Author Topic: A033-ET Riga Timer - experience at Herstmonceux  (Read 152 times)

April 26, 2019, 10:55:54 AM

Matt Wilkinson

  • Administrator
  • Newbie
  • *****
  • Posts: 37
    • View Profile
A033-ET Riga Timer - experience at Herstmonceux
« on: April 26, 2019, 10:55:54 AM »
Hi all

There are now many EvenTech event timers in operation in the ILRS network.  We bought and installed one at Herstmonceux in 2014 and got good/improved performance in comparison to our HxET event timer, which was built inĀ­-house from two Thales Systems timing modules and a clock module. Our A033-ET runs using a server that I coded in C. See here https://cddis.nasa.gov/2015_Technical_Workshop/docs/posters/P4.2_Wilkinson_poster.pdf.

Since then it has been collecting simultaneous laser range measurements, as part of a secondary data stream.  I am now developing this part of our system with the intention of making the A033-ET our primary timer.

Working more closely with the timer i detected occasional 1 second jumps in the reference epoch. This is actually mentioned in the instruments documentation:

Quote
If the initially defined reference time is incorrect, it can be further corrected using the below described procedures of Time monitoring and Setting. Additionally note that successful time synchronization may become incorrect if during 1.5 hours after performing it the A033-ET has not been used either for the measurement or time monitoring (see below). To avoid this case it is recommended to activate the time monitoring during long pauses between measurements or repeat the procedure directly before the next measurement.

At the time, the C server did not put the timer in 'time monitoring' mode during quiet periods, but did run the time synchronisation regularly.  I have since re-coded the server to switch to time monitoring when not taking measurements.  Now in this monitoring mode, I compare every tick to the local PC clock that is kept closely to UTC using the Linux NTP service. If an incorrect epoch is detected, a correction is applied using the 'set parameters' command.  I think this level of monitoring is enough to ensure that the A033-ET works as needed. I still see 1 second jumps in the timer epoch every few days and will continue to monitor this. I will also be watching to see if any jumps occur during measurements.

How does everyone treat this? Do you see any 1 second jumps? How much of a problem is it?

Also, I wanted to upgrade the A033-ET Windows PC from XP to Windows 7.  This caused me and the guys at EvenTech a long series of headaches so I'll describe here my final set-up.  This parallel port card was recommended to me https://www.delock.com/produkte/G_89445/merkmale.html and it was easy to install. There are W7 and W10 64-bit drivers available. EventTech supplied me with a new version of the A003-ET Server application to work in 64-bit Windows.  An important next step was to change the "EPP PORT = XXXXX/YYYYY" line in the param.cnf file.  These are the DEC numbers of the Control Register addresses of the Parallel Port and Extended Port, which are found in the Port tab in the PCI Card properties in Windows Device Manager. These are listed in Hex and should be converted to DEC and DEC+2 for the Extended Port number.

I did all this and still could not connect to the timer.  First I found that Windows was creating a new param.cnf file in a new folder as part of some virtualisation.  It was this file i had to change to set the correct port.  This still did not work and it wasn't until i copied the complete 'A033-ET Server 1' folder out of the 'Program Files (x86)' folder and on the Desktop that i finally got it running.

I hope this is useful and  I'd be interested in the experience at other stations and any tips you might have.

Thanks
Matt


 

« Last Edit: April 26, 2019, 12:05:02 PM by Matt Wilkinson »