In early 2008, an engineer named Kent Falkenstein posted to the time-nuts mailing list, indicating that he was interested in developing a modern control program for the HP 3048A phase noise measurement system.

Although exceptionally powerful and capable of very high measurement performance, the 3048A system is controlled by obsolete computer hardware that's becoming harder by the day to set up and maintain. Having done some tentative poking at the HP 3048A hardware myself without much to show for it, I contacted Kent offline and volunteered to help test any software he developed.

It soon became clear that Kent was serious about supporting the 3048A with a professional-level development effort. He began by creating a modified version of HP's original "Rocky Mountain BASIC" system software that was suitable for use with the HTBasic emulator from TransEra. This version has been available for some time on hparchive.com's Equipment page; you can find it by searching for 3048 and downloading the current release of the "3048A Software Zip Repack" file. It will run for up to 30 minutes at a time on the trial version of HTBasic.

While the RMB release works well, it has the same awkward user interface as the original 1980s-era HP software. In addition, the full HTBasic license required to run the program without interruption costs over USD $500.00. Consequently, Kent began to port the RMB source code to Visual BASIC 6, a language that despite its name bears little resemblance to HP's BASIC dialect. Kent soon proved up to the challenge, delivering his first alpha build in April 2008.

At the same time, Kent travelled frequently as part of his day job. His employer dealt with seismograph sales and support, focusing on clients in the petroleum industry. Kent would spend months at a time in strange ports of call without regular net access, so I often had to wait several months between updates. But he always eventually resurfaced with a new and improved version.

The last build I received from Kent arrived in March 2011. It's a really nice piece of work, easily 95% ready for prime time. There are still some bugs here and there, but all of the mainstream measurement code for PLL and non-PLL-based phase noise measurements works well. As far as I'm aware, all of the instrumentation supported by the RMB version is supported by PN3048.

When I received this build from Kent, he seemed to be on the verge of wrapping up the project successfully... but then he dropped out of sight again. Which, again, was not unusual for him, except this time he seems to have vanished from the face of the Earth. I haven't heard from Kent since August 2011, despite repeated contact attempts. It's possible that something happened that caused him to either lose interest in the project or otherwise be unable to carry on further development -- I simply don't know. In keeping with Kent's willingness to share beta releases with other interested 3048A users and his stated intent to release PN3048 as freeware based on the generous example set by Agilent in permitting him to use the original 3048A.BAS code, it seems appropriate to post the last PN3048 build I received in 2011 here.

Download PN3048 V1.416

Important notes

• PN3048 can run on both 32- and 64-bit versions of Windows, but there's no installer as such. To install PN3048, make a folder somewhere and unzip all files and subdirectories into it, then see the readme.txt file before running the setup batch file. This seems to work well on the Windows XP and Windows 7 machines I've tried, but no guarantees. I recommend backing up your system first.

• When running the setup_x64.bat or setup_x32.bat batch files, it's necessary to launch the DOS prompt with "Run as administrator." If you receive error messages like Component 'IGThreed40.ocx' or one of its dependencies not correctly registered, this is the first thing to check. I've been able to install PN3048 on all current versions of Windows (up to and including Windows 10) with these guidelines.

• Additional notes on VB6 application support in later Windows versions can be found here. TL;DR: it appears that it should be possible to run PN3048 on systems up to at least the Windows 11 era, at least in theory.

• You will need a National Instruments adapter to run the software (genuine, not a clone.) Both USB and PCI adapters have been used successfully.

• If you encounter GPIB connectivity problems, it may be helpful to increase your NI GPIB interface's Bus Timing property to 2 us. Also, try to avoid running with powered-down instrumentation on the 3048A bus segment -- either disconnect the unused gear completely, or turn it on.

• If you encounter lockups in residual 2-port measurements with "Measuring Noise" displayed, try disabling the Voltmeter in the instrument configuration dialog.

• You can (and should) label your traces by clicking on the caption line above the graph, but avoid using characters such as \ / : * ? " < > or | in the label. The caption is used to generate the default filename for the plot's .res file, and PN3048 occasionally fails to save the file properly when they're used.

• I do not have the source code to PN3048 and cannot support it beyond the notes I've already provided. I'm certainly interested in hearing from users, but unfortunately there's little I can do to help with bugs or compatibility problems.

• The latest available documentation is in Seistek_PN3048_page.pdf, which is a copy of http://www.seistek.com as of 21-Apr-12. There's also an older .CHM help file with similar contents, accessible from the program's Help menu.

• In general there's a one-to-one correspondence between both the features and limitations of HP's original RMB software and those of PN3048. As a result, the HP 3048A reference, operator's guide, and calibration manual from the HP Archive link (or the Agilent site itself) apply equally to PN3048. (The 3048A manuals on the HP Archive site have the advantage of being text-searchable, which isn't always true for Agilent's own copies.)