HRC Time Tagging
February 5, 1998
This memo addresses how to time tag the different types of HRC data
that appear in AXAF telemetry.
The HRC data format is fairly simple when compared to ACIS data
formats but it still consists of data sampled both synchronously and
asynchronously and at a variety of different rates.
Some of the HRC data, instrument health and status, will appear in
``Engineering'' portions of the telemetry.
AXAF telemetry format 1 is the HRC observing mode telemetry in which
the HRC places event data, rate data, and housekeeping data in the
24-kbps ``Science'' portions of the telemetry; each type of data goes
in a well defined location within the telemetry major frame.
AXAF telemetry format 2 has the HRC operating in net-in-line (NIL)
telemetry format. In this format the HRC science telemetry appears in
the ``Engineering'' portion at a reduced rate of 0.5-kbps. Each of the
the data items will receive a time tag based on its location within
the telemetry and, for the event data, additional time-tagging
information supplied by the HRC.
2 ``Engineering'' Data
There are several HRC instrument health and status data items that
appear in the ``Engineering'' portion of the telemetry.
Sampling frequencies vary depending on the type of data.
Primary bus voltage and current values and analog versions of the
event and shield rates are sampled four times per major frame.
Secondary supply voltages and instrument configuration status bits are
sampled twice per major frame.
Temperatures are sampled once per major frame.
The location of these data within the telemetry major frame are
determined by TRW and should be obtained from the AXAF IP&CL.
All samples of a given item appear evenly spaced within the
In other words, since the duration of a major frame, tMF, is
32.8 s, items sampled once per major frame appear every 32.8 s, those
sampled twice appear every 16.4 s, and those sampled four times per
major frame appear every 8.2 s.
However, the samples are distributed through the major frame so that
the corresponding sample times between two items sampled at the same
rate are not identical.
We will adopt the convention for time tagging the HRC instrument
health and status items of using the time of the start of the major
frame (or fraction thereof) as the time tag for all of these items.
If TMF is the time of the start of the major frame, vi is
the i-th sample of a given datum within the major frame from a total
of N (0 £ i £ N-1), then the time tag for the sample, Tvi,
is given by:
3 ``Science'' Data
When the HRC has the observing mode telemetry allocation, telemetry
format 1, the 24-kbps ``Science'' portion of the telemetry contains
the HRC's serial digital telemetry.
The repeat cycle for the serial digital telemetry is faster than the
major frame rate; I will refer to these sub-frames as science frames.
There are 16 science frames per major frame.
This telemetry consists of three different types of data: event, rate,
Below are the conventions for time tagging each of these data types.
3.1 Housekeeping Data
All housekeeping data is sampled once per science frame (16 times per
These samples all appear within a contiguous set of words within the
The same algorithm for time tagging these items will be used as
described above for time tagging the instrument health and status
data; we will time tag all housekeeping data with the time of the
start of the science frame.
The time tag, Thi, for the i-th occurrence of a housekeeping item
within a major frame is given by:
where TMF is the time of the start of the major frame and i
is the science frame number (0 £ i £ 15).
3.2 Rate Data
There are two samples of the rate data in every science frame; they
correspond to the rates from the first and second half of the
previous science frame. If Trij is the j-th sample (j = 0 or
1) from the i-th science frame within a major frame and TMF
is the time of the start of the major frame, then the time tag is
Trij = TMF + i||
3.3 Event Data
The times of individual events are asynchronous with telemetry, so the
HRC supplies a hardware generated event time tag.
This time tagging information is generated with respect to the start
of the science frame in which the event occurred (note that this is
not necessarily the same science frame in which the event data
The hardware generated tag consists of the number of ``clock ticks''
since the start of the science frame until the event occurred and the
three least significant bits of the number of the science frame in
which the event occurred (sub-frame tag).
Having only three bits for the sub-frame tag cannot by itself
unambiguously label the science frame in which the event occurred;
however, this can be achieved by also using the number of the science
frame in which the event was telemetered.
If Nticks is the number of clock ticks between the start of the
science frame and the event trigger, ttick is the size of
a clock tick (ttick = 15.625 µs), Ntag is the
sub-frame tag of the event, and i is the number of the science frame
in which the event was telemetered, then the event time tag, Tevt is given by:
In this expression, the quantity isf is determined from
Ntag and i by defining a new variable, n, that is equal
to the three least significant bits of i, then
Note that the quantity isf can be negative as is necessary to
cover the case when the event occurs in the major frame prior to the
one in which it is telemetered.
Tevt = TMF + isf||
4 Next-In-Line Data (NIL)
When the HRC is in NIL-mode, the serial digital data that appeared in
the 24-kbps ``Science'' portion during observing mode telemetry is put
into a special section of the ``Engineering'' portion of telemetry
with sampling at a reduced rate.
The repeat rate of the serial digital data is changed from the science
frame rate to the major frame rate.
4.1 NIL Housekeeping Data
Housekeeping data will be sampled once per major frame and can be time
tagged with the time of the start of the major frame in which they
appear. Using the symbols from the housekeeping data section above:
4.2 NIL Rate Data
The two samples of rate data correspond to the rates within the first
and second half of the previous major frame. Expressed in the terms
from the rate data section above, we have
Trij = TMF + (j-2)||
4.3 NIL Event Data
The actual time tag that is supplied by the HRC for each event is
identical in both the observing and NIL modes.
The event clock tick counter is reset at the start of each new science
frame and the sub-frame counter is reset at the start of each new
Unfortunately, there will be an ambiguity in the time tag because the
sub-frame counter rolls over within the major frame and the HRC's FIFO
buffer can hold enough events to span more than one major frame's
capacity to telemeter events.
We currently should implement the same algorithm for both observing
and NIL mode event time tagging.
More sophisticated means will have to be developed if we wish to
attempt to recover better time tags for the event data in NIL mode.
Dr. Michael Juda
Harvard-Smithsonian Center for Astrophysics
60 Garden Street, Mail Stop 70
Cambridge, MA 02138, USA
Ph.: (617) 495-7062
Fax: (617) 495-7356
File translated from TEX by TTH, version 2.01.
On 10 Mar 1999, 16:40.