How to Run a Dual Beam World Day experiment

Transcription

1 March 28, 2005: Note, in the last month, the line feed receiver cal switching has been made the same as the Gregorian. One cable is no longer necessary. Check the diagram. How to Run a Dual Beam World Day experiment This covers the simultaneous use of one or more of pwr, mracf, clp, and dspc (power profiles, F region seven frequency spectral program, E region spectra, D region spectra) Overview of the various steps: Preliminary In any order: 1. Plug in all the cables Note: data taking cannot run without the cables to the samplers; it is best to plug in all the cables immediately to avoid mistakes or forgetting a cable. 2. Make sure that the data taking window on the workstation in the receiver room is present and accepting commands. (For example, type "io <return>") 3. Make sure that you are logged in to the observer workstation in the control room. (the one with two flat screens; the user is "dtusr", password available there) Important note: after steps 1 and 2, you must set up the Gregorian receiver and enable the cal port. In the data taking window type "ifsetup 430_30blk <return>" and "calport 7 <return>". 1 Do not leave disc or tape io on by mistake. Running the Experiment* 1. Start the program to receive data from the VME crate (radariface). 2. Start the programs to compute the spectra and profiles from the raw samples. 3. Start the data taking (receiver room) with all io off. You can do step 3 first if you want, but it seems better to get 1 and 2 out of the way. 4. Bring up the transmitter, set levels, etc. 5. When you are convinced everything is OK, stop data taking. 6. Turn on gio, restart data taking. 7. Start the monitoring. If you really think you know what you are doing, you can start initially with gio "on", but the trial start seems to help avoid problems. *Detailed instructions follow the block diagram on the next page, then the cabling set up and additional explanations.

2 Analog Data Control... Block Diagram of the Computers and Data Paths Interface Sun in receiver room Digital Data This is set up from the Sun computer in the receiver room. See detailed instructions for how to set up each operation. VME Crate EDT Link Network Run and Monitor Line feed data from here Observer Linux PC (aeron) Raw Samples go from here to here and are stored on disk. Run and Monitor Gregorian data from here 2 This machine stores the raw samples on disk (in for example, /share/aeron4) and does the computing. We start the computing from the Sun (observer) in the control room (two flat monitors). Use rlogin to get to aeron. See the detailed instructions.... We also monitor the results of the computations here. For dual beam monitoring: 1. Use one asp program for the line feed results on the left monitor. 2. Use another asp to monitor the Gregorian results on the right monitor.

3 Running the Experiment 1. Start the program to receive data from the VME crate (radariface). 2. Start the programs to compute the spectra and profiles from the raw samples. 3. Start the data taking (receiver room) with all io off. 4. Bring up the transmitter, set levels, etc. 5. When you are convinced everything is OK, stop data taking. 6. Turn on gio, restart data taking. 7. Start the monitoring. 1. Start the program to receive data from the VME crate (radariface). 3 Do this from the left monitor on the observer workstation in the control room. (This is a convention; it could be anywhere, but it is useful to know where to find the program!) In the small console window, type "radariface & <return>".* You need to tell radariface the name of the file where is to be stored. Normally use a name like "/share/aeron4/10oct02". The path is our largest directory for storing data. The date should be the day the experiment starts. This name is a base. radariface will make a numbered sequence of files. Type the file name where indicated in the radariface window. Click on "new file". iface will add ".000" to the file name if there are no files with this name. If there are already files with this name it will use the next higher number (xxxxx.053 or whatever). *Returns are understood from here on.

