What's New in DAPL 2000, version 2.52

Download DAPtools Basic, which includes the current version of DAPL.

New for DAPL 2000, version 2.52

New notations... Clearer configurations
New FFT features... More precision
New FFT window features... More flexibility
New FIRFILTER features... More data types ...see also FGen for Windows 2.0
New data handling features... More data types
DAPL expression extensions... More data types
FORMAT extensions... More expressive
New STATISTICS command... Better monitoring
Commands module... Better packaging
Updated DAPIO32 services... More is better

This DAPL operating system release provides some new software features that can take advantage of the processing power available with the new DAP 5200a and DAP 5400a boards. (Of course, the new command features are available across the product line. It's just that, with the high powered processors, the improved commands really cook.)

Sampling Notations

DAPL input sampling configurations have previously been defined using a command group in the form:

   IDEFINE   NAME  4     // Old and going away...
     other commands...
   END

That dangling "4" means the number of channels in the input channel pipe. Well, sometimes. Things started to get more complicated with the old DAP 3400a and DAP 4400a, for which the number "4" means the number of input channel groups rather than the number of channels. With the DAP 5400a, things become even more complicated, and this trend is expected to continue in the future as even more advanced products are introduced.

We are now moving all input configuration information inside of the configuration command group. All configuration commands appropriate to the Data Acquisition Processor model are accepted there. For a Data Acquisition Processor supporting individual multiplexed samples, the configuration looks like

   IDEFINE   NEWNAME    // New!
     CHANNELS  4        // New!
     other commands...
   END

while for a Data Acquisition Processor model providing simultaneously sampled channel groups, the configuration looks like

   IDEFINE   NEWNAME    // New!
     GROUPS  4          // New!
     GROUPSIZE 8        // New!
     other commands...
   END

While we are in the transition, there is no need to reconfigure existing applications. Existing configurations and existing products will work the same as before. You must use the new notations to access features of the latest new products, however.

FFT

Blazing processor speed rendered the FFT command's "fast 16-bit method" obsolete. Oh yes, it was still very fast, but with the new 32-bit processors, the small speed difference was no longer enough to justify the compromises in accuracy. So now, the "accurate 32-bit method" is the only choice. The FFT32 and the FFT command use the same accurate method.

But for something lost, something better is gained! To preserve even more precision in your transforms, simply provide input and output data streams of a higher precision data type: LONG, FLOAT, or DOUBLE. The FFT command will identify the data type and apply the appropriate transform algorithm automatically.

FFT Windows

For those who design custom FFT windows, or who apply windows to reverse transforms, the FFT command no longer insists on a 16-bit representation. You can specify your window operation using any convenient fixed or floating point data type, and the FFT command will convert the representation appropriately for the processed data type.

FIRFILTER

For processing ultra-quiet data streams sampled by a high precision sampler, you will probably need filtering that preserves the precision of your original data. The FIRFILTER command is updated to select a filtering algorithm according to the type of data stream: WORD, LONG, FLOAT, or DOUBLE. Filter designs must use the same coefficient data type as the data streams, but this should present no problem. The FGen for Windows program can provide filter design coefficients in whatever data type you need.

FIRLOWPASS

If you need lowpass filtering but don't need the complications of a custom filter design, the FIRLOWPASS command is now much improved: more powerful and more accurate. Previously, FIRLOWPASS offered decimation levels of 2, 5, and 10. Now it offers any decimation level from 2 through 12. Plus, the filtering accuracy is extended from 12 bits to 14 bits. Plus, it is faster at each decimation level compared to the previous version.

For those who need more precision, the FIRLOWPASS command also accepts more data types. Present data in LONG, FLOAT, or DOUBLE data stream, and a higher precision filtering algorithm will be selected automatically for that data type. Passband gain accuracy for the higher precision filters is 18 bits signed. That means a 0.00006 dB passband gain error and 102.4 dB stopband rejection.

Data Handling Commands

To facilitate processing with more data types, some fundamental processing commands are updated. These include:

SKIP
COPY
MERGE
SEPARATE
BMERGE
AVERAGE
BAVERAGE

Integrated Command Modules

Taking advantage of the 32-bit command module features introduced in earlier releases of the DAPL 2000 system, this release integrates a number of commands that previously required separate downloading before they could be used. The following commands are now loaded automatically when the DAPL 2000 system is started:

TOGGLE
TOGGWT
TCOLLATE
DEXPAND
COMPRESS
QDCOUNT ( for version 2.52)

An additional note about the QDCOUNT command. Previously, it was a specialty component, a 16-bit binary file that needed to be separately loaded. Now that this command is part of the DAPL Commands Module, your applications do not need to separately download the command. The 32-bit version has a better defined startup process, assuring that there is one output generated for each counter value received, and allowing the starting count and operating mode to be specified independently. Plus, it adds support for low latency response, which can be important for systems that use encoder data for high speed positioning control.

DAPL expression extensions

You need floating point data streams? Or you need to round floating point data streams to integer? No problem. DAPL expression tasks now support conversions between fixed point and floating point data streams. Just assign your data to another data stream of the type you need. Or have the expression stream the data directly to a PC application through a communications pipe in the format that the application needs.

If you need more than simple data streaming, DAPL expressions can do that too. They now support computations with floating point data streams.

FORMAT

No time to develop an attractive (ponderous, unreliable) GUI interface for your application? If you just need to capture a few data sets and go, the FORMAT command might be able to help. It is now able to format streams of FLOAT and DOUBLE data type. This can help you to get the quick report that you need. Use a command line pipe to route the text to your favorite data analysis program, and your application is done.

New STATISTICS command

The aging TASKSTAT command is now replaced with the new STATISTICS command. If you have a demanding application where response time or allocation of CPU resources is important, this new command is what you need. Though similar in purpose to the older TASKSTAT, the STATISTICS command is better in every way. It has better precision, better accuracy, and better output reports.

Application Services

For advanced application developers, the DAPL system has updated some of the configuration query services, accessible through the DapHandleQuery function of the DAPIO32 programming interface.

`DaplSymbol`	Obtain access keys for subsequent inquiries about elements defined in a DAPL configuration.
`DaplInputChannelGroupSize`	Inquire about allowable `GROUPSIZE` in an input configuration on the queried DAP.
`DaplSampleTimeMinAnalog`	Inquire about the minimum `TIME` command interval supported by the queried DAP for analog signals.
`DaplSampleTimeMinDigital`	Inquire about the minimum `TIME` command interval supported by the queried DAP for digital signals.
`DaplWarnNum, DaplWarnMsg`	Obtain the number or text of a warning message queued by the DAPL system.

DAPL 2000, version 2.52, is compatible with the following PCI bus Data Acquisition Processors:

Browse the DAPL Command Summary.

Go to the Software page.

What's New in DAPL 2000 Operating System Release, Version 2.52