Sunday 3 February 2019

The Vacuum Chamber

Further work on TT7B eventually led me to testing sensors with which the tracker was equipped. While the current winter period provided a number of natural test environments for sub-zero temperatures, pressure at ground level stays more or less the same, relatively speaking. The idea of checking how the two sensors intended to measure ambient and gas pressures perform, though, kept entertaining my mind. Several Youtube videos and some time on Ebay later, I decided to build a cheap simple vacuum chamber.
Having seen a number of homebrew designs, the vacuum pump was typically the most expensive part of the setup. I decided to go with a small pump advertised on Ebay as VN-C3 sold for about €22 which was specified to be able to achieve -80kPa at 10l/min. That would give me pressure levels of around 200hPa which in Earth's atmosphere correspond to altitudes of about 12km. Ebay also lists the pump's more expensive cousin VN-C4 specified to -85kPa. The two valves, also from Ebay, were chosen to fit the same tubing as the pump's inlet - 6mm inner diameter. The vacuum gauge was included just to quickly provide rough information about the conditions inside.
Since the objective was to test TT7B, I looked for such a 'chamber' part of the setup that would fit inside itself a whole tracker. The choice, eventually, fell on a 4.25l glass jar generally sold for food preservation. It was 273mm tall and 158mm in diameter. A 100mm screw lid fit the opening at the top.
The idea was to drill three holes through the lid and fixate the two valves and the gauge in with a silicone sealant. Prior to applying the sealant, I filed the metal surface off the lid and the component a bit to help with adhesion. It is not the most robust solution, but it serves its purpose when handled carefully.
The reason for two valves was to be able to allow air in the chamber without the need to disconnect the pump. The chamber can also be sealed once the air is sucked out, and the pump can be disconnected, while the chamber maintains the low pressure.
This is the complete setup. The 12V DC pump is powered and controlled via a bench power supply. The only thing I added was a Schottky flyback diode across the pump's terminals to take care of the inductive kickback when the supply is switched off. When a low pressure level is attained and the supply is turned off, the pump is able to maintain the pressure level even without sealing the valve.
For a test run, I put TT7B equipped with two pressure sensors inside the chamber. The tracker fits in with a temporary wire antenna that can be bent. The data were transmitted in a form of an APRS packet every 2 seconds and received locally with my iGate.
The mechanical gauge eventually settled at -0.07MPa. After that I stopped the pump and left the chamber intact for about 10 minutes before slightly opening the second valve.
The received data can be seen in the chart on the left. From initial pressure of 95300Pa the pump took the insides of the chamber down to 18800Pa in about 5 minutes. Without closing the main valve, the pressure held by the powered off pump crept up by 450Pa in the roughly 10 minute period. The chart on the right then shows pressure vs. altitude according to the International Standard Atmosphere model. The model puts the 200hPa level achievable with this chamber to an altitude of 11.8km. More or less the expected float altitude of a potential flight.

Prostejov: 49.46° 17.13°
min [m] max [m] diff [m]
200hPa 11240 12320 1080
1000hPa -41 328 369
Brownsville, TX: 25.93° -97.48°
min [m] max [m] diff [m]
200hPa 12010 12520 510
1000hPa 23 281 258

To accompany the standard atmosphere model with some real data, I went through a year worth of data from two weather stations that do regular atmospheric soundings, and wrote down the extremes for a 200hPa level and a 1000hPa level - roughly sea level. The first station is quite close to where I am and launches balloons from a latitude of 49.5° in the Northern Hemisphere. The extremes of 2018 for the 200hPa level were roughly 1000m apart. As the second station, I chose Brownsville in Texas which is much closer to the equator at a latitude of 26°. There, the yearly extremes were only about 500m apart with the maximum a couple hundred meters higher than here in the Czech Republic.