Index: firmware/FSC/doc/FSC.aux =================================================================== --- firmware/FSC/doc/FSC.aux (revision 11660) +++ firmware/FSC/doc/FSC.aux (revision 11666) @@ -9,6 +9,6 @@ \@writefile{lot}{\contentsline {table}{\numberline {2}{\ignorespaces composition of the sensor ID - ad exemplum}}{2}} \newlabel{tdl_sensor_id}{{2}{2}} -\@writefile{toc}{\contentsline {subsection}{\numberline {1.3}Sensor Data}{2}} -\newlabel{sec_sensor_data}{{1.3}{2}} +\@writefile{toc}{\contentsline {subsection}{\numberline {1.3}Sensor Data}{3}} +\newlabel{sec_sensor_data}{{1.3}{3}} \@writefile{lot}{\contentsline {table}{\numberline {3}{\ignorespaces sensor data format}}{3}} \newlabel{tbl_sensor_data_format}{{3}{3}} @@ -16,6 +16,6 @@ \@writefile{lot}{\contentsline {table}{\numberline {4}{\ignorespaces FSC commands}}{3}} \newlabel{tbl_FSC_CMDS}{{4}{3}} -\@writefile{lot}{\contentsline {table}{\numberline {5}{\ignorespaces composition of FSC commands}}{3}} -\newlabel{tbl_FSC_CMD_PACKAGE}{{5}{3}} +\@writefile{lot}{\contentsline {table}{\numberline {5}{\ignorespaces composition of FSC commands}}{4}} +\newlabel{tbl_FSC_CMD_PACKAGE}{{5}{4}} \@writefile{lot}{\contentsline {table}{\numberline {6}{\ignorespaces composition of FSCs acknowledgement}}{4}} \newlabel{tbl_FSC_ACK}{{6}{4}} Index: firmware/FSC/doc/FSC.log =================================================================== --- firmware/FSC/doc/FSC.log (revision 11660) +++ firmware/FSC/doc/FSC.log (revision 11666) @@ -1,3 +1,3 @@ -This is pdfTeX, Version 3.1415926-1.40.11 (MiKTeX 2.9) (preloaded format=pdflatex 2011.3.24) 9 MAY 2011 15:24 +This is pdfTeX, Version 3.1415926-1.40.11 (MiKTeX 2.9) (preloaded format=pdflatex 2011.3.24) 16 MAY 2011 08:05 entering extended mode **FSC.tex @@ -133,31 +133,35 @@ {C:/Users/neise/AppData/Local/MiKTeX/2.9/pdftex/config/pdftex.map}] [2] [3] [4] [5] -(E:\FACT\fact.isdc.unige.ch_svn_firmware\FSC\doc\FSC.aux) ) +(E:\FACT\fact.isdc.unige.ch_svn_firmware\FSC\doc\FSC.aux) + +LaTeX Warning: Label(s) may have changed. Rerun to get cross-references right. + + ) Here is how much of TeX's memory you used: 1324 strings out of 494019 16155 string characters out of 3146498 - 80978 words of memory out of 3000000 + 82978 words of memory out of 3000000 4634 multiletter control sequences out of 15000+200000 14064 words of font info for 33 fonts, out of 3000000 for 9000 714 hyphenation exceptions out of 8191 25i,11n,19p,879b,238s stack positions out of 5000i,500n,10000p,200000b,50000s - - -Output written on FSC.pdf (5 pages, 121844 bytes). + +Output written on FSC.pdf (5 pages, 130624 bytes). PDF statistics: - 273 PDF objects out of 1000 (max. 8388607) + 277 PDF objects out of 1000 (max. 8388607) 0 named destinations out of 1000 (max. 500000) 1 words of extra memory for PDF output out of 10000 (max. 10000000) Index: firmware/FSC/doc/FSC.tex =================================================================== --- firmware/FSC/doc/FSC.tex (revision 11660) +++ firmware/FSC/doc/FSC.tex (revision 11666) @@ -53,15 +53,20 @@ \end{itemize} -FSC uses an 8MHz ATmega32L micro controller(MCU) to readout the muxed 24bit sigma delta ADC (AD7719) which is connected to the RTDs. In addition the ATmegas internal 10bit ADC is used to monitor the humidity, LV-voltages and LV-currents. +FSC uses an 8MHz ATmega32L micro controller(MCU) to readout the muxed 24bit sigma delta ADC (AD7719) which is connected to the RTDs. In addition the 10bit ADC, which is part of the ATmegas32L, is used to monitor the humidity, LV-voltages and LV-currents. The muxers on FSC belong to analog devices ADG77xx family, and are controlled by the MCU. -The User has access to FSC via Ethernet Interface, provided by WIZNETs W5100, the younger brother of W5300, which is used on FACTs FAD boards. - -AD7719 and W5100 communicate with the MCU via an SPI bus, which results in the fact, that a lot of user interaction might slow down temperature measurement or vice versa. Additionally the user should be aware, that sigma delta ADCs have a prolonged settling time, when muxed. Redout of all 64 temperature channels takes about 13s, while one single channel might be readout with at a rate of 10Hz. In addition FSC runs a 32bit timer, counting miliseconds since timer init. The timer is initialized to zero, when FSC is powered. User may synchronize FSCs timer to Unix time. +The User has access to FSC via an Ethernet Interface, provided by WIZNETs W5100, the brother of W5300, which is used on FACTs FAD boards. + +AD7719 and W5100 communicate with the MCU via the same SPI bus, which results in the fact, that a lot of user interaction