Interpreting PrecisionPEC's Results

PrecisionPEC can be used to very closely determine the periodic error for your mount, provided the data is interpreted correctly. After you've collected your tracking data and loaded the tracking log file into PrecisionPEC, interpreting PrecisionPEC's graphs is essential to determine the actual periodic error of your mount.

Without careful analysis of the graphical data, it is possible to come to the wrong conclusion about the magnitude of your periodic error. The example below demonstrates one scenario.

What is My Mount's Actual (Uncorrected) Periodic Error?

In the example below, a Paramount ME customer dutifully collected the tracking log, used the correct image scale for his system, and then loaded the log file into PrecisionPEC. The screen below shows the initial Graph of Periodic Error.

Importance of the Image Scale Parameter

Always double-check what you entered for the image scale parameter! The magnitude of the periodic error graphed by PrecisionPEC, in arcseconds, is entirely dependent upon the image scale parameter.

Example of a "Bad Image Scale"

Let's suppose your mount actually has 3 arcseconds, peak to peak, uncorrected periodic error (which is typical).

If you incorrectly enter an image scale of 1.00 for an optical system that actually has an image scale of 0.5 arcseconds per pixel, then PrecisionPEC's Fit graph will indicate that your mount's periodic error is 6 arcseconds peak to peak.

Further, if you were to continue on and upload this incorrect Final Curve to the Paramount ME, then the mount's internal periodic error correction algorithms will be too large (over correcting) by a factor of two!

So, please take great care when entering your image scale!

Coming to the Wrong Conclusion

"Looking at the raw tracking data, my Paramount ME appears to have +/-7.6 (15.2 peak to peak) arcseconds of periodic error!"

If this conclusion was correct (which it is not) then this mount's periodic error would be more than three times larger than the Paramount ME's specification of "7 arcsecond peak-to-peak periodic error or less". So, what's so terribly wrong with the above conclusion?

Let's start by analyzing PrecsionPEC's Graph of Periodic Error.

The x-axis represents time.
The y-axis represents the number of pixels of offset during tracking in right ascension.
The text labels "pixels" and "arcsecs" represent the minimum andmaximum values of the current graph, and do not represent the actual periodic error of your mount!
The "arcseconds" label (next to the "pixels" label) represents the graph's minimum and maximum number of arcseconds, (computed using the specified Image Scale parameter).
Each light blue line represents an "auto-scaled" division for the current graph.

Notice that the raw data shows a progressive drift (downward in this example) over time. This is normal during unguided tracking, and can be caused by polar misalignment, refraction, tube flexure or other sources of error.

The graph showing the "raw" data has not been "normalized" to extract the repeated periodic error (cyclic) within the raw data. The minimum and maximum values on the graph are automatically scaled to show all of the raw data.

No conclusions about the mount's periodic error can be made using only the un-normalized Raw data.

There will always be a drift component in the raw data. PrecisionPEC's Data | Fit command extracts the periodic error from the raw data, excluding the drift.

Click Data | Fit to determine the Fitted Periodic Error Curve for the tracking log data. When this command is complete, the Graph of Periodic Error shows the normalized raw data (jagged blue line), and the fitted periodic error correction curve (the smooth, purple line in the graph below).

What is the actual peak to peak periodic error for this mount? Use the Y axis Zoom Tool to determine the minimum and maximum periodic error for this curve.

Clicking the up arrow on the Y axis (about 8 times in this example) until the fitted curve just touches the top and bottom of the graph shows that the minimum and maximum pixel values (or amplitude of the curve) as +/-1.2 pixels (or +/- 1.6 arcseconds).

Careful inspection of the graph shows that the magnitude of the actual, uncorrected periodic error is very near +/-1.5 arcseconds of periodic error, or 3 arcseconds peak to peak periodic error.

In this example, the mount's uncorrected periodic error meets the Paramount ME's specification.

What Next?

We now need to take the Fitted Curve and compute the Final Curve. The Final Curve represents the optimal periodic error correction curve for your mount.

Clicking the Final option graphs the periodic error correction curve that can be loaded into the Paramount ME firmware's permanent PEC table. Note that curve represents one cycle of the worm (whereas the earlier graphs display multiple cycles of the worm) that is used to offset or compensate for the mount's periodic error during tracking.

The scale of this graph is not displayed by design. The "stair step" portions represent the smallest possible movement that the Paramount ME's control system can make when tracking.

After establishing communication with the Paramount ME through TheSky6's Telescope | Link | Establish command, from PrecisionPEC, click Data | Upload to Paramount. This permanently saves the PEC curve to the Paramount ME's internal PEC table (in flash RAM).

To verify that the PEC curve has been successfully loaded, from TheSky6, click Telescope | Options | More Settings, and then click on the Precision Error Correction tab to display the mount's periodic error correction curve.

The PEC curve in TheSky6 should look similar to the PEC curve in Precision PEC. Again, the number displayed next to the arcseconds text represents the minimum and maximum for this graph.