4 Running the Experiment 4 1. Start the program to receive data from the VME crate (radariface). 2. Start the programs to compute the spectra and profiles from the raw samples. 3. Start the data taking (receiver room) with all io off. 4. Bring up the transmitter, set levels, etc. Left Right 5. When you are convinced everything is OK, stop data taking. 6. Turn on gio, restart data taking. 7. Start the monitoring. 2. Start the programs to compute the spectra and profiles from the raw samples. Two parts: a. Edit the file "compute" b. Start the programs. In the right hand display on observer, open an xterm and rlogin to aeron: "rlogin -l sulzer aeron". You need to know the password (Ask.) Next type "cd asp/wapp" and "vi compute". (You can use another editor if you like.) The correct filename base must be put into three file names. The ".000" must be present (or possibly a higher number if you are restarting) on "aerfile", which is where the raw data go. The ".line" and ".greg" files hold the profiles and spectra. From the left display, make another xterm and log onto aeron. Type "cd asp". Then type "./asp wapp/compute". Finally type "linefeed". asp will print the name of the data file, and then indicate that it has found the end of the file. (It is waiting for data to appear in the file.) From the right display, make an xterm, log onto aeron, and start up asp as above. Now type "gregorian". asp will wait for data. You now have two programs, one for each feed, waiting to begin computing. You can use the first xterm you made (to edit the file) for listing the files, etc.

5 Running the Experiment 1. Start the program to receive data from the VME crate (radariface). 2. Start the programs to compute the spectra and profiles from the raw samples. 3. Start the data taking (receiver room) with all io off. 4. Bring up the transmitter, set levels, etc. 5. When you are convinced everything is OK, stop data taking. 6. Turn on gio, restart data taking. 7. Start the monitoring. 3. Start the data taking (receiver room) with all io off. This is simple unless you need to change the set up for the raw data program. 5 First, make sure all io is off. Type "io"; if any of the three is on type as appropriate : "tape off", "disc off", or "gio off". To start an LCTS World Day type "dbraw mpc". This starts mracf, power, and coded long pulse. For a non-ltcs World Day type "dbraw mp". "dbraw mpcd" starts them all. If you need to change the set up for any of the programs, you need to know that dbraw is a macro which runs the program rd with three different program ids: dbmracf, dbpwr, and dbclp. To change the set up for mracf type "rd" in the left hand window (does not say "data taking" at the top). For "new program id" type "dbmracf". When you are done making the changes, you can type "send" to make sure they get to the VME crate. Do not change the set ups unless you know what you are doing. If there is some doubt as to the state of the set up, there is a command for getting the defaults back. This is the only thing you should do unless you know exactly what to do. The set up is not bullet proof and you can encounter errors that might not be so easy to understand.

6 Running the Experiment 1. Start the program to receive data from the VME crate (radariface). 2. Start the programs to compute the spectra and profiles from the raw samples. 3. Start the data taking (receiver room) with all io off. 4. Bring up the transmitter, set levels, etc. 5. When you are convinced everything is OK, stop data taking. 6. Turn on gio, restart data taking. 7. Start the monitoring. 4. Bring up the transmitter, set levels, etc. 5. When you are convinced everything is OK, stop data taking. 6. Turn on gio, restart data taking The operator will handle the transmitter; you need to learn what the scope display should look like so that you can set the levels. 5. To stop data taking, type "<control> c" in the data taking window. You will have to wait until the end of the next integration period before it stops. 6. Now type "gio on". This means that when data taking is started the output buffers will be sent over the EDT link to the Linux PC aeron and radariface will receive them and write them in the raw data file that you specified earlier. The two computing jobs will read the records and make the profiles and spectra, writing the results in the output files. (In step 7 you will start the monitoring in order to look at the results.) Type "dbraw mpc" or "dbraw mp" or "dbraw mpcd" to start the process.

7 Running the Experiment 1. Start the program to receive data from the VME crate (radariface). 2. Start the programs to compute the spectra and profiles from the raw samples. 3. Start the data taking (receiver room) with all io off. 4. Bring up the transmitter, set levels, etc. 5. When you are convinced everything is OK, stop data taking. 6. Turn on gio, restart data taking. 7. Start the monitoring. 7. Start the monitoring. (If you do not do this, the data still should be fine, but how would you know?) Go back to the control room (observer). 7 On the left hand display of observer, start another xterm; type "doasp". Then type "181.5 >> /pwr/caltemp". (Use the current linefeed cal. temperature.) Now you need to tell it which file to read the profiles and spectra from. Type, for example, "store{/share/aeron2/10oct02.line} df". Use the actual file name you put into the file "compute". Note that "line" means that this is the data from the line feed. You monitor it on the left display, because that is where you started the computing. Now type "olmon". Two or three windows will open displaying the different types of data. Arrange them as you want. Now do the same thing on the right hand display, but the file name should end in ".greg", and the Gregorian cal. temp is about 95. Two or three plots of Gregorian data will appear. The monitoring programs occasionally stop (and wait for keyboard input) when they have problems reading over the network. When this happens, type "cd / clr" and then "olmon".

