Gas System Main Page
On this page information about the (TPC) gas systems shall be collected.
LPTPC Gas System Manual
A long, long time ago, back in 2004, that is to say in the middle ages, the dark (or golden?) age of FLC-TPC group there was no such thing as a gas system. People would connect a gas bottle via reductor and manual flowmeter to the chamber, carefully trying not to blow it up. They guessed from data what the gas inside the chamber might be doing. In 2005 a fist basic system had been constructed and now one could measure water and oxygen content of the gas, pressure sensors were implemented and the flow could be regulated electronically. In 2006/7 it was begun to migrate the system to DOOCS as a new control system scheme.
The water content is currently measured by means of mirror dew point instruments DP3-D-SH from the swiss company MBW. A mirror is cooled until dew or frost forms on its surface. The temperature where this happens is a measure for the water content, which can be computed from the mirror temperature. The formation of dew or frost is detected by reduction of intensity of light reflected from the mirror.
Mirror Cleaning of the Dew Point Instrument
Water is not the only substance that will condensate on the mirror, other organic substances might do the same. Thus, apart from a clean design of the overall gas system, it is necessary to clean the mirror from those more permanent layers from time to time. The dew point instrument as an automatic cleaning function for that (AMC). For details of its configuration see the instruction manual in the Instruction Manual Filing Cabinet (there are different versions of the instrument!). This function is also treated inside the corresponding DOOCS-Server for the instrument. The instrument will also warn about pollutions by a blinking indicator on its front panel. You can check if everything is still clean by pressing the Standby button on the instrument. The mirror will heat up to room temperature and the indicator should move back into the middle of the red region. If it does not go back completely, follow these steps to clean the mirror:
- Reduce or stop the gas flow.
- Unscrew the mirror holder (big metal cap on front panel). This requires some force, but as well care. The cooling of the mirror will stop.
- Take out the plastic measuring head.
- Have a close look to the mirror, so you might be able to see some pollutions other than water droplets.
- With a clean-room wiper and some Isopropanol very carefully clean the metall mirror and the surrounding plastics. Take care not to scratch the surface, as this will lead to stray light that needs to be compensated by a higher lamp voltage, which leads to a shorter life time of the lamp (90,- € last time I bought one).
- When done replace the measuring head.
- Fasten the metal cap again firmly but not with excessive force. This makes the system gas tight.
- Repeat the Standby check.
If the indicator is in the right-hand region of the red area, everything is very clean and you can even reduce the lamp voltage to increase lamp life. For this carefully adjust the lamp voltage with a small screw driver at the set screw in the vicinity of the indicator. The indicator should be in the middle of the red region.
If the mirror is badly scratched, one is forced to increase the lamp voltage to reach the middle of the red range. However, cleaning is the prefered method to assure normal operation.
- Measurement head of the newer instrument (Inv. 1050344, white arrow marking) drifts up towards room temperature reading within one to a few days. Must be a problem in the very simple measurement head, as the head from the older instrument (Inv. 354261, yellowish arrow marking) works fine.
- Measurements in T2K are very unreliable. It seems that isobutane condensates on the mirror as well and disturbs the measurement by different evaporation behaviour. Result are slow oscillations of several K. The same setup works fine when flushed with Nitrogen or Argon.
- A different measurement principle should be used for water in T2K, without cross-sensitivity to component gases.
Oxygen is measured by semi solid fuel cell devices. It follows from the measuring principle, that the cells are spent during the measuring process. The expected lifetime is 2.598*107 ppmV*h. So after opening the packaging the fuel cell should be exposed to gases with high oxygen content (namely air) only for a short a time as possible. The instruments are calibrated with ambient air (20.9 %V) to cover the entire measurement range.
If the cells are spent, some measurement instruments will indicate this by a special signal, but it can also be deduced by negative readings, or if 20.9 %V cannot be reached anymore during calibration. One cell is about 300,-€, prices increasing due to the electrolytes used. Beware of the production date and shelf life!
Strange switching oscillations in the PLC auto range readout of the OA1500. Reason is unknown, the effect happens arbitrarily. Quick-fix is to power-cycle the PLC.
Diode Laser Spectroscopy
An extremely attractive method for gas analysis is spectroscopy with tunable diode lasers. It is virtually non-invasive and has in first order no spendable components. However, commercial devices are still very expensive. For water it is simple, for oxygen long paths are needed, thus multi-range instruments possess some difficulties.
O-Rings, Seals and their Operation
A very basic and important factor for the quality (tightness) of a seal is the mounting procedure. The mounting pressure on the seal shall be applied homogeneously. If several screws are used to apply this pressure they shall be tightened in a cross over fashion, where each time the diagonal between the screws splits the non-fastened section into two equal halves. Otherwise the sealing material will be shoved along the groove and be distributed inhomogeneously. Thus the number of screws should be preferably greater than and a multiple of four.
When mounting a seal also the applied torque is important. During one round all screws are fastened to a fixed torque in a cross-over fashion:
- In the first round 50% of the nominal torque shall be applied .
- In the second round 80%.
- In the third round the full nominal torque.
- After 24 hours the sealing material has settled and another round with the full torque should be performed.
It is helpful to number the screws in a clock-like way and write down the fastening order. For symmetric, circular distribution:
- For 8 screws: 1-5, 3-7, 2-6, 4-8
- For 12 screws: 1-7, 4-10, 2-8, 5-11, 3-9, 6-12
- For 16 screws: 1-9, 5-13, 3-11, 7-15, 2-10, 6-14, 4-12, 8-16
If the screws are not homogeneously arranged around the flange or there are corners in the groove, this should be regarded in the fastening order as points where the "flowing" of the seal-material is hindered.