card) with a PICAXE + LCD to control the Yaesu G5500 controller and the radio uplink and downlink frequencies from the orbit calculations done on the PC. This isn't hard in principle but any experience I could tap into would be helpful.

 

[BrendanP]

Manuka is the man for this.

Manuka is the man for this, hes a ham. He'll be all over this like a cheap suit next time he looks at the forum

 

[Tom2000]

Before you can evaluate the hardware you're going to replace with your homebrew external hardware, you need to gather some information and do some analysis.

A place to begin might be this:

1. What info does the PC feed the controller - for the rotors, and for the radios. What is the PC/controller interface? What are the data formats? Is the communication one-way, or does the PC need rotor positioning feedback or some sort of return from the radios (such as S meter readings, confirmation of frequency and mode change commands, etc.)

2. What are the physical interfaces to each radio and to the rotors? And to each rotor position sensing system?

3. What are the data formats between the PC and the controller, and between the controller and each peripheral?

4. What calculations, data translations, etc, does the controller need to perform?

You don't need to answer these questions in this forum, but you darned sure need to get them at hand and write them down.

I'm sure that there are more questions that might pop into my mind if I gave this some serious thought, but it's stuff like this you need to learn and analyze before you can start making design decisions.

Just off the top of my head, one or more Picaxe processors might be of some use in your project, but I think that you'll need some more powerful hardware in there somewhere. However, I wouldn't even hazard a guess until I'd done some detailed analysis and derived a detailed system definition and specification.

If the Kansas City Tracker is very well defined and specified, and if you can get your hands on such a document, the hard work is done. You can design a system from those specifications.

Good luck with your project!

73,

Tom

 

[manuka]

Bravo -things have been pretty much answered above, but which OSCAR ?

For those who've just come in - OSCAR is an acronym for Orbital Satellite Carrying Amateur Radio,with the first launched back in the 1960s. Scores have since been launched, with many of these microsats. now quite powerful (thanks to solar PV advances), & the beacons of Low Earth Orbital (LEO) ones can often be heard just with a hand held transceiver & rubber ducky antenna! See=> http://www.amsat.org/amsat-new/satellites/status.php

In spite of Emily Clarke's enthusiasm ( => http://www.projectoscar.net/symposium/Intro_sats.pdf ) I confess terrestial wireless communications presently take my fancy, but I'm happy to chip in for the heavens. Stan - ZL2AJZ

 

[tikeda]

It's been done with a Basic Stamp 2 (SAEBRTrack by N8MH). Also described on ARRL pages here.

The project seems to require a couple ADC inputs (for each axis) and four connections for up, down, cw and ccw movement. That, and an interface to the computer.

What would be neat is having the ability to download pre-set paths from a PC and run the rotators offline while in the field.

 

[geoff07]

next question re fpu: Technical?

Thanks for all the replies. I think I have answers to the points that have been noted, and clearly noone has identified any gotchas that I have to think about up front. I'm aware of the Stamp and other implementations, but I want to use PICAXE for reasons of rapid progress and upgradeability in what could become a very complex system (see below).

The input is 'up a bit/left a bit' signals from the PC plus acsii frequencies, and position signals (potentiometers) from the rotator. The output is control signals for the radio and control of either four relays or perhaps direct control of the motor windings.

Stage two is intended to be a webpage front end to avoid the need for switches etc and to allow the device to be mounted on the antenna mast to reduce power cabling.

Stage three will be an autonomous feature as mentioned by tikeda. This is driven by the Kepler elements (orbital equation data) issued by Norad which would be uploaded periodically to the unit. To track a satellite you solve this equation in real time and it is the solution that I want to program, so there is no other need for processing. However for this I will need to be able to do floating point calcs (celestial mechanics requires a lot of coordinate conversions and other high precision sums). The code in general use is the PLAN13 program writted by G3RUH in 1990 and oft-plagarised since. I'm aware of the fpu device but I find that the usage of that is very complex. I might be able to handle the logic of the calls but I wonder if the chip memory would be large enough for the code. So, perhaps Technical could say whether Revolution has any plans to include floating point primitives in the PICAXE armoury to make driving the fpu easier? I would hope that this could make it much simpler (=quicker) to code and help it fit the space. The fpu seems to me to need some more integration than it currently has. If this proves too much for a PICAXE then I would do as tikeda suggests which is to precalc the orbit info and follow the path. But then it isn't autonomous!

