33rdA nnual Precise Time and Time Interval (PZTl)M eeting A SYSTEM TO COMPARE AND EVALUATE THE QUALITY OF PRECISE FREQUENCY AND TIMING SYSTEMS Werner R. Lange President and CEO Lange-Electronic GmbH Gernlinden, Germany Abstract Larger scientific and commercial sites like military test ranges or satellite ground stations very ofen use several sets of independently operating PTFS (Precise Timing & Frequency Systems). Very often these systems are separated by several hundreds of meters and synchronized to different sources that make it diffiult to compare the quality in terms of on- time accuracy and frequency precision between these independent systems. This paper describes a “tool” to measure and evaluate up to eight independently operating PTFS by measuring the differences between the Ipps-signals and the phase relations of frequencies. This 66t~~ils ’d7e signed for a satellite ground station with six independent PTFS, most of them based on GPS receivers using UTC as time scale. The PTFS are located on a campus; the distance between the systems is up to several hundred meters. The frequency part of the “tool” continuously measures the phase difference between the multiple frequencies of the external PTFS, which must not necessarily be of the same nominal frequency and outputs the data to a PC. The accuracy of the measurement is about IO to 50 ps. The timing part of the system compares the difference between I pps signals with an accuracy of IO0 ps and also outputs the data to the same PC. This unit has some - more functions distribute the information to external sites for monitoring and alarm functions as well as act as an NTP-server. The data derived from the system can be used as well for immediate control, as well for long- term evaluation of the behavior of each independent PTFS. THE PROBLEM At a modern satellite ground station, a scientific research laboratory or a military test range, there are usually a lot of different time and frequency sources to be found, all of them a part of a single system, dedicated to a specific task. Very often it turns out that the “single, specific task” is not so single. Just an example: A satellite ground station contains six different subsystems, each of them controlling a single satellite at a specific position. All of these control systems are delivered at different times and each of them represents the latest technology at time of delivery (better to say, at time of quote, but this does not really matter here). Thus, each of the stations have another generation of computer systems, software and, very often, time and frequency generators. We will not go very far back in time and will assume they are all synchronized to a common time scale by GPS or, perhaps, one of the terrestrial radio systems like Omega, Loran C, DCF 77, or other long-wave radio systems. Comparing the different time and frequency systems, significant differences can soon be found. The differences can be so big that, at one of our customers sites, data which have been sent to a satellite, then transferred to the next satellite and from there to the customer could not been decoded at the final user at the usual speed due to a small frequency difference between the two control stations of the two 37 1 Report Documentation Page Form Approved OMB No. 0704-0188 Public reporting burden for the collection of information is estimated to average 1 hour per response, including the time for reviewing instructions, searching existing data sources, gathering and maintaining the data needed, and completing and reviewing the collection of information. Send comments regarding this burden estimate or any other aspect of this collection of information, including suggestions for reducing this burden, to Washington Headquarters Services, Directorate for Information Operations and Reports, 1215 Jefferson Davis Highway, Suite 1204, Arlington VA 22202-4302. Respondents should be aware that notwithstanding any other provision of law, no person shall be subject to a penalty for failing to comply with a collection of information if it does not display a currently valid OMB control number. 1. REPORT DATE 3. DATES COVERED NOV 2001 2. REPORT TYPE 00-00-2001 to 00-00-2001 4. TITLE AND SUBTITLE 5a. CONTRACT NUMBER A System to Compare and Evaluate the Quality of Precise Frequency and 5b. GRANT NUMBER Timing Systems 5c. PROGRAM ELEMENT NUMBER 6. AUTHOR(S) 5d. PROJECT NUMBER 5e. TASK NUMBER 5f. WORK UNIT NUMBER 7. PERFORMING ORGANIZATION NAME(S) AND ADDRESS(ES) 8. PERFORMING ORGANIZATION Lange-Electronic GmbH,Gernlinden, Germany, REPORT NUMBER 9. SPONSORING/MONITORING AGENCY NAME(S) AND ADDRESS(ES) 10. SPONSOR/MONITOR’S ACRONYM(S) 11. SPONSOR/MONITOR’S REPORT NUMBER(S) 12. DISTRIBUTION/AVAILABILITY STATEMENT Approved for public release; distribution unlimited 13. SUPPLEMENTARY NOTES See also ADM001482. 