[time-nuts] HP Z3805A com port and monitoring questions

[Magnus Danielson]

On 10/22/2012 07:13 AM, Tom Van Baak wrote:
>> Trimble and HP took different approaches on their GPSDO's.
>> Trimble lets you get into things and fiddle around. HP is much more a black box.
>> I have never seen anything that lets you "fiddle' with one of the HP units.
>
> Hi Bob,
>
> I'm glad you asked. True, the Trimble TSIP binary command set lets you fiddle and the HP SCPI command set is much easier to use but more limited by comparison. But there's also a rich under-the-hood layer on the HP units. Spend some time with this file and you'll see what I mean:
> http://leapsecond.com/museum/z3801a/z3801a-bin.txt
> I explored the functionality in some detail during my hp GPSDO DAC dither investigation some years ago. Contact me offline.

I spend time to reverse engineer it, and there is a whole bunch of 
"hidden" features which isn't in the documentation. Among others, you 
can get the list of the pSOS processes.

There's also a pForth in there, so if you kick the right escape sequence 
you get into a Forth prompt, but I don't remember how much of the inner 
gut you get through it.

There is *MUCH* more under the Z38xx hood!

I would need to have pSOS manuals for the 68k variant to get further.

The Trimble approach is a tuneable PI-based PLL. Nothing wrong with 
that. HP went for the Smart Clock approach out of NIST. The Trimble 
approach is a platform where the openness allowed for the customer to 
adapt to its application (and alternative oscillators) while the HP 
approach was to achieve a well-engineered solution. What Trimble did 
right is that they tuned up the GPS receivers oscillator while HP spent 
more on the control algorithms. It's not all, but a quick stab at the 
differences.

Cheers,
Magnus