Stage four is gps input for exact time and antenna location (useful if mobile), plus wind speed.

Whether I will ever get beyond stage two is a very good question, but if a PICAXE has the capacity to do it in principle then it should not be too hard.

I used to be busy on the pacsats ten years ago but then I worked abroad in a busy job and all that had to be parked. Back then I ran stages 1 and 2 plus the download of all the email signals from the sat, all using Wisp/KCT on a 16MHz 386 VLB PC. Now I am semi-retired I fancy a new challenge to get up and running again.

G8ZNW

 

[tikeda]

A 1-4 MBit serial flash chip could hold a lot of motor movement data for passes, perhaps more than enough before the old set of orbital elements go stale.

Other options to the FPU chips... The Arduino supports floating point variables with its math library (But that capability can be costly in terms of performance and memory). There are systems like the Netburner line that include 32-bit Coldfire processors with plenty of computing power. The lowest priced Netburner development kit is about $100 and includes a C/C++ development environment, TCP/IP stack and webserver code. You probably could port Plan13 or even SGP4/SDP4 subroutines developed by others to the Netburners.

In any case, I'd look forward to hearing more about your plans as they progress. I don't have a lot of time for sats but I've always been interested in seeing what others are up to.

 

[Clive Wallis]

Hi,

I came across this old thread while browsing for something else

Just wonder whether you have done anything more on your project?

I think it should be fairly easy to do the first part of your project, ie to control the antenna azimuth & elevation, and display the angles on an LCD display.

I have a tracker designed by Howard Long G6LVB, which does this, and is driven by my tracking program (SAT_Explorer) to control my G-5500. Details are available on the web, including C source code.

More difficult, would be calculating the tracking angles from the Keplers, although IIRC this has been done with PICs.

Cheers

Clive G3CWV

 

[hippy]

For tracking, I would personally design the antenna controller separate to the tracking instructor so that could be a PC to start with a PICAXE later. Although two chips it's not a lot more and lets you keep the antenna controller as is and working while developing the tracking instructor.

The umFPU interface can be a bit complicated. I've not done a lot with it but I found one way to simplify things was to create my own library as PICAXE subroutines and then just call them as required ...

w1 = ?
w2 = ?
Gosub W0_becomes_W1_times_W2
SerTxd( "Result is ", #w0 )

etc.

 

[brucerob]

Single-chip satellite position processing

As a follow-up to this rather old thread, I'd like to point out that I've ported the Plan13 algorithm to the Arduino/ATMega platform. The library, freely available at http://code.google.com/p/qrptracker/, will provide az/el lat/long and doppler-shifted frequency when provided with keplerian elements, location, frequency and time. Included also is a means of storing and accessing 13 keplerian elements in the EEPROM of the AtMega328. I've made a website to discuss the idea at http://sites.google.com/site/qrptracker, and you'll find an introductory video there, as well.

I'm posting all this to a picaxe board because I'd enjoy collaborating with someone who has a picaxe tracker and would like to have one of these on-board chips do the satellite position calculations, rather than a computer. My chip could output easycomm lines on a serial port, for instance. You wouldn't need any Arduino equipment: I'd just send you a chip with your favorite keps on board. Eventually the keplerian elements would get rusty, but it might be fun for a Field Day exercise or whatever.

I'd also like to say that I've found your conversations on this forum amazingly helpful in my last two years of having fun with the picaxe. I don't believe I've ever posted before, simply because I've always been able to find the answers I need by searching! And, of course, working with the picaxe was a wonderful way to get used to the concepts that helped me do the Arduino project.

Finally, I'm working on a 28x2-based robot for Christmas for my son. You can see a google sketchup version of it at my blog, http://ve9qrp.blogspot.com/2009/10/robot-plans-in-google-sketchup.html.

Regards,
Bruce