A few weeks back I had the DIY Sky Quality Meter demo setup working (see here). My primary target in building a Sky Quality Meter was to have a complete all-in-one ambient conditions logging device. So to finalize this project, I added a micro SD-Card reader, real-time clock, barometric pressure sensor and a rechargeable battery. I could manage to squeeze all the code into an Arduino nano :-). With the small footprint of the Arduino nano, I could build a case box (3D-printed) with only 123x68x34mm external dimensions.
One essential part of the case is a chamber for the TAOS TSL237S sensor. The status LEDs of the Arduino did alter darkness readings severely. So any light apart from the night sky has to be shielded from the sensor!
Furthermore, the sensor requires a IR-block and color correction filter, to work comparable to the Unihedron Sky Quality Meter. As described here, Unihedron uses a HOYA CM-500 filter. I could find an almost identical filter, which is now included in the case as front cover of the sensor chamber.
My Arduino code may lack some fine tuning (forgive me, but I will not publish my source code. It would not be fair to Unihedron, who had all the development to build the original SQM device!). But the sensor readings are comparable to the second fraction digit in most cases to the SQM unit I could use for testing. This is sufficiently precise to me. I do refresh all the values (darkness, sensor frequency reading, temperature, humidity, pressure, dew point, calculated altitude, battery voltage, presence of SD card as well as time and date) every 5 seconds. This is someway insane, as a refresh rate of 30 to 60 seconds would still be very high 😉
With a freshly charged 18650 battery, my device may record for more than 50 hours. So even a weekend trip would be no issue.
For all curious folks out there, this is the parts list:
– Arduino nano v3
– DS1307 RTC module
– BME280 – temperature, humidity and pressure module
– Micro-SD Card interface
– 1.3″ OLED Display (128×64 Pixel)
– 18650 Lithium battery carrier with charger and 5V output
– TAOS TSL237S sensor
– 8mm UV-IR Cut filter
– 3D printed case
And this is the device:
A few times in the past I have seen charts of the dark sky quality for astronomical use. Especially during a visit of the University astronomy department I had the chance to talk with one of the facilities operators. This made me curious about how they quantify the night sky quality.
Upon further research I found the widely used Sky Quality Meter by Unihedron. It is a small box measuring the sky brightness by means of a light sensor with corresponding frequency output. When I could find a distributor selling these sensors for a good price, I was up to building my own one.
My concept was to use a arduino style micro controller and build a sky quality meter together with weather data (temperature, pressure, humidity) recording device. The device should save every 10-30 seconds the data collected. And it should last for at least one whole night running on a small(er) battery like AA or 18650 lithium type.
The hardware part was quite easy to accomplish. I simply had to connect a couple of wires from the micro controller to the sensors and other components. The harder part was to create a proper software, fulfilling all my needs. I had to find a way to
1) cope with the frequency range of 0.01Hz to 1MHz
2) fit all the code within the small memory
3) calculate sky quality and calibrate device to a reference device
Luckily, apart from a reasonably well described device on the Unihedron web-site, I could borrow one Sky Quality Meter from a fellow astronomy club member. So I had a reference device to compare and calibrate my device.
Here is the test setup. The readings are already very close to the original device!
After several hours of designing, 3d printing, drilling, soldering and assembling, my camera slider is ready to use! All components are neatly packed within cases, so that the only wires visible are the power supply from the LiPo battery pack and the shutter release cable to the camera. I have several extensions in mind, like a pan-tilt unit. But for the moment, I will test and use the setup as is. The next improvements will be more on the firmware, for more features: non-linear movement, slow start, slow finish, pre-defined profiles, … for more impressive movies.
