Weekly report of problems

08nov06	15nov06	22nov06	27nov06	06dec06	13dec06
23aug06	30aug06	20sep06	27sep06	04oct06	11oct06	18oct06	25oct06
31may06	05jul06	12jul06	19jul06	26jul06	02aug06	09aug06	16aug06
22mar06	29mar06	05apr06	12apr06	19apr06	26apr06	03may06	10may06
07dec05	04jan06	11jan06	15feb06	22feb06	01mar06	08mar06	15mar06
28sep05	05oct05	12oct05	19oct05	26oct05	02nov05	23nov05	30nov05
03aug05	10aug05	17aug05	24aug05	31aug05	07sep05	14sep05	21sep05
25may05	01jun05	08jun05	15jun05	29jun05	06jul05	13jul05	27jul05

13dec06

14dec06: large birdies were seen at 418.625 and 419.2 Mhz in data taken 01dec06 and 06dec06.

06dec06

06dec06: az swings at za=10 with the tilt sensors were done to try and measure any residual platform tilts. The Pointing checks showed a 9 asec 1az term that could come from a platform tilt. The tilt sensors did not measure any residual tilt. The 9 Asecond tilt is probably true but that would be a .0025 degree 1az term. The tilt sensors fits usually give rms fits around .005 degrees.
10dec06: refit UT1 - UTC dec06 thru Jun07.
11dec06: fit 1 az and sin(za) terms to pointing errors and update the pointing model.

27nov06

04dec06. Cell phone detector birdie at 1548 Mhz Data taken back on 08nov06 showed that the new cell phone detector has a birdie at 1548 Mhz that is about 15 times Tsys in a 390 hz channel width when the cell phone detector is in the parking lot adjacent to the control room.

22nov06

28nov06. Pointing checks after the az 1 encoder failure show a .19 " offset at the azimuth encoder and a 1 azimuth pointing error term probably caused by the tiedown cable lengths being different from when the model was made (do to tiedown encoder failures??).
28nov06: The 3rd amplifier in the downstairs if/lo has a gain of 14 db rather than 10 db. See checking the programmable gain linearity.

15nov06

21nov06: look at pointing check after the 14nov06 az encoder 1 failure. 1 source has a 20" offset. Need more data to determine whether it is an encoder offset or something else.

08nov06

12nov06: a2010 saw beam1a noiser than the other beams on 01nov06. This looks like it is a problem with the fiber transmitter for beam1a.

25oct06

26oct06: 430 transmitter had rf on before ack light active. Problem was the monoplexor in the ch was taking 3 useconds longer than the monoplexor in the carriage house. The delay was initially measured and the software limits set to 12 usecs from tx ipp to rf on. When the software limits were set i think it took 3 usec (signal up) 3 usecs turnon, 3 usecs (signal down) = 9 usecs. This gave 3 usecs of slop for the software limit. With the added 3 usec delay of the ch monoplexor, it is now possible for the rf request to come before the ack. Jon's box will hold off the rf until the ack gets back.
28oct06. 14oct06 az1 encoder failed from lightning. Problem was the encoder board port 1.
30oct06: lbw cable test upstairs iflo showed loose cables in 750 mixer chassis.

18oct06

18oct06: axis 212 ptz web cam. Looking at its birdies in the shield room and on the telescope.

11oct06

12oct06: dewar monitoring failed. Problem was the upstairs ethernet repeater.
12oct06: added dewar monitoring page.
17oct06: looking at the rfi from the axis 212 ptz web camera.

04oct06

04oct06: 614 khz comb in 327 receiver. dynamic spectra and average.
05oct06: 614 khz comb in 327 rcvr. az,za, and time dependence. We found that the comb was coming from the azimuth 2 encoder (carriage house side) on the platform. It is one of the new encoders that was installed 02aug06.
09oct06: added rfi page for encoders. Included comb at 430 Mhz.
09oct06: compare 614 khz comb strength using new and old encoders.

27sep06

28sep06: added alfa stability page. Uses TsysPolA/TsysPolB for a2010 drift scan data. Added months aug06, sep06 (thru 20sep06). Shows when and how much beam 5b and 6b jumped.
01oct06: alfa stability page. Added months jan06 thru jun06.
03oct06: The alfa stability monitoring (polA/polB) for sept shows that different beams were drifting and jumping during the month:

beam 5B was jumping around on (03,09 thr 14,) sep06. It was especially bad on 9sep and 12sep. On 11sep the jumping started. The 12 second radar birdies/compression seemed to be magnified on 11sep06.
6B was moving on 05sep.
3A,4A,5A drifted together on 16-17 sept and30 sep. This is also seen by looking at the first ten strips of each day for sept for a2010 on these days.

