13:00 UT Kjell changed the VPol optical transmitter in Ant 10 on Tuesday (Oct 3), and I finally had time to check it. I find that it solved the problem of the rapid fluctuations we had been seeing for some time (present at least since the Sep. 10 X-flare, but possibly well before. However, the levels for Ant10 VPol were low, so I have updated both the FEM and DCM level settings (for all antennas), which I hope will solve the problem. All gain/attenuation calibrations will now be different for data after this time, although it should all be handled automatically.
15:45 UT I checked the Ant 10 delays, and strangely the HPol delay was changed by 3 ns, but not VPol. That means there is an new X vs. Y delay. I updated all of the delays (minor changes to a few antennas, but no other X-Y delay differences). Ant 10 data will be compromised since Tuesday, Oct. 3. I restarted the solar scan at 15:45 UT. I will redo the X-Y delay calibration tonight.
20:05 UT Last night I loaded a new design of Jack's 300 MHz correlator. It did not work--some sort of sync problem in collecting the X packets, but at least all of the packets showed up. I reloaded the 200 MHz correlator and went through the procedure to update the delays again. An overnight test shows the delays are okay. On another topic, Ant 12 has some sort of pinion gear failure, so that it did not track correctly in Az (apparently it was oscillating wildly). Kjell is working with InterTronics about it. It will be out of the array until fixed. This all started on Oct. 10, I believe.
04:41 UT I loaded yet another new design of the 300 MHz correlator. Again it does not work, but in a different way. The sync problem I identified in the last design is gone, but a lot of X packets are missing and the slot number is not matching the accumulation number. I will leave it running in case Jack wants to look at it tonight. We may miss some or all of tomorrow's observations, depending on when I can get the 200 MHz correlator loaded and the delays measured.
14:34 UT I loaded the 200 MHz correlator design, and everything is set up to do a calibration and set the delays. However,there is no source available right now. I have set up the schedule to start on 3C273 at 15:10 UT. We may get onto the Sun in a couple of hours, but not before. I also adjusted the schedule to go back to 3C273 as a phase calibrator. It is about 1 hour west of the Sun now, and 10 degrees north of it, so it should be a usable source again.
14:30 UT I loaded yet another 300 MHz design on Monday night (Oct. 15), and we ran it all day Oct 16. It still has problems, but the sync problem seems to be solved, so we get all packets in the correct subchannels. The off-by-one slot number is still there. We had some rather mysterious pattern of coherence and non-coherence on certain baselines, but the design did not meet timing, so it is possible that this caused the problem. Jack will fix the slot problem, and try to get it to meet timing, and we will try again. The data solar for Oct. 16 will be no good. I have switched back to the 200 MHz design for today, and set the delays this morning, so today's solar observations should be okay.
15:31 UT Jack produced the same design as above, but tweaked to meet timing and also adjust the slot number. I tested it last night and found it to have the same coherence issues, which we find changes on each reboot. Jack did some tests where he zeroed coefficients for slots 0 and 1, and we do see some packet distribution anomalies. This morning, I switched back to the old correlator, and updated the delays at 15:25 UT or so. But the delays were quite large, so I am doing another 3C273 calibrator observation to check them. Unfortunately, I see that the High-Hpol drain current is low on Ant 14, so I guess the X-channels will be no good. This seems to be an intermittent problem--the same one we had some time ago, after which Mark replaced the power cable. I should say that the Low-VPol drain current is also bobbing up and down, which is also a problem we had earlier. I'll have to notify Mark.
12:08 UT This morning the Tohban reported Ant 10 as not stowing. I rebooted the crio (it went into safe mode), then reset it. After booting, I had to issue a sync command to get it to respond to the ACC. Then it still would not stow--some sort of problem with the HA axis. I entered a debug session, and finally got it to work--not sure what I did. It began tracking around 12:07 UT, so there may be 20 min of good data on the refcal, but no more.