Now, concluding the build, I have the following parts in the final setup:
- 12 U-groove wheels, matching the carbon fiber tubes (3D printed plus ball bearings)
- 4 wheel carriages, each holding 3 U-groove wheels (3D printed)
- 1 case for the microcontroller and motor driver (3D printed)
- 1 case for shutter release, power supply and connectors (3D printed)
- 1 hand-controller case (3D printed)
- 1 battery bracket (3D printed)
- 2 end-assemblies (5 parts each, 3D printed)
- 2 supports (3 parts each, 3D printed)
- 1 NEMA17 stepper motor
- 1 A4988 stepper motor driver
- 1 Arduino Nano (Atmel 328 microcontroller)
- 1 4×20 LCD
- 1 Joystick module
- 1 DC-DC converter
- 2 micro switches with long lever as end switches
- GT2 10mm timing belt with wire reinforcement
- 1 GT2 10mm 20 teeth pulley wheel
- 4 GT2 10mm idler wheels
- A whole bunch of M3 to M6 screws and nuts
- 40x3mm flat aluminium
- 30x50x3mm L-shaped aluminium
- carbon fiber tubes
The project was really a fun to make. Even more, the resulting slider provides flexibility and transportability! I may configure the slider in any length, depending on the available tubes. The tubes I use, are 37cm in length and have aluminium screw-in adapters to fit one to the other. I have a bag, which I used as personal item in air travel. The bag holds the complete setup for up to 5 meters (exkluding tripods). The bag weighs in at approximately 5kg – so it is a light weight setup for the length possible.
At night, with long(er) exposures, the slider concept to stop for each frame, proofs to be perfect!
A test at night with 10 second single exposures:
Crisp and clear images due to stop-motion. But not yet fluid motion (there seems to be a bug in the motor control).
To be continued 🙂
After several days of building and coding, the slider was ready to perform a first real world test. Even if the electronics have not yet been encased (all the wires are going wild around the platform), this was proof of concept…
For quite some time I would like to create night time movies in hyperlapses. For me, the most stunning results may be created by moving the camera along a linear path by the use of motorized sliders. Motorized sliders, which are more than 2 meters long, have an impressive price tag. Further more, these tools are bulky and heavy, especially when the setup attached to weighs in a few kilos.
Therefore I decided to build my own with a few goals in mind:
- light weight
- variable length
- suitable for a load of a few kilos
- wider range of speeds
- extendable for rotation axis
- direct control for camera(s)
To achieve all or most of these goals, I came up with a design built around carbon fiber tubes with aluminum screw-in fixtures. Appropriate tubes may be built from scratch or are readily available for camera gimbals. I chose the camera gimbal extensions, as there is no big price difference to buying stock material. Further more, they come in a handy size of +/- 40cm in length.
The end supports will have to hold the tubes as well as a gear belt, along which the slider cart will be driven. For long setups, I created supports, to prevent bending and excessive stress to the tubes. Both types of support will have legs as well as tripod mount screw holes (3/8 UNC thread)
The slider cart consists of 4 blocks holding 3 pulley wheels each. The blocks are attached to a base plate (in test setup a plywood sheet). In the middle of the base plate lies the motor unit consisting of a steper motor and 4 guiding wheels to create enough tension for the gear belt to be driven by the motor.
All in all, the shopping list is really limited, as most parts were 3D-printed. What I had to purchase or use (most parts were already to be found in the workshop) was:
- carbon fiber rods (at least 8)
- 24 ball bearings type 626 2RS (6x19x6 mm)
- GT2x10mm belt matching the desired length
- GT2 20 tooth drive gear
- 4 guide wheels without teeth for 10mm belt
- 1 NEMA 14 stepper motor, <3V nominal voltage
- several M5 and M6 screws, washers and nuts
- 3/8 UNC thread taper
- approximately 0.5m of 40x3mm Aluminum sheet
- 25cm of 30x50x3mm Aluminum L shaped profile
- Arduino, Stepper motor controller like A4988, 12-18V (lithium) battery
- 1 can of rubber spray like Plasti-Dip (c)
Most of the time I spent was in CAD constructing the parts. Printing took about 3 days. The pulley wheels have to be sanded for a smooth surface before coating with rubber. The remaining time was spent in cutting, drilling and tapering the aluminum parts, before all parts could be attached together.
The first test run was more than pleasing. See for yourself:
The next thing to do is to create a control box with all the features implemented for every day use 🙂