freebsd-skq/contrib/ntp/html/driver6.htm
1999-12-09 13:01:21 +00:00

254 lines
11 KiB
HTML

<HTML>
<HEAD>
<META HTTP-EQUIV="Content-Type" CONTENT="text/html; charset=iso-8859-1">
<META NAME="GENERATOR" CONTENT="Mozilla/4.01 [en] (Win95; I) [Netscape]">
<TITLE>IRIG Audio Decoder II for Sun SPARCstation
</TITLE>
</HEAD>
<BODY>
<H3>
IRIG Audio Decoder</H3>
<HR>
<H4>
Synopsis</H4>
Address: 127.127.6.<I>u</I>
<BR>Reference ID: <TT>IRIG</TT>
<BR>Driver ID: <TT>IRIG_AUDIO</TT>
<BR>Audio Device: <TT>/dev/audio</TT> and <TT>/dev/audioctl</TT>
<P>Note: This driver supersedes an older one of the same name, address
and ID which required replacing the original kernel audio driver with another
which works only on older Sun SPARCstation systems. The new driver described
here uses the stock kernel audio driver and works in SunOS 4.1.3 and Solaris
2.6 versions and probably all versions in between. The new driver requires
no modification of the operating system. While it is generic and likely
portable to other systems, it is somewhat slower than the original, since
the extensive signal conditioning, filtering and decoding is done in user
space, not kernel space.
<H4>
Description</H4>
This driver supports the Inter-Range Instrumentation Group (IRIG) standard
time distribution signal using the audio codec native to the Sun SPARCstation.
This signal is generated by several radio clocks, including those made
by Arbiter, Austron, Bancomm, Odetics, Spectracom and TrueTime, among others,
although it is often an add-on option. The signal is connected via an optional
attenuator box and cable to either the microphone or line-in ports on a
Sun SPARCstation <TT>/dev/audio</TT> audio codec device. The driver receives,
demodulates and decodes the IRIG-B and IRIG-E signal formats using internal
filters designed to reduce the effects of noise and interfering signals.
<P>The IRIG signal format uses an amplitude-modulated carrier with pulse-width
modulated data bits. For IRIG-B, the carrier frequency is 1000 Hz and bit
rate 100 b/s; for IRIG-E, the carrier frequenchy is 100 Hz and bit rate
10 b/s. While IRIG-B provides the best accuracy, generally within a few
tens of microseconds relative to IRIG time, it can also generate a significant
load on the processor with older workstations. Generally, the accuracy
with IRIG-E is about ten times worse than IRIG-B, but the processor load
is ten times less.
<P>The program processes 8000-Hz mu-law companded samples using separate
signal filters for IRIG-B and IRIG-E, a comb filter, envelope detector
and automatic threshold corrector. Cycle crossings relative to the corrected
slice level determine the width of each pulse and its value - zero, one
or position identifier. The data encode 20 BCD digits which determine the
second, minute, hour and day of the year and sometimes the year and synchronization
condition. The comb filter exponentially averages the corresponding samples
of successive baud intervals in order to reliably identify the reference
carrier cycle. A type-II phase-lock loop (PLL) performs additional integration
and interpolation to accurately determine the zero crossing of that cycle,
which determines the reference timestamp. A pulse-width discriminator demodulates
the data pulses, which are then encoded as the BCD digits of the timecode.
The timecode and reference timestamp are updated once each second with
IRIG-B (ten seconds with IRIG-E) and local clock offset samples saved for
later processing. At poll intervals of 64 s, the saved samples are processed
by a trimmed-mean filter and used to update the system clock.
<P>Infinite impulse response (IIR) filters are used with both IRIG-B and
IRIG-E formats. An 800-Hz highpass filter is used for IRIG-B and a 130-Hz
lowpass filter for IRIG-E. These are intended for use with noisy signals,
such as might be received over a telephone line or radio circuit, or when
interfering signals may be present in the audio passband. The driver determines
which IRIG format is in use by sampling the amplitude of each filter output
and selecting the one with maximum signal. An automatic gain control feature
provides protection against overdriven or underdriven input signal amplitudes.
It is designed to maintain adequate demodulator signal amplitude while
avoiding occasional noise spikes. In order to assure reliable capture,
the decompanded input signal amplitude must be greater than 100 units and
the codec sample frequency error less than 250 PPM (.025 percent).
<P>The program performs a number of error checks to protect against overdriven
or underdriven input signal levels, incorrect signal format or improper
hardware configuration. Specifically, if any of the following errors occur
for a timecode, the data are rejected. Secifically, if any of the following
errors occur for a time measurement, the data are rejected.
<OL>
<LI>
The peak carrier amplitude is less than 100 units. This usually means dead
IRIG signal source, broken cable or wrong input port.</LI>
<BR>&nbsp;
<LI>
The frequency error is greater than +-250 PPM (.025 percent). This usually
means broken codec hardware or wrong codec configuration.</LI>
<BR>&nbsp;
<LI>
The modulation index is less than 0.5. This usually means overdriven IRIG
signal or wrong IRIG format.</LI>
<BR>&nbsp;
<LI>
A frame synchronization error has occured. This usually means wrong IRIG
signal format or the IRIG signal source has lost synchronization (signature
control).</LI>
<BR>&nbsp;
<LI>
A data decoding error has occured. This usually means wrong IRIG signal
format.</LI>
<BR>&nbsp;
<LI>
The current second of the day is not exactly one greater than the previous
one. This usually means a very noisy IRIG signal or insufficient CPU resources.</LI>
<BR>&nbsp;
<LI>
An audio codec error (overrun) occured. This usually means insufficient
CPU resources, as sometimes happens with Sun SPARC IPCs when doing something
useful.</LI>
</OL>
Note that additional checks are done elsewhere in the reference clock interface
routines.