20sep06

25sep06: 12 second radar appears at 1378 for 430 seconds on 05sep06.

30aug06

31aug06: new lbw cal values installed. Measured on 06jul06.

23aug06

23aug06: check the rf radiation monitor at the visitors center with the 430 transmitters.
29aug06: new cband cal value installed (measured back in 06jul06).

16aug06

16aug06: 430 dome fit gain vs frequency.
20aug06: added daily monitoring of tiedown tensions.

09aug06

09aug06: power dip caused dewar monitoring to hang up. Had to cycle the power on the gpib to ethernet converter up in the rfi box in the dome. The ups for this box is probably not working very well
09aug06: Tsys/gain stability of cbh using the wapps and 100 Mhz bandwidth, Compare this with the data taken with square law detectors.
10aug06: dewar monitoring stopped working after a power dip. Brought system down to control room to test it out. On 11aug06 it started working again with a 4 % packet loss. It was using fiber c1-11 and c1-12 (main cables rcvr room to dome tx). We found that main cable c1-11 had 5 db more loss than c1-12. We switched c1=11 to c1-13 (which was unused. It used to be the dome velocity monitor of hagens.). Still had 4% packet loss (but we left it connected to c1-13 ...c1-11 is now unused). Dewar monitoring system is running in normal config with downstairs repeater installed. The 4% packet loss is more than normal (should be 0). It may be a xcvr or xcvr power supply problem upstairs or downstairs. Should check the platformnet connection in vertex shelter with pc to see if its packet loss is high.
12aug06: td12 encoder stopped working around noon after a power dip. The encoder always read 9" (without giving an error). When the td was asked to move, it would move up or down till it hit the limit switch (since it was looking at the encoder for the position feedback). I manually positioned td12 to 12" (using a tape measure) and arranged the other 2 tiedowns to be at the right position. I checked the voltage on the encoder board and it was ok (5 volts). The tiedowns should be left off till this is fixed.
12aug06: 430 yagi performance after moving the antenna closer to the line feed (before the low loss cable was installed).

02aug06

03aug06: test mmic receiver in cryo lab for stability. include a noise src and cbh.

26jul06

30jul06: test mmic receiver in cryo lab for stability.

19jul06

19jul06: punta salinas blanking test. data taken 09may06.

12jul06

13jul06: added sefd curves page and sefd fits for lbw.
14jul06: 1188 Mhz birdie from kronos timecard reader.
17jul06: 1200 Mhz narrow birdie.
18jul06: az, za dependence of birdies transmitted from the control room and the parking lot.

05jul06

07jul06: Tsysall would not work. Upstairs power meter has been shut off. Someone had been working in the rfi shielded box where the power meter was installed. When the reinstalled the box cover, it pushed the off button of the power meter. After finding the problem we moved the power meter farther into the box to not colide with the cover.
10jul06: birdie near 1320 Mhz appears in a2049 data on 06jul06.

31may06

01jun06: system performance 430 yagi after fixing the broken capacitor on the driven element. The performance did not change.

10may06

12may06: compare focus curves sbtx old and new horns.
16may06: for IF monitor display added the alfa filter bank. It will say fbAlfa wideband or fbAlfa 1390-1490.

03may06

03may06: distomat birdies after the windows were replaced.
07may06: performance of new sband tx horn.

26apr06

26apr06:. trendata 1gb ethernet switch causes rfi that can be seen in the 327 receiver.
30apr06: power failure caused little star in the turret box to lose its program. Had to reload the program.

19apr06

19apr06: the 750 and 40 Khz combs in the 327 receiver are coming from the sband console in the dome.
21apr06: one of alfa's refrigerators fails on 18apr06.

12apr06

13apr06: testing the trenddata gigabit switch for rfi.
16apr06: 750 Khz and 39.45 Khz combs in 327 receiver.

05apr06

05apr06. ntpd time jumps on the wapps 2003-2006. Where they come from.
09apr06: the improvement in the 430 ch yagi after the last focus move.
11apr06: sbn cal value remeasured.

29mar06

30mar06 echotek configuration error caused IQ imbalance in atm data.
01apr06 when the ntpd on aeron lost sync 2003 through 2006.

22mar06

28mar06: look at the new ac units to see if the 1 Mhz comb is still present. Switch the old Ac unit on and off to make sure we can see it.

