RE: Sky Scout for observing satellites? Long and slightly OT

From: Bill Frost (
Date: Tue Jun 20 2006 - 22:02:25 EDT

  • Next message: Ted Molczan: "TJM obs of 2006 Jun 21 UTC"

    Brooke Clarke wrote . . .
    Celestron is introducing a new product called the Sky Scout. ... It is a low power hand held telescope that's intended for use to identify whatever you point to.
    Hi Brooke,
    This sounds a fascinating piece of technology and I've been trying to work out how it might do it. According to the manual, you just point - in other words there is no initial state (no calibration, apart from telling the device if you're observing daylight saving or not).
    So as an experiment, I should be able put it in a clamp, orient it in my backyard at 76.1 d (azimuth) and upwards at 47.9 d and insert batteries. After warmup and GPS trilateration, it should tell me that at 10:00 am this morning (21/06/2006 0:00 UT), I am looking at Sirius!
    Pretty impressive and with some software additions it would make for a very nice LEO satellite spotting scope using the device's 'Location' feature. However, I doubt that satellite spotting would be high on Celestron's agenda, given that it would need to educate its users that unless they downloaded TLEs every two weeks or so, the predictions would not be accurate.
    I hope the rest of this post is still on topic - it is my guess how this device might work, based on downloading the fairly sparse user's manual and trying to work out what chipsets you need to integrate to build a $US400 device that does all this.
    Note that everything below this line is pure guesswork - the device is unreleased and I've read no sensible reviews on it:
    In essence the device would consist of:
    *	integrated gravitational and magnetic field sensor
    *	dual axis accelerometer
    *	GPS
    *	embedded SBC (single board computer)
    *	Linux OS
    *	LiOn battery
    The magnetic field sensor uses a gravitational sensor to establish the z axis and then computes x and y values that are accurate to within a tenth of a degree depending on the chipset used. Some of the Honeywell chips are sensitive to a magnetic field of 1 x 10 to the minus 6 tesla. This establishes the observer's azimuth.
    At first I thought Brooke was incorrect referring to an accelerometer, as there is no initial state - ie the observer does not locate a known object, then move the pointer to an unknown object allowing the device to sense the acceleration and the time elapsed. However, I was the one who was wrong. A dual axis accelerometer uses the 0 value to establish the static gravitational field, then calculates the slant angle using a fascinating concept of MEMS (ie the chip is fabricated in the traditional way and a micro electro mechanical system is built in at the same time). The angle of the flap is used to find the slant angle, hence the observer's elevation.
    So far we have two devices probably each the size of an SD card you put into your digital camera.
    The GPS is only used at power on to find the observer's location and establish an accurate system time. To cut costs, the GPS could be an older style SiRF II chipset under the assumption that any user wanting to observe the heavens would avoid foliage, urban canyons or other sources of multipath interference. Thus an external aerial or signal enhancing SiRF III chipset are not required and the chip should easily fit on an Australian 20 cent piece. Once trilateration had occurred and the device clock was synchronised, the GPS could be shut down to save battery.
    The SBC's role is to pool readings from all onboard devices and compute the observer's El Az, project it onto the celestial sphere and find the brightest corresponding object in the FOV. For the 6,000 supported celestial objects a relatively simple lookup would suffice and for the solar system objects, I'm guessing you'd use Meeus' algorithms. The SBC would have USB input, which would presumably double as a means of charging the battery. The SBC would have embedded Linux as the operating system and I guess the integration is done via programs written with a micro C++ compiler, for small footprint. (Unless someone knows of a good Assembly language implementation of SPD4 and Meeus!) If you've seen a laptop memory module (SODIMM), then this is roughly the size we're talking about.
    This setup whould be powerful enough to run LEO predictions and should easily cope with deep space satellites as well.
    If you've read this far, then I can't wait for Kevin Fetter to buy one and give us some feedback! It would be a wonderful way for a new observer to spot an Iridium or ISS or for a dummy like me to track something hard to spot like Suitsat.
    Links to chipsets:
    Compass chipsets
    GPS on a chip
    General MEMS devices
    Subscribe/Unsubscribe info, Frequently Asked Questions, SeeSat-L archive:

    This archive was generated by hypermail 2b29 : Tue Jun 20 2006 - 22:05:50 EDT