Somat eDAQlite, Руководство пользователя, Страница: 188 / 228

Somat eDAQ
lite
188 I2773-4.
4
en
Data stored in the SIF format can be corrupted in the unlikely event of one or more
power failures during a test run when there is insufficient charge in the backup battery.
Understanding that these corruptions do happen, HBM does attempt to evaluate and,
if possible, recover SIF data file contents. This is handled on a case-by-case basis and
is subject to a service charge. The SIF component file set (i.e., all of the SIF#####.SIC
files) must be sent intact. There is very limited hope of recovering data from a
consolidated corrupt SIF file alone.
If TCE reports that any test runs have ended abnormally or there is any other reason
to believe that the SIF data file is corrupt (after using the TCE SIC consolidate task),
proceed as follows.
For internal Flash memory, the only way to copy the eDAQ
lite
resident SIC files is to
use the eDAQ lite web browser. Select the Explore eDAQ Files option in the system
tab and copy the SIC files under the /hd/eDAQ folder to the PC. Note that this can be
very time consuming if there are a large number of SIC files.
10.5 Tips on Eliminating eDAQ
lite
OverFlow Errors
The eDAQ lite generates a DeviceOverFlow or QueueOverFlow error when the
defined test exceeds the eDAQ
lite
unit’s processing capabilities. Following are some
suggestions for test setup changes to avoid a DeviceOverFlow or QueueOverFlow
error.
Use integer data types for transducers
Significant improvement in processing throughput can be achieved by using integer
output data types instead of the floating point data type in the transducer channel
definitions. 16-bit integer data requires only half the storage space as 32-bit floating
point data. The main drawback with integer data types is that they are not supported
by all computed channels and DataModes. There is no loss of accuracy when using
the 16-bit integer data type for any of the data sources that originate from eDAQ
lite
A/D converters since these are all 16-bit A/D converters.
Use Engineering Scaler computed channels
If transducer channels that originate as 16-bit integer data types (e.g., ELHLS and
ELBRG channels) are used in a Desk Calculator computed channel, it is usually more
efficient to use the 16-bit integer data type in the transducer channel definition and
create an Engineering Scaler computed channel to generate the floating point data
needed for most calculator expressions. The main advantage of this is that the
eDAQ
lite can store the original transducer channel as a 16-bit integer instead of as a
32-bit float. For more information on the Engineering Scaler computed channel, see
“Engineering Scaler” on page 133 .
Use the large pipe frame size option
Using the larger pipe frames significantly improves throughput performance
particularly when there are a large number of transducer and/or computed channels
running at slower sample rates (i.e., 500 Hz or slower). Using the larger pipe frames
modestly improves throughput performance for most tests with sample rates below or
around 10000 Hz. At sample rates above 10000 Hz, the small pipe frame size option
generally results in the best throughput performance. This option is selectable on a per
test setup basis using the ELCPU configuration options (see
“Pipe Frame Rate” on
page 90 ).
HBM: public