15mar06

16mar06: td12 does not always start when it is commanded to.
16mar06: description of a/d signals used in the tiedowns.

08mar06

08mar06: az 2 encoder (ch side) jumped on 07mar06.
09mar06: Measured Tsys for the 9 filters of lbw. Loop through measuring the filters 6 times. The 1280 hipass had a variation of Tsys the first time we measured it. After that it was stable.
11mar06: Saw total power oscillations in cbh while tracking the crab pulsar/nebula. Added Oscillations page.
14mar06: azimuth encoder difference and the azimuth encoder rack gear.

01mar06

02mar06: td 12 releases for about 10 minutes.
07mar06: the jun05 warmup of lbw actually caused resonances in polA and polB.
07mar06: lbw polB cal cable. Wiggling cables external to dewar do not cause cal deflection to move by the 20% jump that we've seen before.

22feb06

22feb06: Measured the channel number for the center of the alfa/wapp bandpass.
24feb06: compute the harmonics in the alfa/wapp receiver with cfr=1385.
28feb06: lbw polB cal value jumped by 20% on 04dec05.

15feb06

15feb06: joanna rankins had problems with her polarization data taken on 26nov05. When i look at her data, i see no correlation in the polA, polB cal values (as if it was taken with 2 diodes rather than 1). We were running with the single pol B diode till 23nov05. So the new diodes were installed when she took her data. It may be that the software that was forcing the polB 90deg phase shift was not updated till 28nov05 (monday). If she took her data with the 90deg phase shift it still seems like there should be correlated power (but maybe not in the cross spectra were i was looking). Need to understand this a little better.

01feb06

07feb06: The radar multiplexor (calixto's box) needs to be worked on.

25jan06

27jan06: checked the azimuth encoder offset using the calibration data nov05-jan06. The nov05 encoder failure and then resetting did not introduce an offset.

18jan06

21jan06: It is no longer possible to send the carriage house signals to the wapps. This caused a problem with dan stinebring when he tried to take 50 Mhz, 430 Mhz data simultaneously. The loss of functionality occurred when bill sisk removed the external inputs of the buffer amplifiers.
23jan06: The 50 Mhz receiver has a frequency offset of about 150 hz when tested with the birdie maker. The synth used to generate the frequency was checked and it is locked to the station clock.

11jan06

13jan06 xband receiver check after reinstallation in dome looks ok.
13jan06: hipass and band pass filter placed in front of the 327 rcvr 1st amp to test for intermods in the receiver.
14jan06: 3rd harmonics of fm stations showing up in 327 receiver. They are not being created in the receiver.
15jan06: wide band look at dome ac birdies in 327 Mhz receiver.
16jan06: 28oct05 327 az swings showed birdies with 1 minute periodicity.
17jan06: 30nov05 data. compression in the alfa receiver with the 100 Mhz filter installed.

04jan06

09jan06: on 26dec05 murray had no power in polB for lbw for 1 file. A little later in the day things started to work again. This looks like it was a problem in the switches in the downstairs if/lo polb.
10jan06: 1 Mhz comb of ac units in dome is also seen at 430 Mhz (data taken 06dec05).

07dec05

12dec05: 9 level wapp pulsar folding is doing the 9 level correction incorrectly.

30nov05

02dec05. galfa observers had reported at ripple in beam 2a for the narrow band spectra. The wide band spectra did not have the ripple. On 02nov05 we switched the galfa 2a,2b inputs. The ripple stayed in pixel 2a so the problem was in the digital card (we then returned the cables to their normal order). The system was rebooted by cycling the power (on the power supply). After this the problem went away.
05dec05: A2125 was doing mapping with the 430 Mhz reciever. Periodic birdies with periods of 1 minute were evident in their data. On 02dec05 i took 2 hours of data to characterize the periodic birdies.

23nov05

24nov05: lbw polB filter 3 (1230 to 1470) is drifting. It is also showing some jumps in total power when we reselect it.
28nov05: lbw cal values for new diodes installed.

02nov05

04nov05. Alfa beam 1a looks like it jumping around a little (.5 to 1% of Tsys). This was not happenning on 08oct05 and was seen on 17oct05. Take a look at the total power versus strip for oct05 and compare the red polA trace (top plot) with the red polB trace (bottom plot)
05nov05 azimuth dependence of birdies in the 327 rcvr.

26oct05

27oct05 lbw filter bank. PolB 1230 to 1470 Mhz filter (filter number 3) is drifting in gain. The polA filter is ok. Using some of the other filters in the filter bank shows no drift. So the problem must be inside the box with one of the connections for that filter.
29oct05: murray sees calibration jumps in his OH data aug-oct05.

