Mon., Feb. 13th

We came in to find that our wired network access was no longer working.  EIT came and poked around.  Turned out the story is:  Due to an extended power outage over the weekend, which exceeded the battery-backup window for the network switch in the closet, and it forgot some configuration changes (from the last time we added machines) which hadn't been saved.  That was all straightened out, and I asked Antony to email me the list of IP addresses so that we will have them for reference in the future (which machine is supposed to use which IP address).  Basically, the address range we should use is 168.223.31.5-14.  A couple of IPs have special rules to allow VPN access - .5 for the Xserve & .6 for the scope.

Anyway, back to work.  David is downloading Quartus onto his Mac laptop, which is running Windows under Parallels, so that he can work on Quartus when Ray is on the iMac.  Meanwhile, Ray is ordering a new SATA hard drive for the Acer machine.

Back to server coding.  I corrected the syntax errors that I introduced in my code changes from over the weekend so that I could get to the point of running tests again.  (None of the new code is actually being dynamically exercised yet, however...) Anyway, here's the current output (same as before):


----------------------------------------------------------------------
At Mon Feb 13 17:14:07 2012 + 586 ms opened node0.uart.trnscr transcript...


Mon Feb 13 17:14:12 2012 + 528 ms: < 
Mon Feb 13 17:14:12 2012 + 529 ms: < HOST_STARTING,CTU_GPS,1.9
Mon Feb 13 17:14:12 2012 + 530 ms: < HOST_READY
Mon Feb 13 17:14:12 2012 + 531 ms: < $PDME,21,OK*1B
Mon Feb 13 17:14:12 2012 + 531 ms: < $PDME,22,OK*18
Mon Feb 13 17:14:12 2012 + 532 ms: < $ACK,WIFI_STARTING,v0.18*66
Mon Feb 13 17:14:12 2012 + 532 ms: < $ERR,BAD_CHK,[$GPRMC,212106.014,V,3025.676,N,08417.112,W,0.0,0.0,161211,4.1,W
Mon Feb 13 17:14:12 2012 + 533 ms: < $GPRMC,212107.020,V,3025.676,N,08417.112,W,0.0,0.0,161211,4.1,W*70]*3D
Mon Feb 13 17:14:12 2012 + 533 ms: < $ERR,BAD_CHK,[$GPGGA,212107.020,3025.67587,N,08417.11218,W,0,00,99.0,051.73,M,-29*65]*0B
Mon Feb 13 17:14:13 2012 + 973 ms: < $ACK,
Mon Feb 13 17:14:13 2012 + 974 ms: < *68
Mon Feb 13 17:14:13 2012 + 976 ms: < $ERR,UNK_CMD,*00
Mon Feb 13 17:14:13 2012 + 977 ms: < $ACK,WIFI_READY
Mon Feb 13 17:14:13 2012 + 979 ms: < *6D

I wonder if the PDME 21/22 (turn on TRAIM/POSHOLD) commands are causing the bad checksums, say because we are sending them at the same time the GPS is sending stuff?  I need to take out the code in the CTU firmware that sends them, anyway, because in the new architecture the server will take care of initializing the GPS module.  Did that.

Also added a new "START" command which calls start_counter() which starts the OCXO clock cycle counter.  This command will be sent by the server when it's ready to start the run.   Note to self:  Remember to someday add a PLL to multiply the OCXO frequency so that we can time arrival of the PPS edges more accurately.

Added a command "HOST" to the Wi-Fi script to relay arbitrary command lines from the server to the sensor host through the script.  Still need to test.