might slow down temperature measurement or vice versa. Additionally the user should be aware, that sigma delta ADCs have a prolonged settling time, when muxed. Redout of all 64 temperature channels takes about 13s, while one single channel might be readout with at a rate of 10Hz. + +In addition FSC runs a 32bit timer, counting miliseconds since timer init. The timer is initialized to zero, when FSC is powered. User may synchronize FSCs timer to Unix time. \subsection{Measuring} \label{sec_measuring} The user may send a command telling FSC, what to measure. FSC will measure it and return the result, right after the measurement was finished. It is possible to tell FSC to measure several sensors at a time. -Users may issue an 'activate' command (see table \ref{tbl_FSC_CMDS}), to specify which channel should be measured during the next measurement phase. In case one is not sure which channel was activated beforehand users may issue a 'status' command, to gather information about FSCs current status. Whenever the user completed the channel activation process the 'measure' commands may be issued. Since the high resolution ADC AD7719 used for resistance measurement and the 10bit ADC used for voltage measurement run independently on different time scales, different 'measure' commands are defined. 'measure resistances' will start a measurement possibly taking some seconds, while 'measure voltages' will be ready almost instantaneosly. -Since a measurement might take some time, FSC generates an answer, once the 'measure' command received, informing the user about the command reception. The results are then submitted in a single message to the user, which is called a telegram. +Users may activate (see table \ref{tbl_FSC_CMDS}) beliebige sensor channels, by writing into the aproriate 'enable register', to specify which channel should be measured during the next measurement phase. +%In case one is not sure which channel was activated beforehand users may issue a 'status' command, to gather information about FSCs current status. +Whenever the user completed the channel activation process the 'measure' commands may be issued. Since the high resolution ADC AD7719 used for resistance measurement and the 10bit ADC used for voltage measurement run independently on different time scales, different 'measure' commands are defined. 'measure resistances' will start the measurement of all activated resistance channels possibly taking some seconds, while 'measure voltages' will start a measurement of the activated voltage channel and be ready almost instantaneosly. +Since a measurement might take some time, FSC generates an answer, once the 'measure' command received, informing the user about the command reception. The results are later submitted in a single message to the user, which is called a telegram. +Since the maximum readout data rate for the MCU from W5100 memory (~100kb/s) is substantially smaller than the data rate over ethernet, is is possible for the user to fill the W5100 receive memory, while FSC is still working on the first command. Although not extensivly tested, the reaction of W5100 to a completly filled memomry is not critical. It will stop acknowledging incoming TCP/IP packets, which ususally causes the sender to keep sending them. Eventually a time out on user side will occur and the connection will look like lost, but actually it was just given up by the sender. During a 'measure resistance' FSC is not checking for incoming commands. So incoming commands will pile up in W5100 input FIFO. Which means, sending a lot of on demand commands will delay FSCs reaction to the following commands consecutivly. The W5100 transmission FIFO can recieve 4kbyte. I am not sure what happens in case this FIFO is completely full, but I guess W5100 will reject incoming TCP/IP packets and the user side PC will constantly keep on (re)sending them. Since a typical command consists of less than 10bytes it is necessary to send a lot of commands without receiving any answer before the FSC FIFO is full. Index: firmware/FSC/doc/FSC.toc =================================================================== --- firmware/FSC/doc/FSC.toc (revision 11660) +++ firmware/FSC/doc/FSC.toc (revision 11666) @@ -3,5 +3,5 @@ \contentsline {subsubsection}{\numberline {1.1.1}RTD switching}{2} \contentsline {subsection}{\numberline {1.2}The Telegram}{2} -\contentsline {subsection}{\numberline {1.3}Sensor Data}{2} +\contentsline {subsection}{\numberline {1.3}Sensor Data}{3} \contentsline {section}{\numberline {2}Ethernet Interface}{3} \contentsline {subsection}{\numberline {2.1}FSC Registers}{4}