19oct05

21oct05: output of 327 Mhz rcvr after the first amp and before the rf filter.

12oct05

14oct05: cal stability for mathews 430 feed checked. Now looks ok. Tsys is 165K if cal is 42K.
16oct05: The 327 Mhz receiver has 2 separate 1 Mhz combs. They are coming from the AC Units in the dome. They span the entire receiver band and get up to 8*Tsys in a 95 Hz channel width.
15oct05: A power dip caused the downstairs synthesizers to all go to rf off. Turning rf on was not enough to get them working again. I had to cycle the power on each of them before the gpib bus would start talking again. Observer lost about 1/2 hour...
17oct05: location of 1350 faa radar aliases in alfa.

05oct05

06oct05: data taken 28sep05 shows that a large fraction of the 1300/1400 mhz birdies are being radiated by the wapps.
07oct05: The Tsys jump of sbw in aug05 was from the cal value changing.

28sep05

29sep05: added 430 carriage house cal values. Looks like there is a bad match in polA starting at the diode. The reflection is 3.75 meters in air, (2.75 m if velocity is .7c).

30sep05: at midnite (start of 30sep05) a 1 second tick was inadvertantly inserted into the station clock. The clock was off by 1 second (The station clock provided a time that was 1 second behind the correct time). The offset was corrected about 8:30 am 30sep05. During the morning of 30sep05 the tiedowns were acting strange and the rotatary floor vibrated a bit. I think this was because of the 1 second time difference. The little star computers did not jump with the 1 second tick insertion. The time stamped positions that they received were one second late, so most of the tracking commands sent to them would have be ignored.

03oct05: Early morning, the sband radar run had trouble with the cal program that fires the cal. This program runs on the corp1 computer. Programs were not able to communicate with it. Rebooting the cor crate fixed it. The backplane network for the correlator crate has been subnetted. This problem has occurred in the past. That time a computer using the "other" part of the subnet sent arp packets saying that they also belonged to the "cor backplane" part of the sub net. This caused requests to go to this "other" computer rather than the cor backplane (there is no evidence that this is what actually happenned this time..).

21sep05

26sep05: 11aug05 atm digital receiver glitches in test data.

14sep05

14sep05. td 4 both load cells out of calibration (they had been calibrated onn 04sep05). ldcell1 had a dc offset of 1.5 volts in the amplifier. Both load cell amplifiers had oscillations in the output of many millivolts. Replaced both load cell amplifiers.
14sep05: polarized tsys in alfa.
15sep05: the cal deflection in alfa is not a function of the rotation angle.
16sep05: td 8 had trouble with it's analog input.
20sep05: cell phone harmonics in lbw when transmitting from inside the dome.

07sep05

07sep05: dewar monitoring failed 06sep05 after power dip. Had to cycle the power on the upstairs fiber to gpib transceiver. The fiber repeater in the racks downstairs for the platform ethernet is not on a ups. This is probably causing the problem. Luis put the repeater on the ups in the racks.
07sep05: tension in tiedown cables after td4, td8 were recalibrated on 02sep05. Td4-2 looks like it's gain needs to be recalibrated.
12sep05: 1 section of rolling surface of turret floor replaced 06sep05. On 12sep05 did two turret floor spins with the tilt sensor to see if there were any bumps when the old to new rolling surface passed over a roller.

31aug05

31aug05: 34970 multimeter for dewar monitoring continued to fail. Brought the device down to the labs. Had trouble turning it on with no cards plugged in. Will try the spare 34970 upstairs tomorrow.
31aug05: looks like lbw tsys has jumped up 4 or 5 K after the 26aug05 warmup then cooldown.
01sep05: replaced the 34970 with the spare and the dewar monitoring now works (used the old plugin cards).
02sep05: replace spectrum analyzer in rfi shack with spectrum analyzer that was in the ant test range. rfi monitoring working again.
02sep05: td 8 and td 4 the amplifiers for 4 loadcells had their voltage regulators broken. replaced the voltage regulator chipes and recalibrated the loadcells. tensions on tiedowns now working. The tension on td 4 loadcell 2 is drifting by a few kips (with no load present). The db25 filter for td8 was bad. It was removed (we need to order more of these). td is also missing the db25 fitler.
02sep05: lbw is only using the 2nd diode (since diode 1 was replaced and the spare diode has not been calibrated). This configuration will not work with the lbw hybrid in to give circular polarization. When diode 2 feeds both pols, there is a 90 deg phase shift between polA and polB. When this is passed through the hybrid to create the circular, no cal power gets to one of the circulars. Use linear pols until we fix diode 1.
02sep05: p2030 alfa pulsar search. Observers complained about large spikes in data. Showed them that it was from the out of band faa radar.
04sep05: td12 db25 filter for the pcu shorted out the 24 volt supply. Observations done with tiedown tilted by about 8 inches (need to check the pointing error versus tilt angle).
05sep05: 16khz comb at 327Mhz. Grounded the motor shields 29aug05. Improved things for awhile but then it go worse.
06sep05: lbw increase in Tsys is coming from the resonance at 1400 Mhz. The 27aug05 problem of murray lewis with no signal in polB was caused by the h90cal and his selection of the circular polarization option. The signal was present, but no cal got through to one polarization (because of the h90cal 90degree phase shift and the 90 degree phase shift of the hyrbrid to create circular).

