Automating the LHiRes III Calibration Lamps

The recent upgrades to the LHiRes III spectrograph include an improved calibration lamp assembly that basically has three positions:

  1. Straight through (normal data acquisition)
  2. Flat Calibration Lamp
  3. Neon-Argon Wavelength Calibration Lamp

There is a multi-pin connector that allows the observer to send 12 volts to the device and to switch to any of the three positions. Normally one would just send 12 volts to the spectrograph and use the switches mounted on the side of the instrument to select the desired mode. But the multi-pin connector was designed with remote operation in mind so I decided to assemble a simple hardware and software solution that would allow me to do so over my home’s wi-fi network.

The first challenge is to get the necessary connectors and to solder a 4-coductor wire to the connectors’ tiny pins. The connectors are “tiny xlr” style occasionally employed in music gear. It was a sore test of my non-existent soldering skills!

The main piece of the system is a model B Raspberry Pi. It’s certainly a bit of overkill as the Pi is a full-blown computer running Denbian – a Unix flavor operating system. But I figure that I may also use the Pi for other automation tasks such as building a dome controller. I found a 4-relay “hat” that fits over the GPIO pins of the Pi, allowing me to switch power to four devices. I use three to select the three positions of the calibration assembly, while the fourth currently switches the power going to the AP1200 telescope mount. I bought a small plastic box and mounted the Raspberry Pi and relay board inside, and installed two connectors, one for power input and the other to send power to the LHiRes III. For about ten dollars I bought a “WiPi” adapter for the Pi which allows control over my home WiFi network. Finally, to make it all work from the warm confines of my upstairs study in the house I wrote a simple Visual Basic program that communicates with the Pi and actuates the switches. It all works great!

The Raspberry Pi board plus 4-relay “hat” installed in a small plastic box. The cable coming from the bottom of the picture is the power in – using the Shelyak-supplied power cable, while the output cable to the LHiRes can be seen at the top of the image. The blue glowing appendage seen inside the box on the left side is the WiPi module.


The first “improvement” I need to make is to mount a switch on the box to allow me to bypass the Raspberry Pi in order to use use the switches on the LHiRes to turn the lamps off and on when I’m in the observatory.


The next task is to create a small class library for the control so that I can incorporate the automation into scripts that can be run from Maxim DL, which is what I use for controlling the cameras and data acquisition.





A view of the small control used to remotely control the calibration lamps on the LHiRes. Since there were four relays I decided to use the extra one to turn the mount on and off. The spectrum is of the star Alnilam the center star in the belt of Orion


Buffy the Lab Assistant checks the work before moving the spectrograph to the telescope






Clearing and cool.  It looks to be an ideal night – lows around 60F, no wind.

Rig #1:

More spectra of the Si lines in Deneb’s spectrum, H-alpha for P-Cyg.  For the U Montreal BRITE campaign.  Most definitely have to get going on reducing all the data from this project.

Rig #2:

A long sequence on GD552, per request of Joe Patterson.  I did a series on this star two nights ago and got some decent data.  In fact the standard deviation in the comp star differential magnitudes was the lowest I’ve ever achieved, around 0.003 magnitudes.  Due in no small measure to having really solid guiding and not having to flip the telescope at the meridian.  At that declination I should be able to track about three hours past the meridian.  So far this year I’ve gotten two nights on this target – which Joe claims is refusing to show any orbital signal:

GD 552 = Cep 1 (no clear periodic signal yet). The latter is sort of baffling – we should be able to find that orbital signal (revealed by spectroscopy, so we know where it should be). It’s a star of enormous significance – probably the oldest CV in the sky – so anything we learn about it would be great. Davud Cejudo has been observing it from Spain; some long runs by USA observers would greatly help to set a stringent limit on periodic signals – or better yet, find one.

Out at 2:30am.