Scopebox: a telescope controller
I have a telescope. Unlike most telescopes, mine is not computerized. No motors to track the stars or slew to an object punched into a keypad, not even shaft encoders (aka “digital setting circles”) to read out the current position. Finding objects is a full-manual experience, involving the Telrad sight and/or the 10x finder scope with lots of switching back and forth between the finder and a star chart. This process is called “star hopping,” and I wouldn't have had it any other way when I first started out. It's pretty inefficient if your goal is to check off the most galaxies/nebulae/etc. possible in the shortest time, but it's a great way to get to know the sky.
I've been doing astronomy like this for ten years, and finally decided it might be time to join the 21st century and add some encoders and a computer. (This is fairly straightforward on my Dobsonian telescope, whereas adding a motor drive system is not.) I think the tipping point came when I weighed my finder scope and my Nexus 10 tablet, and found that the tablet weighed less: I could replace the finder with the tablet, right up at the eyepiece.
This being a home-made telescope, and me being a EE, of course I had a list of requirements no commercial product could quite match:
A wireless digital-setting-circle interface: altitude and azimuth optical encoders, with a bluetooth link to apps on the tablet. You can buy this off the shelf, but my telescope has some mechanical imperfections (axes not perpendicular), and with access to the code I could correct these in software.
Fan control. My telescope has a fan behind the mirror to quickly bring it into thermal equilibrium, and fans blowing across the mirror to mix the boundary layer there. These are controlled with toggle switches, but with motion awareness the new system should turn them on and off automatically as desired during times of telescope use and inactivity.
Battery management. My old, heavy sealed-lead-acid battery is wearing out, and the electrical cable dragging on the ground is a continual hazard. This could be replaced with a much lighter lithium-ion pack, and the whole thing mounted on the telescope rocker box. In fact, it would be really convenient to have the charger built in so I can just plug the telescope into a 12V source to charge.
I decided to stay with a 3.7V, parallel-connected battery pack. The fans need up to 12V which could be supplied by a boost converter or four cells in series. I don't like worrying about cell balance for series-connected packs and most everything else in the system needs a low voltage, so I opted for the boost converter (see below).
The battery consists of ten 18650 li-ion cells in parallel. The number was chosen somewhat arbitrarily to fit into the plastic enclosure I had available. My cells are LG high-capacity types with an unusual 4.35V charge voltage. The total capacity should be at least 100 Wh, enough to run everything for several nights of observing. I did not add a separate protection circuit. The 18650 cells have basic short-circuit protection (PTC thermistor) and the battery manager and microcontroller will protect against over- and under-voltage conditions.
The battery-management IC is a bq24193 from Texas Instruments. It can act as a standard switch-mode battery charger with inputs up to 17V, or run in boost mode to supply 5V at up to 1.3A for USB OTG applications. The scopebox circuit takes advantage of this to supply 5V to the optical encoders and to a USB standard-A receptacle, which serves as a USB power outlet for a tablet or cell phone. When a [potentially much higher voltage] DC supply is plugged in for charging, the switch in the DC power connector CONN2 turns off MOSFET Q1, isolating the 5V rail from the charger voltage. (The digital setting circle functionality is disabled while charging.)
The bq24193 is highly configurable over its I2C interface, permitting the use of my high-voltage li-ion cells, setting precharge and termination currents, IR drop compensation, and so forth. PD2 on the AVR registers faults and status changes on the bq24193 INT output, and PD5 toggles the switch-mode circuit between buck (charge) and boost (5V OTG—normal scopebox operation) modes.
The unregulated VSYS power rail follows the battery voltage and supplies the microcontroller, the fan-voltage boost circuit, one-half of the level converter, and the bluetooth module, the latter through a 3.3V LDO.
An LTC3122 boost circuit provides an adjustable 6-12V at up to 500 mA for the fans. The same voltage is supplied to one or the other or both fans via the dual MOSFET U103. A high-frequency PWM signal from the AVR is low-pass filtered to around 1 Hz and injected into the LTC3122 feedback network to adjust the output voltage. This is useful to lower the fan speeds, which permits leaving them on while observing; at higher speeds their vibrations disturb the view.
Everything fits onto a 2"x3" PCB.
- gEDA design (gschem schematic and pcb layout): scopebox_hardware.tar.gz
- AVR Eclipse C project: scopebox_code.tar.gz
This work is licensed under a Creative Commons Attribution-ShareAlike 3.0 Unported License.