24aug05

25aug05: measured tsys for mathews feed (22aug05). Tsys is about 151 K. There is a 220 second ripple in the cal
27aug05: lbw . murray lewis took some oh data in the morning and he got no signal in polB. He was using the hybrid and the 1550-1820 filter. The upstairs power meter showed power in both pols (-40 dbm). The correlator got no data in polB. Later users of lbw had not trouble (although they had slightly different setups). The downstairs synths are not (rf off) since they are shared between the two pols.
27aug05: the 16 Khz comb in the 327 Mhz receiver is coming from the alfa motor controller.
29aug05: td8 . power cycled on td8 around 3pm. Later that evening (8pm) it was noticed that the encoder readout value for td8 was not changing (even though the jack was moving). Cycled the power on td 8 (ups and cb1,cb2) and the encoder started reading correctly.
30aug05: on 26aug05 about 14:00 the rfi hilltop monitor stopped working. Turns out the spectrum analyzer is no longer communicating on the gpib bus. Brought the analyzer down to the lab.
30aug05: td8 has been reading motor over temp. The circuit for this bit was wrong. Updated the circuit to read correctly. The other tiedowns have the same circuit. they were updated to the new version.
30aug05: the power cable for the 34970 used for the dewar monitoring was bad. Moving the cable caused the power to the device to go off. Reseated the cables but it failed about 15 minutes later.

17aug05

18aug05: lightning smoked td 12. Tensions went bad for awhile and then the encoder position went to 24Inches. This usually means the little star is not reading the encoder values correctly. When jorge,luis looked at the little start, the front panel display was blank. Could not change little star since lightning started up again. will do it tomorrow.
19aug05: td 12 started working after encoder interface board was changed. Looks like the kips on once load cell of td 8 are now off by 20 kips. this occurred during the lightning ...
20aug05: datataking hung up after the power changeover from generator to comercial saturday evening. All 4 of the downstairs synthesizers had cycled their power. You could see this since all 4 of them had the rf off indicator set. After the power cycling the gpib bus that controls them was hung up. I cycled the power on the 4 synths and then restarted the if2Prog that controls them. These synthesizers need to be put on a ups (if they are currently on a ups, then the battery/load needs to be checked). The observer probably lost over an hour of observing time...
21aug05: the galfa piggy back on alfalfa (they're calling it togs) complained of an rfi at 1421.6 Mhz. This is an intermod of the 1350 faa radar created in the galfa mixers. 1350 is at 285If1, lo1 is 195.845. The intermod is 3*192.845 - 2*285= 17.53Mhz at baseband. baseband dc is 1439.16 and it's flipped so this comes out to (1439.16-17.53=1421.63). Alfalfa is probably not going to move their sky center frequency. I guess if we have a filter bank before the galfa mixer we could cut the 1350 radar...

10aug05

10aug05: problem with distomat 1 was the fiber connection. Fixed it and the distomat started working again.
12apr05: during 430 run, saw interference for awhile. Some of it was a 2Mhz comb. Was not able to trouble shoot since radar run was ongoing. Went away after awhile.

03aug05

03aug05: distomat 6 has been failing more often then the other distomats (see distomat failures for jul). Electronics checked all 6 distomats and cleaned the windows (when distomat 6 was cleaned, the coating on the front of the window slid off when glass cleaner was a>

Transfer interrupted!

