I borrowed a C-64 from John Perkins for several months until I bought my own. Ray Thompson who does variable star observing with an Optec SSP-3 had a C-64 program which captures the data from the SSP-3 via the user port on the C-64. The variable star program was a good base to start developing a "high speed" data capture routine. It already contained a clock setting procedure and an assembler routine to accumulate the counts from the SSP-3. In order to get a linear relationship between the accumulation period and the programmed "rate" setting, the final basic program had to be run from a compiled version rather than using the interpreted version. (Thanks go to John Morriss of Toronto for assisting with this and other C-64 advice.) The fastest effective "rate" achieved by this program is about 20 counts per second. Although this is hardly "high speed" (compared to a professional setup) it is more than adequate for the Galilean mutual events. We used a rate of roughly 5 per second,although 1 per second would have been fine according to Dr. Franklin.

In order to obtain observations of scientific value we had to synchronize the internal clock to within 0.1 second accuracy. The C-64 clock records time to 1 tenth of second, so what we needed was a way to trigger the clock to start at the moment a time signal minute pulse is received. I enlisted the aid of Bob Sandness who was able to construct a circuit to isolate the minute pulse from WWV in Colorado which is a 1000 hertz tone. (I believe WWV Hawaii is 1500 hertz). Bob designed and built a two part device which includes an active filter and a phase locked loop (PLL) tone decoder which detects the 1000 hertz tone and signals the computer through a pin on the user port (see figure 1). The program logic was designed to look for the synchronizing pulse over a short two second interval triggered by the operator. The operator anticipates the upcoming minute mark and "arms" the program through the keyboard. The program then confirms receipt of a synchronizing pulse. Dan Driscoll who has both an SSP-3 and a C-64 was instrumental in getting this whole combination to work together. Dan also wrote a data graphing program for the C-64 so we could view our results before converting them to IBM PC format. All this high tech equipment would have been of little use without a telescope that could track accurately and keep the glare of Jupiter from entering the photometer's detector (and swamping the signal). Fortunately, Bob Sandness graciously allowed Dan and I to hookup this gear to his Celestron 14 with Byers drive (1 - 2" periodic error!) that more than did the job (see the sidebar about the C14 and the SSP-3). We observed our first mutual event (an eclipse) on March 17. However, on March 31st we observed an occultation and an eclipse which produced much better results due to the superior sky conditions and the WWV time synchronization achieved on the latter night.

The results shown in the following graphs were produced on an IBM compatible PC. The C-64 data files (containing about 5000 numbers per observation) were converted to IBM PC format by Wayne Hayes. Figures two and four show the entire light curves collected while figures three and five are close-up views of the events themselves.

The digital data were sent along with observation and equipment notes to Dr. Franklin to be used along with the results of other groups. Their objective was to improve the orbital elements of Io in time for the arrival of the trouble-prone Galileo spacecraft.

In future we may use this technique to observe asteroid occultations or possibly grazing occultations of bright stars by a very thin crescent moon. As the cost of used laptop computers continues to drop we may consider stepping up to a more portable system and eliminate the data conversion issue.

FIGURE ONE: Clock synchronizing technique for the Commodore 64 computer.

Go to Top

Partial Occultation of Io by Europa

• The Blue filter was used by accident instead of the Visual filter.
• The C-64 Clock was synchronized using a PLL circuit feeding a start pulse into the User port on the WWV Minute pulse. This was done only a few minutes prior to this run. Several time checks were made later (see the "Time Checks" sidebar).
• Sky conditions were excellent. The temperature was about -5oC but dropped several degrees over the night.
• Subjects in order of observation: 30 seconds dark current; 60 seconds of sky O .. x (x = photometer, .. = Io and Europa); the event, along with one brief dark current to check tracking, the sky again O .. x ; a long stretch of dark current while moving to the opposite side of Jupiter; and finally the sky on the other side of Jupiter x O ..

Partial Eclipse of Io by Europa

• Order of observation: dark current, sky opposite, dark current, sky at 90o (below), the event, dark current, sky opposite, dark current, sky at 90 (below).
• Note that the time checks made after the run indicate some drift of the clock, making it difficult to obtain a precise time for the mid point of the event (see the "Time Checks" sidebar on this page.

Figure two: Light curve showing eclipse of Io by Europa.

Figure three: Light curve showing occulatation of Io by Europa.

Go to Top

Two mutual satellite events were observed on March 31,1991. The 5000 measurements collected for each event were used to prepare the light curves accompanying this article. As can be seen below, all of the measurements were collected in little more than half an hour of observing time!

1. Partial Occultation of IO by Europa:

  
     Start           3:25:00.1 UTC  
     End             3:41:15.5 UTC  
     Duration        16:15.4 = 975.4 seconds  
     Data points     5000  
     Effective Rate  5.126 counts/sec  

2. Partial Eclipse of Io by Europa:

  
     Start           5:15:01.7 UTC  
     End             5:31:16.8 UTC  
     Duration        16:15.1 = 975.1 seconds  
     Data points     5000  
     Effective Rate  5.128 counts/sec   

Go to Top

OPTEC gives a formula for image scale as:

  
     s =  (pi) f / 180           (1)  

  
     fov = 3600a/s               (2)  

The Byers drive accuracy was so fine that no guiding corrections were made during actual observations. Occasionally the flip mirror was used to verify the tracking. The periodic error is on the order of 1 to 2 arc seconds based on CCD images Bob has taken with an SBIG ST-4 CCD camera. Bob says the declination drift over the 15 minute observation runs would have been not much more than 7 arc seconds.

Go to Top

As the time checks made after the first observation run seemed reasonable , I unfortunately decided not to re-synchronize prior to the eclipse observation. The time checks after the last event show that the clock was not keeping consistent time , I suspect due to the changing temperature. The response by the PPL circuit to a WWV pulse (when signal strength is good) is much faster that the 0.1 second clock interval.

  
   WWV Minute       C64 Clock  
   Pulse (UTC)      Reading        Drift  
  
   4:27:00.0  UTC   4:26:59.9            + 0.1  
   4:28:00.0  UTC   4:28:00.1            - 0.1  
   4:29:00.0  UTC   4:28:59.9            + 0.1  
   4:30:00.0  UTC   4:29:59.9            + 0.1  
   4:31:00.0  UTC   4:30:59.9            + 0.1  
   5:54:00.0  UTC   5:54:01.3            - 1.3  
   5:55:00.0  UTC   5:55:00.7            - 0.7  
   5:56:00.0  UTC   5:56:01.1            - 1.1  
   5:57:00.0  UTC   5:57:01.1            - 1.1  
   5:58:00.0  UTC   pulse not received   n/a  
   5:59:00.0  UTC   pulse not received   n/a  
   6:00:00.0  UTC   pulse not received   n/a  
   6:01:00.0  UTC   6:01:00.7            - 0.7  
   6:02:00.0  UTC   6:02:00.7            - 0.7  
   6:03:00.0  UTC   6:03:00.7            - 0.7  

Go to Top