33rd Annual Precise Time and Time Interval (PTTI) Systems and Applications Meeting, 27-29 Nov 2001, Long Beach, CA 14. ABSTRACT see report 15. SUBJECT TERMS 16. SECURITY CLASSIFICATION OF: 17. LIMITATION OF 18. NUMBER 19a. NAME OF ABSTRACT OF PAGES RESPONSIBLE PERSON a. REPORT b. ABSTRACT c. THIS PAGE Same as 6 unclassified unclassified unclassified Report (SAR) Standard Form 298 (Rev. 8-98) Prescribed by ANSI Std Z39-18 satellites. Our customer lost a lot of money - commercial high data rate links are expensive, we are talking about thousand of dollars within a few minutes. He asked us about a solution that constantly measures the frequency and time difference between all of his “islands,” and we designed the system described here. A POSSIBLE SOLUTION As said above, each of the satellite control centers contains a PTFS which outputs 10 and/or 5 MHz, 1 pps, and RIG-B. We are continuously monitoring the main frequency outputs, usually 10 MHz, and the on-time pulse, which is a 1-pps digital signal. One of the stations is used as dedicated reference station. Phase difference linearly changes with time, as long as the drift rate of two frequencies is constant. Thus, the phase difference of two nominally identical frequencies forms a more or less linear curve. The gradient of the curve is a measure of the ratio of the two frequencies. The six satellite control facilities are located at one campus, but the distances between the single centers is up to several hundred meters. To avoid crosstalk and common mode problems, the 10 MHz (usually a sine wave) is formed into a square wave and both signals, the 1 pps as well, are transferred as balanced outputs via simple CAT 7 twisted pair lines as used in Ethernet or similar network environments. The system contains four parts: the phase comparator part for detecting frequency anomalies, the lpps comparator for detecting timing errors, the computer overhead, and a laptop as a man-machine interface for programming the system and displaying the results. PHASE COMPARISON The phase comparison is done by two four-channel comparator boards supplied by a small German company named K+K. The principle of these boards was described by Kramer and Klische 111. We are using modified boards, because we presently relate to the basic frequencies of 10, 5, and 1 MHz. We have built a phase-lock loop into the input to avoid spurious signals that may arise through the distribution path. - - serial data shiftregister serial data -..--___I data latch A A read / I comm-and ’32 ’12 - period counter counter fx control fraction - 1 T counter analog pulse expander Figure 1. Beat frequency counter (one channel). 372 All frequencies are sampled with a synchronous strobe signal, named “read command” in Figure 1; each of the boards compares four inputs to the MHz reference input, even if the input frequencies 10 are not MHz. The reference input is also used as a clock source of the board, thus assuring all 10 internal frequencies are derived from the same source, thus being phase-coherent. The theoretical solution of the phase difference is 50 ps or better than 0.2 degrees related to 10 MHz. TIME COMPARISON The time comparison boards compare up to eight 1-pps inputs, one of them is used as reference input. The theoretical solution of the timing board is also 50 ps. OVERHEAD The overhead consists of a single board computer with all kinds of interfaces, RS232, Ethernet, USB, etc. Its operating system is Linux. It collects the data from the frequency and time boards described above and forms a data stream to the laptop via a USB port. DATA STORAGE The actual data are derived at a rate of 1 to 20 samples per second. They are stored and displayed for a programmable time (we usually use 60 to 100 seconds), then the data are condensed by a factor of 5 to 30 and stored. In case of a failure or significant anomaly, the actual data several seconds before and after the failure are stored. After a few hours, the data are again condensed and stored in a separate file. This file shall be archived in the customers central system. As already said, the above times are programmable by the customer. ETHERNET CONNECTION We have added an Ethernet access into our Alarm & Control Unit, which is the core of our dual redundant frequency and time systems. So we also have added a program to display the status informations of the external “islands” to get a whole picture of the status of all systems at a single site. We also get “raw” time information via NTP into the laptop, which helps to compare the central station with distant stations via the Internet or fixed networks (this is, of course, a total different quality of comparison, but it is at least an exchange of health and status data and provides a lot of trust for the operations manager). DISPLAY As a display we use the TFT-display of a laptop computer, but any other display can be used, too. As we are using LabView as the data-handling program, we are limited to the possibilities of this program, but we found that this is not really a limitation. 373