LBA_DAS; Some Additional Notes =============================== Bin loading is (view from front) N2 N3 N2 N3 D3 N4 Beware, the slots are not keyed, some modules are physically interchangeable so getting them wrong will cause a nasty brown smell. There needs to be some space above and behind the bin so the modules can vent to ambient. If a fan fails, power off and remove the module, otherwise it may cook. If only one complete N2+N3 pair is operational, put them in the left-most slots, where they constitute IFP 1. There's a floppy disk in with the DAS which contains some sample setups. In practice it's easy to overwrite these accidentally, so it would be wise to use DOS or Windows to make them read-only so they are preserved as a reference point. When you create new configurations these can always be preserved as files with your own names. Apart from these samples we have always used the extension .PRO (for 'profile'). If you have problems we can create new files here and fax them through for you to type up - they are only text files - or just copy the parameters onto the various setup pages and then save as a new profile. We have moves afoot to provide a high-level interface integrated with PCFS, but that's more like moves amillimetre at this stage. As you can see from the "DAS to S2 Interface" document there are some configurations which cannot be recorded without swapping IF cables or connecting the S2's C1 cable into the S2 AUX socket instead of the normal S2 RECORDER OUT socket (and plugging the free end of the kludge cable into the latter). We had hoped for a number of new S2 recording modes to cover these, but this is not to be, so later this year we plan to develop a jockey board to add to the system. This will provide the necessary multiplexing to match our signals into the standard S2 mode set. Fortunately all the important configurations we are likely to need right now are covered already. The screwdriver adjustment on the front of the N2 modules adds delay to the 1PPS signal in 25ns chunks so you don't have to muck around with lengths of cable and a CRO to avoid edge clashes with the 5MHz signal. The monitor circuit which both drives the LEDs and notifies the control program, is driven by the 1PPs signal so if that fails the LEDs freeze (in fact both may turn red), the program will complain and so will the S2! To set DELAY, push in RESET with a blunt thing, slowly turn DELAY with a small screwdriver (2.5 to 3.0mm blade) until you see green flashes. Wait for two flashes in each position to determine its proper state. There will always be 3 adjacent positions giving green (occasionally 4) - choose the middle one and release RESET. nb DELAY is a 16 position switch but is wired modulo 8. You can turn in either direction and there are no stops. The LEDs 5MHz and 1PPS normally flash green at 1pps. If the monitor circuit detects the absence of either signal, or excessive jitter between them, they will latch red. Only RESET or an equivalent pulse from the program will clear them, and then only if the 1pps signal is present. nb.1) For manual reset hold RESET in for a full second. 2) Manual reset restarts the equipment but does not clear the flag in the current program. Do another S)etup. 3) If the program option AUTOS is ON then it will automatically clear flag, reSetup the system and increment ASCNT. Changes made to parameters in the program have no effect until S)etup is executed (and you can chose not to bother). At this point the filter is stopped, a RESET pulse is sent to the N2 modules and all parameters are reloaded. On the next 1pps edge the filter starts again. ie it is not possible to make even the slightest change without disrupting data completely for a few seconds. This is a feature of the development software currently provided. The production software will only disrupt the data flow for major reconfigurations. nb it is worth setting the auto-restart bit in the S2 config file to ease it past these and other glitches. The S2 log file notes such events and is worth examining afterwards. Good Luck! Questions, suggestions, gripes, (praise?) etc. to Dick: Ph 61 2 9372 4262 (4222 = switch) Fx 61 2 9372 4298 dferris@atnf.csiro.au Also check ftp.atnf.csiro.au/pub/people/dferris/lba_das/ occasionally for the latest documents (ignore any thing older than 7/97). ______________________________________________________________________________ file: addnotes.txt Rev 21/7/97