<P>Unlike other drivers, which can have multiple instantiations, this one
supports only one. It does not seem likely that more than one audio codec
would be useful in a single machine. More than one would probably chew
up too much CPU time anyway.
<H4>
IRIG-B Timecode Format</H4>
The 100 elements of the IRIG timecode are numbered from 0 through 99. Position
identifiers occur at elements 0, 9, 19 and every ten thereafter to 99.
The control function (CF) elements begin at element 50 (CF 1) and extend
to element 78 (CF 27). The straight-binary-seconds (SBS) field, which encodes
the seconds of the UTC day, begins at element 80 (CF 28) and extends to
element 97 (CF 44). The encoding of elements 50 (CF 1) through 78 (CF 27)
is device dependent. This driver presently decodes the CF elements, but
does nothing with them.
<P>Where feasible, the IRIG signal source should be operated with signature
control so that, if the signal is lost or mutilated, the source produces
an unmodulated signal, rather than possibly random digits. The driver will
automatically reject the data and declare itself unsynchronized in this
case. Some devices, in particular Spectracom radio/satellite clocks, provide
additional year and status indication in the format:
<PRE>&nbsp;&nbsp;&nbsp;&nbsp; Element&nbsp;&nbsp; CF&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; Function
&nbsp;&nbsp;&nbsp;&nbsp; -------------------------------------
&nbsp;&nbsp;&nbsp;&nbsp; 55&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; 6&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; time sync status
&nbsp;&nbsp;&nbsp;&nbsp; 60-63&nbsp;&nbsp;&nbsp;&nbsp; 10-13&nbsp;&nbsp;&nbsp;&nbsp; BCD year units
&nbsp;&nbsp;&nbsp;&nbsp; 65-68&nbsp;&nbsp;&nbsp;&nbsp; 15-18&nbsp;&nbsp;&nbsp;&nbsp; BCD year tens</PRE>
Other devices set these elements to zero.
<H4>
Performance</H4>
The mu-law companded data format allows considerable latitude in signal
levels; however, an automatic gain control (AGC) function is implemented
to further compensate for varying input signal levels and to avoid signal
distortion. For proper operation, the IRIG signal source should be configured
for analog signal levels, NOT digital TTL levels.
<P>The accuracy of the system clock synchronized to the IRIG-B source with
this driver and the <TT>ntpd</TT> daemon is 10-20 microseconds with a Sun
UltraSPARC II and maybe twice that with a Sun SPARC IPC. The processor
resources consumed by the daemon can be significant, ranging from about
1.2 percent on the faster UltraSPARC II to 38 percent on the slower SPARC
IPC. However, the overall timing accuracy is limited by the resolution
and stability of the CPU clock oscillator and the interval between clock
corrections, which is 64 s with this driver. This performance, while probably
the best that can be achieved by the daemon itself, can be improved with
assist from the PPS discipline as described elsewhere in the documentation.
<H4>
Monitor Data</H4>
The timecode format used for debugging and data recording includes data
helpful in diagnosing problems with the IRIG signal and codec connections.
With debugging enabled (-d -d -d on the ntpd command line), the driver
produces one line for each timecode in the following format:
<PRE>00 1 98 23 19:26:52 721 143 0.694 47 20 0.083 66.5 3094572411.00027</PRE>
The first field containes the error flags in hex, where the hex bits are
interpreted as below. This is followed by the IRIG status indicator, year
of century, day of year and time of day. The status indicator and year
are not produced by some IRIG devices. Following these fields are the signal
amplitude (0-8100), codec gain (0-255), field phase (0-79), time constant
(2-20), modulation index (0-1), carrier phase error (0+-0.5) and carrier
frequency error (PPM). The last field is the on-time timestamp in NTP format.
The fraction part is a good indicator of how well the driver is doing.
With an UltrSPARC 30, this is normally within a few tens of microseconds
relative to the IRIG-B signal and within a few hundred microseconds with
IRIG-E.
<H4>
Fudge Factors</H4>
<DL>
<DT>
<TT>time1 <I>time</I></TT></DT>
<DD>
Specifies the time offset calibration factor, in seconds and fraction,
with default 0.0.</DD>
<DT>
<TT>time2 <I>time</I></TT></DT>
<DD>
Not used by this driver.</DD>
<DT>
<TT>stratum <I>number</I></TT></DT>
<DD>
Specifies the driver stratum, in decimal from 0 to 15, with default 0.</DD>
<DT>
<TT>refid <I>string</I></TT></DT>
<DD>
Specifies the driver reference identifier, an ASCII string from one to
four characters, with default <TT>IRIG</TT>.</DD>
<DT>
<TT>flag1 0 | 1</TT></DT>
<DD>
Not used by this driver.</DD>
<DT>
<TT>flag2 0 | 1</TT></DT>
<DD>
Specifies the microphone port if set to zero or the line-in port if set
to one. It does not seem useful to specify the compact disc player port.</DD>
<DT>
<TT>flag3 0 | 1</TT></DT>
<DD>
Enables audio monitoring of the input signal. For this purpose, the speaker
volume must be set before the driver is started.</DD>
<DT>
<TT>flag4 0 | 1</TT></DT>
<DD>
Enable verbose <TT>clockstats</TT> recording if set.</DD>
</DL>
Additional Information
<P><A HREF="refclock.htm">Reference Clock Drivers</A>&nbsp;
<HR>
<ADDRESS>
David L. Mills (mills@udel.edu)</ADDRESS>
</BODY>
</HTML>