In the quest for round stars I recently had an idea I wanted to test. Ihave noticed that even after careful drift alignment my LX200 has a residual drift in RA. Several posts on MAPUG echo this observation. This drift occurs after periodic error correction is used. Whether the drift is caused by backlash effecting the PEC training process by 'Swallowing corrections' resulting in an unequal number of east vs west corrections, A quirk in the Meade PEC implementation, or that that the PEC training is done with a less than perfect alignment I do not know. I feel that it probably isn't the latter because nearly everyone reporting the problem also reports almost exactly the same amount of drift.
For several months now I have known that my roundest star images lie somewhere between 60.3 and 60.4hz. The LX200 only allows the tracking rate to be set to the nearest 0.1hz. Contrary to the 59.4hz to 60.1hz adjustment range stated in the manual I discovered the actual range of adjustments available is 55.0-65.0hz.
The idea I had was over a short period of time to select back and forth between the closest available rates with a duty cycle that would give the correct average rate. The math for this dithering turns out to be really nice, I think it comes from something I vaguely remember from school called partial fractions. Basically:
Fout=( f1*t1 + f2*t2) / (t1+t2)
Here is an example: 1 sec of 60.1hz=60.1 cycles ; 9secs of 60.0hz = 540 cycles ; total number of cycles =600.1 ; total time = 10s ; average freq = 60.01hz
I wrote a program to experiment with how doable this is and have made 1 test to date.
|
This is a screenshot of a program I wrote to test this idea. It talks to ACP and in turn the telescope via ASCOM and steps the tracking rate between the nearest two available on the scope (F1,F2 ) with times (T1,T2 ) required to get the desired rate. The times T1 and T2 sum to 10000 msec which is the cycle time over which the frequencies are averaged. |
|
|
|
|
|
|
| 60.20hz 8 minutes | 60.30hz 8 minutes | 60.33hz 8 minutes | 60.35hz 8 minutes | 60.37hz 8 minutes | 60.40hz 8 minutes |
Note without the program I would have had to choose either 60.3hz or 60.4hz. As you can see 60.35hz worked best. I will as they say continue the research to see if this is consistant.