04aug05: Observer using cband high had intermittant ripples in the total power. Period of 5 hz. Lasted for a few minutes and then stopped. Happened 2 days in a row. Amplitude was a few % of Tsys. When looking at the data saw a ripple in polB of about 1% of Tsys. Checked the cables and found 2 bad cables: dewar to postamps polA, and postamps to If/Lo cabinet polB. Replaced both cables. Vibrations in the system were probably causing the gain variation.
03-04aug05: maintenance is installing some beams in the feed tower to fix some beams that are deflecting because of the kevlar cables. This may affect the pointing ??? need to check. Took some calibration data 05aug05 (sbhi) and 08aug05 cbhi. The pointing did not look that bad. There may be a small ramp in azerr,zaerr vs za but it will take more data to nail it down.
05aug05: td12 trouble. motor disabled with overspeed condition. Mainly occurred when trying to pull down. Jorge went down and turned the cmd speed pot 1/8 of a turn and it started working. There is something funny going on with this td. Also found that the jack was full of water.. would push it out when you pulled the jack down to around 18 inches. Maintenance said they would look into trying to seal it better.
05aug05: pulsar observers complained that alfa was a lot less sensitive than lwide when looking at known pulsars (they were using pixel 0). Checked Tsys for alfa and it looked ok (see alfa tsys by month). They did say that the 12 second radar may have been affecting their ability measure the sensitivity.
05aug05: tim hankins complained that he didn't see the crab pulsar today. Also saw a large 1.6 second ripple in the total power (probably a few % of Tsys). Saw it in cbh and in sbh. Could be pointing error since we are looking at the crab nebula?? took some pointing data. Took some data on 05aug05 (sbhi) to check the pointing and on 08aug05 (cbhi) to check the system performance. The measurements showed that cbhand high was working ok.
09aug05: tim hankins saw crab pulses at cbhi and xband.
09aug05: 8pm. distomat 1 failed. remained off till next day.

27jul05

29jul05: dewar monitoring failed again after a power dip. They went up and checked and the ups has a problem. This needs to be checked out since it also holds up the monoplexor driver for the 430.
29jul05: still problem with td12. overspeed.
28jul05: td 12 failed after power dip. Changed little star, dac, signal conditioning. Still have problem with overspeed. The signal conditioning board might not be configured correctly. There is confusion between the differential version (old ones) and the single ended version (new ones). Not obvious what the old configuration was.
28jul05: dewar monitoring stopped after a power dip. Jorge and luis went up and found that the power strip in the rightmost rack in the rotary floor room is not on a ups. After talking to tony noya he told me that only the leftmost rack is on a ups. The rfi box in the right most rack contains the power meter used to measure the if1mp power for the fiber optics. It also holds the dewar montoring. The power dip was screwing up the ethernet transceiver for the dewar monitoring. Luis connected the power strip of the rightmost rack to the ups that is used for the monoplexor driver. This explains why we have always had problems with the dewar monitoring after a power dip. I had always thought that both racks were on a ups. The second rack also contains some fiber transcievers and the 20 Mhz buffers. I don't know whether they were also plugged into the power strip in the rightmost rack, or they had power cables that went to the left rack. We need to verify that the 2nd ups that we plugged into can hold the combined monoplexor/power strip load. We checked and it holds the power strip load just find (but the monoplexor was off).
29jul05: power dip and dewar monitoring hung up again. Since rfi box now on ups, problem must be in the fiber ethernet repeater in the transmitter room upstairs. The repeater was put on ups 01aug05.
01aug05: People reported that pixel 2b in galfa has it's power moving around again. Drifts off for awhile and then comes back. This sounds like the previous problem with pixel 2b.
01aug05: p2030 alfa pulsar search saw large 12 second compression in 01aug05 data. Turns out to be compression from the faa radar.
02aug05: Upstairs if/lo, user selected the straight thru option (if1 "if1_ifnum 5"). This should bring the rf down the fiber (no 1st lo mixing). The signal appeared on polA, but their was no signal on polB. Checking the frontpanel of sw3b,sw5b upstairs, the lights had position 5 selected correctly. Turns out that someone had taken sw3b, pos5 and routed it to an unused connector. The signal was not passing thru the amplifier. This must have happened when someone was doing a test upstairs and didn't reconnect the cables correctly. Plugged the cables in correctly and there is now a signal in both polA, polB for the straight thru (ifnum 5) position.

