HP 53305A Phase Analysis Software

[maxwell3e10]

I've been poking around in this program that can be still downloaded from Keysight website http://www.ko4bb.com. The program is written in Visual Basic 3 and can be decompiled easily to original code.  Ostensibly it is for the HP53310A modulation domain analyzer, but it actually has commands to also talk with HP5372/5373 counter as well as HP E1740A Time interval analyzer. It works by reading the time stamps from these instruments and converting them to frequency and phase deviations and Allan variance. HP53310A does not have time stamp capability from the front panel, but apparently they can be extracted by using some diag:poke commands to the ucic controller and reading with diag:bre? command. (what does bre stand for?)

I've been debating if its worth trying to recompile the program with modern VISA drivers, as is it currently only works in 16-bit windows and requires really old GPIB drivers. Has anyone used it at some point and found it useful?

[Tony_G]

Interesting - This is listed as demo software on the site so I never bother to try it out.

I'll spin up a VM and give it a run (if possible) - Will let you know.

TonyG

[Tony_G]

Having taken a look, it seems they only have the demo version on the site so you can't actually take measurements.

That said, I probably wouldn't bother trying to bring it up to current tech unless you don't want to do the graphing yourself. I think it'd be better to just use the examples in the Programming Manual to get the raw data and then use modern graphing code (controls or D3.js if you don't like being oppressed by strong typing</joke>).

I ran this under Windows XP as there is apparently a VMWare bug with its Threadripper support and what these older OSes are doing with paging. If you have the unlocked version then I think you should be able to get by with installing the Keysight IO Libraries 17.1 release, turning on 488 and then setting up the appropriate SICL address.

Let me know how that goes. I have a 53310A myself but the bench is still full and I'm trying to get back into the YouTube swing.

TonyG

[maxwell3e10]

I must have used the version from http://www.ko4bb.com, its the same file size but the Keysight version must be disabled.
http://eww.ko4bb.com/getsimple/index.php?id=download&file=HP_Agilent/HP_53310_Modulation_Domain_Analyzer/HP_53310A_Modulation_Domain_Analyzer/HP_53305A_Phase_Analysis_Software_for_HP_53310A.tar.gz


I also have HP53310 and like to get the time stamps to work. Not sure if its easier to get the program to run or to get a debugger and figure out exactly what codes they send with the diag:poke command.

[Tony_G]

I don't think you need to do anything special to get the time as it is based on the pixels and timebase. Section 5 of the programming manual goes into this.

TonyG

[maxwell3e10]

I downloaded version G.02.02 of Keysight/Agilent IO Libraries. As far as I know its the last version that supports 16-bit applications and has sicl16.dll and gpib16.dll files. I am running it under Windows XP on an old laptop, so the only way to hook up a GPIB interface is with USB (which probably won't work but I haven't tried), or find an old LAN-GPIB or RS232-GPIB adapter.

53310A can measure time intervals, but not absolute time stamps. One way we get time stamps is to plug a reference waveform of similar frequency into channel B and measure time interval A->B. That will measure time differences between zero crossings of signal and reference waveform.

[Gerhard_dk4xp]

I also have an 53310A, but I have used it only to see the locking behavior of PLLs up to now.
Watching with interest 

regards, Gerhard

[maxwell3e10]

Maybe NI-488.2 software will work. It appears it can support 16-bit applications, even in Windows XP emulation mode, in version 3.0
https://forums.ni.com/t5/Instrument-Control-GPIB-Serial/NI-488-2-does-work-on-Windows-XP-Mode-within-Windows-7/td-p/1035002/page/6

[maxwell3e10]

I got the software to work on an XP machine with NI-USB-HS adapter and NI-488.2 drivers. The only slightly tricky thing is one has to keep open an NI-MAX window to the instrument, otherwise it gives GPIB error 28 because of some energy-saving feature.

The most useful feature of the 53305 software is that it can directly get time stamps between successive zero crossings. Here is a plot of the time stamps with the input given by its own 10 MHz reference (my unit has an ovenized oscillator). It measures the time stamps at 1MHz, getting every 10th negative-going zero crossing. The time stamp scatter is about 95 psec.

[maxwell3e10]

If anyone is interested, here is the capture of (undocumented) GPIB commands used by the 53305 software to record the time stamps with the 53310A modulation analyzer.

[Tony_G]

Interesting. What were the full diag poke commands? Were they just the same command with different return data?

TonyG

[maxwell3e10]

You can see the full buffer content after each command. It sends some kind of program to the UCIC (Universal Counter Integrated Circuit). The program is mostly the same, but a few parameters change depending on the setup.

Btw, if I measure 100 MHz sine from a Rigol AWG, I get time stamp r.m.s of 27 psec.

[maxwell3e10]

Someone asked about sicl16.dll file that is used for old HP GPIB communication. Here it is attached