One night last week promised almost ideal imaging conditions after nearly two months of dust, wind, jet stream, and cloud cover woes. Aside from the first-quarter Moon, the jet stream was calm, and for the first time in nearly eight weeks, it looked like there would be no cloud cover at any altitude, and atmospheric “seeing” would be near perfect. Imagine my excitement to finally get a good night’s imaging done, only to find that for some reason, focusing failed no matter what I tried. Even with manual adjustments of the focuser motor position, all I got were blurry little doughnuts on my test images.

2022/04/07 00:04:58 349 - CRITICAL  - [RoboFireLocalField           ] - [GENERIC_ACQUIRE_STAR_Code                    ] - Find Field Stars : Found 2 stars. Number of stars found is lower than requested of 5 [KO]
2022/04/07 00:04:58 352 - INFO      - [RoboFireLocalField           ] - [GENERIC_FOCUSER_GOTO_Code                    ] - LocalField VCurve AutoFocus Error : Return Previous Position
[SENDEVENTO_VOYAGER_Signal                    ] - [S00021] SEND EVENT => Signal = RUN_FOCUS
2022/04/07 00:05:00 554 - CRITICAL  - [Focus                        ] - [Funzione                                     ] - Action Aborted for Error : Focus Async Error (Error executing LocalField VCurve AutoFocus : Find Field Stars Error, maximum retries reached ... action failed)
2022/04/07 00:05:00 558 - EVENTO    - [ServerClientWebSocket        ] - [SENDEVENTO_VOYAGER_Signal                    ] - [S00021] SEND EVENT => Signal = FOCUS_ERROR

The imaging automation software’s automated intelligence focusing routine could not identify enough stars to lock onto and create a local field measurement image to use during the focusing process. The AI evaluates the stars in the local field image to determine the star brightness compared with those stars’ known brightness values. However, if the AI cannot lock onto the stars correctly, it cannot do an accurate “plate solve” to match the star coordinates to a known star database. If it does not know what stars it is looking at, it cannot determine the expected brightness values for those stars. Therefore it cannot tell if the focus is at its optimal level by adjusting the focuser position to get the brightness within accepted tolerances. Visually, I could get it pretty close to a point where the stars looked like single pinpoints. When the software took over to fine-tune the focus to the most statistically optimal position, it would fail, as seen in the log transcript above.

At first, I thought something was loose on the imaging train, that perhaps the focuser stepper motor was not moving the focuser precisely enough. It is attached to the metal focuser housing by a 3D printed adapter, and I was concerned that it was perhaps because of the admittedly hard plastic that there might be an issue. However, Dave at Pixelskies confirmed that everything was good, so not the adapter bracket then.

I use a Pegasus Ultimate Powerbox for power distribution and control on the telescope, and it has an environmental sensor that measures ambient temperature and relative humidity of the air. The answer came when I looked at the measurement graphs, and it became clear that the likely cause of the problem might be dew forming on one or both of the mirrors of the optical tube assembly (OTA). Luckily the Pegasus Powerbox has built-in controls for dew heater collars. Still, since I have not run into problems with dew yet since moving my telescope to the observatory in Spain, I did not buy suitable dew heater collars for the OTA that could connect to the Pegasus Powerbox.

Finding a heater element for the secondary mirror assembly was also problematic because most suppliers in Europe only carry secondary heater strips for up to 10-inch Ritchey Chretien OTAs. The 12-inch OTA is just too large, and I confirmed that the 10-inch heater strip would not fit with a few suppliers. Luckily I found a solution with Kendric Astro Instruments, who recently released a new version of their secondary heater collars that works better with a larger RC OTA. The new collar design was not available on the market until a few weeks ago, so I had not addressed the possibility of a dew problem on the secondary mirror yet. I ordered the new collar, but it might take time to arrive in Spain since it is shipping from Canada.

That leaves only the primary mirror to address. Since I do not want to introduce anything bulky inside the OTA that could cause image quality problems with light reflections or interfere with the field of view, I opted to get an extra heating collar to go outside the primary mirror housing. I hope it would be sufficient to create a significant temperature differential in the lower part of the OTA to keep the ambient air temperature in contact with the primary mirror just above the dew point when necessary. We’ll have to wait and see if this works effectively enough. Running the primary mirror fans might also help move air through the OTA and reduce the probability of dew forming. Still, I will have to be careful that it does not introduce vibrations that could affect my images.

As I mentioned before, the Pegasus Powerbox has an environment sensor to measure temperature and relative air humidity. However, it only has one sensor probe for this, and I would like to get an idea of what is happening inside the OTA. It has been some time since I tinkered with Arduino, but I plan to get an Arduino Uno and attach two Grove SHT31 Temperature & Humidity Sensors to it. The hope is that the highly accurate SHT31 sensor will be small enough to slip inside the edge of the primary mirror housing of the OTA so that it does not interfere with the light path. Attaching the other sensor outside the OTA, close to the one on the inside, should give me a differential comparison of temperature and humidity near the mirrors. Then it is relatively trivial to write a python script on the imaging computer that will send a Telegram message to my phone when dew might cause potential problems. I can then log onto the Pegasus Powerbox to ensure that it is adjusting heating collar temperatures to compensate effectively. At least, this is my geeky plan; we’ll see how well it works in practice. Luckily the Arduino components are not expensive to play with and find out.

Improved guiding statistics after training the mount for PPEC

Despite not being able to focus the primary imaging camera, one good thing was that I could spend time using the near-perfect “seeing” conditions to run several tracking periods with my mount to analyse periodic errors in the worm gear. My guide scope did not appear to have a problem with dew to the extent it affected guiding. Using a software tool called PEMPro, I generated a reliable correction curve over several worm gear periods for my mount. Using this correction curve with my mount’s permanent periodic error correction (PPEC) function drastically improved my guiding accuracy. It took me several hours to do this since I had not used PEMPro before and needed to figure a few things out, but the final improvement result in the guiding statistics looked very promising. Until now, I have not been able to get right-ascension guiding error rates below 0.5 arc-seconds reliably, so 0.26 arc-seconds marks a new best for my mount. I think I can still improve on this somewhat, but I am pleased with it for the time being.

This hobby always brings challenges with it. It is frustrating at times, like last week when one environment variable (the dew) ruined near-perfect imaging conditions. To add insult to injury, cloud cover and jet stream activity are back up again this week, though it seems like Friday and Saturday nights might be worthwhile (if dew doesn’t scuttle everything again). At least one of the new heat collars should arrive this week in time for the coming weekend.