13jul05

13jul05: The power level in pixel 2b of the galfa spectrometer is drifting/jumping. Data taken at the same time with the wapps shows no drift/jumps. The cables for galfa pixel 2b were all tight.
14jul05: The dewar monitoring gpib-enet converter was replaced. The device that broke was a 10Mbit gpib-enet. We replaced it with a 100Mbit gpib-enet. I had tried to use the 100Mbit device before with vxWorks and could never get it to work. It turned out that the programs i had to configure the ipadress on the 10Mbit device did not work for the 100Mbit device . Jorge downloaded some windows software and got the addresses setup ok. This means that we can use the 100Mbit devices with the vxWorks software (if we want to).
14jul05: a2055 lost about half there run. Gui said waiting 86402 seconds to be on source. The basketweave routine had trouble crossing midnite if the pattern started too close to midnite. Wrote some tcl code to fix the problem. Arun was going to install it when he got a chance...
15jul05: During the day the tiedowns were disabled because people were working with the distomat temperature sensor. The log book says that the tiedowns should be turned off till further notice. I think there was a mixup in communications. After working on the distomats there was probably a thunder storm and the person probably told the operator to leave the tiedowns off "until the weather clears up.." .. Anyway the tiedowns were left off for the entire weekend. On monday morning i turned them back on.
19jul05: galfa pixels 4a and 6a were interchanged. This occured at the output of the distribution amps in the control room (the wapp data was not affected). The pixels were swapped on 11jul05 and were put back in the correct order on 19jul05.

06jul05

06jul05: cima was configuring the cal multiplexor to fire the 25Hz cal (if2_calsrc 2 ) rather than the interrim correlator (if2_calsrc 1) When the interim correlator was initialized by cima.
07jul05: The alfa Pixel 4A 174 Khz ripple is coming from the fiber receiver.
08jul05: We replaced the pixel 4a fiber receiver for alfa and the 174 Khz ripple went away. Pixel 2a still has some ripple.
09jul05: dewar monitoring stopped working. The upstairs fiber receiver did not have any lights on. The problem is either the fiber receiver or the gpib to ethernet converter.
09jul05: the distomat temperature sensor has been failing intermittently. This has been going on for awhile.
12jul05: lbw channel B gain or tsys was changing with time. The 1370HiPass and 1320 HiPass were in. When the 1370 hiPass was taken out, it was more stable. 13jul05 found that the digital line that controls the switch for the 1370 hipass polB was broken. Fixed the cable.

29jun05

01jul05: sbn dewar warms up during sband radar run.
01jul05: Looked at the 174 khz ripple in pixel 4a of alfa. (more info).
05jun05: Td12 is not starting up consistantly. The commanded power on sequence uses a two step process (see td power on sequence). The first power1 sent by the little star is not always sending the 24 volts to close the master contactor relay and give power to the pwrsup/amp. Tracing the cables it looks like there is a missing cable and the signal drive is floating..

22jun05

22jun05 - cband polarization switch hooked up. cbpol (lin,cir) switches it back and forth.
23jun05 - took on/off cals of lbw at 1400 Mhz to see the resonance in polA (more info).
25jun05 - tower 12 hit by lightning. Td 12 releases, stays released for about 1 day (more info)

15jun05

17jun05 - The old method of computing the cals would find a single value for each 25 Mhz band and then fit an Nth order polynomial across the entire band (1120=1720 for lbw). This polynomial was then evaluated at 25 Mhz steps and placed in a lookup table for later use. The lbw cal data had lots of rfi (on the sky) and a large ripple in polA with a period of around 75Mhz. The old method did not do a very good job of fitting the data so i switched to a new method that fit the entire bandpass at once.
18jun05 - distomat 4 stopped working at 13:00 (sat). Stayed off the air until 20jun05 (10am monday). Luis murray went down and cycled the power on the main box for distomat 4 and it started working. There was no rain storm at the time it failed since the other distomats remained functional. If all that is needed is to cycle the power on the distomat, then this is something that an operator could do. It probably would be a good idea for someone to understand why cycling the power should make a difference??? For 2 days the tracking of the telescope was not "optimal" since the program switched to using the temperature sensor (and model) rather than the distomats.
21jun05 - transfered power from generator to public power in the morning. After the switch, the 4 synthesizers that drive the iflo (the original 4) would not function. The "rf off" label was selected which implies that the synthesizers saw a power dip. Turning the rf back on made no difference. The gpib controller could not communicate with the 4 synthesizers. Turning all 4 of the synthisers off and then on caused them to start working again. It looks like the ups for these 4 synthesizers is not holding up under the power dip. The synths are getting a power glitch and they are coming back in a funny mode where the gpib communications is hung up. We need to check the ups to make sure it is large enough (and that it's batteries are ok).
21jun05 - finished reducing the lbw cals. Installed them in the Tsys area and the idl data directory. The cal for polA looks like it has more ripple than it used to (about a 1.2 meter cable). (more info: lbwcals).
21jun05 - polA lbw has a resonance at 1400 Mhz. It is about 30% larger than tsys. The tsys values that have been taken are reading 35 K. Moving to 1403 Mhz, the Tsys is only 29-30K. The