8 Functional Block Diagram of the Dual Beam World Day Set Up 8 Line Feed Gregorian Controller (SPS) IPP, beam, rf, and phase The receiver set up takes an if signal and makes filtered baseband (I and Q). The sampling line runs to a buffer on the way to the RI to allow it to feed the scope as well. Cals Interface Transmitter Control panel IPP and Gate (Sampling) lines The beam and rf go to buffers in order to feed the scope. The PN code generator is used only for the coded long pulse (the E region program used in LCTS). Its output feeds the transmitter. The SPS phase runs to it instead of directly to the transmitter and is switched through when the CLP is not used. It requires other inputs as well; see details. PN Code Generator Note: As of March 2005, the Gregorian and line feed cals are done together as was previously done for the Gregorian alone.

9 Block Diagram of the Set up 9 Input Signal Line Feed Gregorian Controller (SPS) Distribution Amplifier Note: the inputs must be the two ports that are symmetrically placed. The output must be the "odd" one. The linefeed comes from cable comp. amp. output 7-2. Switch on 3-5 must be on 400 MHz. Greg. comes from 5-13 output #1. Must do setup! (ifsetup 430_30blk) Functional Block Diagram of the Dual Beam World Day Set Up Interface Transmitter IPP, beam, Control panel rf, and phase Cals Transmitter Sample(see page 12) IPP and Gate (Sampling) lines PN Code Generator One setup each for linefeed and Gregorian Combiner Filter Characteristics Freq.: 250 KHz Input gain: 0 db output gain:20 db All CH 30 MHz I Atten. 12 db Inputs Baseband Mixer Q Krohn-Hite Filters Atten. 12 db Outputs Why 250 KHz? We sample at a 2 µsec interval, and so we have a 500 KHz clear bandwidth. The two 250 KHz baseband filters give this. CLP uses this bandwidth to get 300 m range resolution. mracf uses it to assure a very flat bandpass in the central 250 KHz it needs. PWR uses it to get a superior match for its 4 µsec baud and a properly sampled power profile. These cables run by the scope on the way to the Interface. Use Ts to connect to scope. To scope and interface

10 SPS and Transmitter Control Set up, No CLP Note: the first cable is no longer necessary. To 9-2 Cal 1 and 5-11 TTL input Port To RI-IPP (10-10) 50 MHz RF SPS PWR Start CLP code switch SP0 SP1 SP0+SP Phase TX-IPP Beam RF RI-IPP CAL SMP0 ext. trig. ext. 20MHz IPP PW RF Phase MHZ Transmitter Control Panel Ohm Drivers In Outputs Beam RF To Scope input 4 (trigger) Connections the transmitter control panel can come through the 50 ohm drivers; this is not indicated in the simple block diagram below. In 9-14 Outputs To RI-GW (10-10) SP0 To Scope input 3 Line Feed Gregorian Functional Block Diagram of the Dual Beam World Day Set Up Interface IPP and Gate (Sampling) lines Not used in this set up Notes: The top oscilloscope is used to display the Gregorian; the bottom for the line feed. Only a single output to the scope is shown for each line above. Actually, two outputs are used, one for each scope. The rf pulse triggers the scopes and is also displayed to help show that the correct SPS buffers are cycling. SP0 (the sampling pulses for the A/D converters) is also displayed for the same reason. It really is not necessary to put it on both scopes, and so one of those inputs is free for other purposes if needed, such as the beam. Controller (SPS) Transmitter IPP, beam, Control panel rf, and phase Cals PN Code Generator The digital inputs to the scope should be terminated by the internal 50 ohm loads. Lines 1 and 2, the I and Q signals, should not be terminated; they are terminated at the RI, and just "pass by" the scope.

