Schaller Observatory (84 W, 42 N) is dedicated to high speed stellar photometry of variable stars and Optical SETI (Search for Extra Terrestial Intelligence). The detector, amplifier and acquisition equipment are able to detect amplitude variations in a photon signal data stream at rates up to 22 KHz. The astronomical challenge is acquiring stellar photons at sufficient data stream rates to make the high frequency variability measurement meaningful. Using a moderately sized scope, the study becomes limited to a small number of very bright stars.
This site describes the high speed photometry equipment, techniques, analysis tools and results.
The main telescope at Schaller Observatory is a Dobsonian Reflector with a 55.9 cm (22 in) diameter primary mirror. The telescope is computer controlled using stepper motors. The telescope's automation was per plans published on the WWW by Mel Bartels: Motorize Your Telescope. I found the plans easy to follow and well done. Here are a few tips, hints and photos from my incarnation of the plans: Schaller Observatory Telescope.
The equipment is housed in an observatory building with a 10 foot diameter dome roof. The observatory dome was purchased as a kit from: Technical Innovations. Here are a few tips, hints and photos describing the construction and use of the observatory: Observatory Information.
The stellar photons are detected with a photomultiplier tube (PMT) at the focus of the telescope. The PMT is an RCA-7764, and it is generally operated with -1000 V bias voltage. Additional information about selecting and using a PMT detector is available here: Photomultiplier Tubes.
The output current from the photomultiplier tube is converted to a voltage signal by the Front End Amplifier (FEA). The FEA is a special form of a current-to-voltage converter. The amplifier is special because it amplifies incredibly small currents while preserving the signal's high frequency fidelity (<50 KHz). The FEA sensitivity is 0.2 Volts per Nanoampere, with a 200 Megaohm feedback resistor. Additional information about the FEA, including a circuit diagram, is here: Front End Amplifier (FEA)
Terrestrial noise sources are ubiquitous, especially 60 Hz AC power. The output of the FEA is filtered twice: First through a passive low pass filter with a 100 KHz cutoff, and second, with a tuned 60 Hertz notch filter. Both filters use off-the-shelf components, as described in detail here: Filters.
The output of the notch filter is routed to the data acquistion system. The data is acquired using off-the-shelf tools for recording sound files in Windows. The files are recorded in 16 bit mono at 44 KHz sampling frequency. The recordings are saved as Wave files: that is with a ".wav" suffix. Wave file format structure is described by a Programmer here: Wave Format
Data analysis is performed using Fast Fourier Transforms (FFT). The Fourier transform is a mathematical technique for calculating spectral information. A typical star recording, lasting 10 seconds at 44 KHZ sampling frequency, contains 440,000 individual data points. The observatory operates a Linux based computer system to run an optimized Fourier Transform code known as FFTW (Fastest Fourier Transform in the West!). The FFT of large wave type files can also be computed (on Window's platforms) by using a commercially available program known as SigView. Additionally, the FFT analysis of incoming starlight data streams is viewed in realtime using Analyzer 2000.
Limits on the Search
The telescope mirror collects photons of starlight from the stellar source. The individual photons have traveled incredible distances, before impinging on the mirror and reflecting to the active surface of the detector. The physical size of the mirror and the visual magnitude of the star are important factors in determining the quantity of photons detected. Photon shot noise is a dominant data feature at these high time resolutions, particularly with fainter stars.
The frequency response fidelity of the PMT, FEA, Low Pass Filter, Notch Filter and Data Acquistion Card has been tested together as a system. The system's fidelity is tested by using a Light Emitting Diode (LED) at the main telescope mirror. The LED is driven by a square wave frequency generator. A square wave is advantageous because the decay of odd numbered upper frequency harmonics is known theoretically. Here is the system's response to a 1000 Hz Square wave:1000 Hz Square Wave Response (Image, 46 Kb). Here is a spectral analyis of the system's response to a 1000 Hz Square wave, with discussion:Spectral Analysis (Image, 46 Kb).
The main telescope is fitted with a 50mm finder scope from University Optics. The finder scope is equipped with a miniature video camera, rated at .05 Lux sensitivity from Resources Un-Ltd. This hardware combination effectively displays stars of the 6th magnitude on the video monitor.
The observing session is started by aligning the scope to a conveniently placed bright star. The audio output of the amplifier chain is monitored with headphones. A bright star impinging on the detector creates a clear increase in the signal's audio level. The telescope is then "locked in" on the stellar coordinates. Finally, the telescope is adjusted slightly ahead of the star's track and the drive system is stopped. The recording is made as the star drifts across the sensitive field of view. This method avoids issues with drive system vibrations and audible stepper motor noise. It is important to minimize ambient noise in the observatory during recordings, because at these extraordinary amplification levels the coaxial cable to the photomultiplier has some microphonic sensitivity.
Additionals stars for observation are selected using a computerized star atlas: Lodestar. The atlas displays objects as seen on the video monitor. Locating the next target then becomes a simple matter of inputting the coordinates and engaging the computer controlled drive. The file's name is the HD designation (eg HD187642.WAV). The file's date/time stamp is the record of when the recording was made (in UTC).
Here is an example of what photons arriving from stars "sound" like. Altair.wav (Sound, 263 Kb) You will notice that it sounds like noise! What did you expect??* This sound file represents only 3 seconds of the total recording, to keep the file a reasonable size. Here is a spectrum analysis of the same wave file : Spectrum Analysis (Image, 50 Kb).
To date (February, 2001) 250 star recordings have been collected, with more being added as weather conditions permit. Each recording (of approximately 10 seconds duration) fits conveniently on a standard 1.44 Megabyte floppy disk.
I will be adding more material to this website soon. I hope you have enjoyed my description of this Hobby Project.
*What I'd like to hear! (Sound, 100Kb).