08jun05

10jun05 - measured lbw cals sky/absorber. Saw a punta salinas like radar around 1305 Mhz. Punta salinas says it was not them and it was a bit off of their frequency. Also took cband absorber data for cband cals (so cheo didn't have to come in monday morning at 6am).
11jun05 - riccardo reported large birdie (alfa) at 1428Mhz for the may05 a2010 data. (see 1428 Mhz birdie). I've asked riccardo to send me the scan numbers for his bad data. When i look at it I see the large birdies but i don't see it spreading over all of the band.
11jun05 Riccardo reported a wandering birdie n alfa in jun02 data. I looked and it is there. Need to investigate this when we get some telescope time.
13jun05 - took cal lband sky data over again since radar at 1305 was gone. Also took lband data to measure ratio of other cals to High correlated cal. Took cband sky data with high correlated cal (but it was cloudy and drizzly so probably not that great). Had no time to do the ratio of cband high correlated cal to the other cals. Looked at lbw cal data and the absorber data was a bit unstable at the beginning. Probably because the absorber was removed from the air conditioned room right before the measurement started. The cal measurements need more that 2 hours to get the most out of the time (so we can to cal, absorber and measure the ratio of hcorcal to other cals at the same time ... and so phil doesn't have to get up at 5:30 am so often).
13jun05 There is something funny going on with the lbw cal data. The lbw cal value for pol B increased by about 1.5 Kelvins after the polB cal cable was fixed. This makes sense and means that the cable was really broken. The Tsys for polB using the old cal values and the new cal deflection is 35 K where it used to by 28 K. If the amount of power in the cal deflection has increased and i am using a smaller value for the cal in K then the Tsys should be less. I did notice that there seems to be a bump in polB at 1400 Mhz that is about 30% of Tsys and about 2 Mhz wide. This needs to be looked into.
13jun05 - installed cband pol switch software. tried testing it but it turns out the cables haven't been connected.

01jun05

02jun05: fractional occurence of rfi for may05 updataed
03jun05: 1am nevis server fails. This is the yp server (providing info on discs, users, permissions,etc for the entire network). There is a backup server but the computers didn't switch to it. Datataking (observer2/observer) hangs, telescope offline.....Spent 1am thru 8am with arun (on the phone from washington dc) getting services moved around so things work. Remote observer got back online about 7am. Arun then worked on the mailserver while i went to bed. Manny found that the problem with nevis was that the 2 cooling fans in the computer had stopped so looks like cpu overheated...
05jun05: updated a2010 monitor data for may05. The problem with the faa radar causing large jumps in the total power and then a long recovery time (many seconds) is still a problem. This is also causing a problem with the alfa continuum group (desh, chris, russ taylor).
07jun05: 16Khz comb in 327 receiver. problem with P1979,a2057. Still looking into this..

25may05

26may05: alfa pixel4a still has 173Khz ripple with amp of .09K. (more info).
28may05: lbw stokes calibration table included data for 4 sources taken jul04,sep04,oct04,apr05 (more info). The delta G parameter shows that the cal values for lbw are not constant. The drift is probably in polB. The yearly plots of TsysA-tsysB also shows a drift in Tsys (TsysA-TsysB (see lbw 2005)).
30may05: distomat 4 loses power 18:00. next day willie finds that it is an ac power filter at the input of the distomat canister that is blown (it wasn't the surge protector after all). He didn't have time to change it because of the sband transmitter.
30may05: chris salter reports a 16 khz comb at 327 Mhz. I took some data on 31may05 and saw the comb close to the 15.85 khz that was measured from the cameras 13oct03 (15.85 Khz comb from cameras). We couldn't easily test the cameras because the switch to power off the cameras in the control room has not been working for a long time. They had trouble making the relay upstairs work so it was bypassed and power was provided directly to each camera. We need to get this fixed. The operators new about this, but it was never communicated to me, rey velez, or anyone one else in electronics (that i am aware of..).
31may05: galfa pixel6a and pixel4a cables were switched at the output of the receiver room distribution amplifiers. It has been switched since at least 02may05. It was not switched on 12dec05 (i've got no data between these two dates). It was fixed at 11am 31may05 (more info).

 home_~phil