11 SPS and Transmitter Control Set up, Using Code Gen. for CLP (LTCS) 11 To 9-2 Cal 1 and 5-11 TTL input Port To RI-IPP (10-10) 50 MHz RF SPS PWR Start CLP code switch SP0 SP1 SP0+SP Phase TX-IPP Beam RF RI-IPP CAL SMP0 ext. trig. ext. 20MHz IPP PW RF Phase MHZ Transmitter Control In Ohm Drivers Outputs Beam RF To Scope input 4 (trigger) In Outputs SP0 To Scope input To RI-GW (10-10) Line Feed Gregorian Functional Block Diagram of the Dual Beam World Day Set Up Interface 9-15 Pseudo-Random Code Generator int./ext ext. code clock PN out Controller (SPS) Transmitter IPP, beam, Control panel rf, and phase Cals IPP and Gate (Sampling) lines PN Code Generator Line 6 of the SPS is used to switch between the code coming from the sps (ext.code) for programs other than coded long pulse, and its own internally generated code for the coded long pulse program. Thus PN out feeds the transmitter control, passing the SPS code on through except when the coded long pulse progrm is running. The code generator gets its clock from the sampling pulses.

12 rack no. Chasis no. Transmitter Sample (30 BEAT) Distribution Amplifier Adjust attenuator as required Use one output for each receiver To power splitters in receiver setups Other Things to Remember switch must be in 400 position. (3-5) 2. Load/Ant. SW on CRR side. (9-2) 3. Put selector switch in PRF gate position. (7-2) 4. On chasis 8-1: a. CRR 1,2,3 selector switch should go on the highest temperature position. (The number will be typed into the monitor program.) b. Move both CUTOFF switches to the "on" position. c. Move the "Cal 1" switch to the off/ext (middle) position. d. Set the thumbwheels to 210. e. Move the transmitter disable switch to the up position (allows the 430 MHz tx to operate). Increasing rack no. Increasing chasis no. The racks in the background are numbered.

13 13 More Details 1. Restarting the computing and data transfer: The data transfers and computing rarely have problems. The exception is during close lightening storms. In this case the computing encounters some bad records, but it usually manages to skip them and find the next good record. Once in a while, the bad records are continuous. (The computing asp processes might complain or the monitoring programs start beeping; they do this when not fed new records for a while, and so if you have stopped data taking, expect it; otherwise find out why.) If the transferring/computing needs restarting: a.first, stop the data taking. b. Quit radariface using its menu. c. abort both asp computing processes; "<control> c", followed by "quit" will do it, or "<control \". d. restart radariface; enter the same file name (unless you have a reason for changing) and do "newfile". radariface will start using the next file in the sequence. If there is any doubt about the files, look in the directory where they are located. e. On aeron, edit wapp/compute; if you are using the same file name, you will need to change the.000 to the number that radariface has decided to use. f. Start the computing f. Start the data taking. 2. If the datataking programs will not run correctly: It is sometimes necessary to quit datataking by typing "exit" in the window. Typing "datataking" in the window on the sun restarts it. If it will not start up, it might be necessary to reset the VME crate. This is rare. 3. The setup program is run on a window other than the datataking window. Type "rd". Next type "new <program name>. The program names are dbpwr, dbmra, dbclp, and dbdspc. Then change the appropriate paramters; you need to do a "send" if the datataking is running. 4. On any non-datataking window you can look at macro definitions. You need to cd to directory: /home/online/tcl/proc/atm/gen. "goproc" gets you to here if you are logged on as dtusr, as you are in the receiver room.

14 1. To test the EDT link: a. make sure that radariface is running and select a new file b. On the datataking Sun at a pointing prompt type rundiag c. 0 dap2 d.?? (to get the gio test) e. 4 to select a test pattern (Select one of them.) f. 1 to set number of times to loop. g. 88 to write the file. h. Now look at the file with od -x <filename>. i. Repeat with different test pattern as required. Important Note on the Khronhite filters: Diagnostics The filters have two modes: L.P. (lowr pass filter) and GAin(straight through amplifier with no filtering. However, there is only one display, and so if the bandpass is displayed (the normal situation), then you do not see the mode. Push the mode switch to see the mode; (push the frequency switch to see the frequency again). If you have appaernt low signal to noise ratio, the filter might be in the wrong mode. Use the mode switch to put in in the L.P. mode.