[time-nuts] Z3801A - notes on the hp58533a GPSDO

David I. Emery die at dieconsulting.com
Fri Jan 12 23:26:48 UTC 2007


On Fri, Jan 12, 2007 at 03:29:15PM -0600, Didier Juges wrote:
> Regarding the Thunderbolt, I got some information from the Trimble sales 
> manager for timing products:
> 
> The Thunderbolt was designed and built in-house (of course from parts 
> bought from outside sources). So, the software and IP are Trimble, not 
> OEM'd. Any resemblance with the Symmetricom 58540A unit is pure 
> coincidence (or more likely purposely made to look the same in order to 
> try and capture some of the same market, since the Thunderbolt was there 
> first).

	I stand humbly corrected.  Sorry to have spread such gross
misinformation and mortally embarrassed about it. My information on the
possibility they both had a Japanese OEM origin was based on a web site
I was directed to four years ago by someone else who I thought to be
knowledgeable that very clearly showed a Japanese OEM product that was
very similar (but perhaps itself a knockoff or made to look alike).

	And I have to admit I have yet to carefully dissect either a
58540A (I have a couple of them in service here) or a Thunderbolt (don't
currently own one, but want to based on what I've heard recently) to
reach my own conclusions.

	Which leaves the other question,  from reading the 58540A
literature, manual and other material available I concluded four years
ago that it adjusted the 1 PPS timing to track GPS once locked to GPS by
adjusting the EFC for the OXCO (I quite distinctly remember a statement
being made in the documentation that the 1 PPS and 10 MHz phase
relationship is fixed with no abrupt jumps of one clock tick in the 1
PPS or 10 MHz once locked). I have not dug up the documentation recently
to see if I can find this statement again, but I am pretty sure that it
is true.

	I know the same has been said about the Thunderbolt...

	And without dissecting a 58540A I shall have to say I am
uncertain as to the origin of the 1.023 MHz GPS C/A timing  - it is said
the Thunderbolt derives this from the 10 MHz, and obviously the Motorola
UTs seem to derive this from a 19.xxx MHz TXCO, but I guess I cannot
intelligently comment about the 58540A without myself dissecting it.


> 
> A cursory look at the HP 53540A and Thunderbolt specs, particularly for 
> holdover, will confirm the units are not the same, not even in the same 
> league (Thunderbolt: +/- 1uS/2 hours, HP 53540A: 100uS/1 hour, which is 
> hard to believe)

	On one last point, it is obvious (or was to me 4 years ago) that
the 58540A was offered with quite a variety of different spec OXCOs
depending on the OEM requirements for phase noise and holdover versus
price.   The PCB clearly has provisions for mounting several different
ones.   This would mean that the actual holdover performance of a
particular 58540A would depend on which OCXO it contained.   I don't
know any particular details of what was shipped and how to tell from
model or serial number (looking inside is the best answer I realize).

	As for performance, I am curious myself.   Certainly the
Thunderbolt is better documented in respect to performance, but whether
in fact it actually IS a lot better when GPS locked is something I'd
like to know.   Its firmware does seem to supply a lot more data about
the quality of the lock than then 58540A does - that box is basically a
simple power up and forget box for applications where tweaking for
ultimate timing or frequency performance isn't needed.

	And in a last trailing remark on a not very related topic - I
have convinced myself that the RFTG-M-RBs (the L105As with the LPRO)
very clearly have the C field EFC control pin tied to the same circuitry
that controls the OCXO EFC on the RFTG-M-XOs.    This clearly implies
that they can be locked to GPS - I hope soon to try to discover if this
can be made to happen with a random external 1 PPS or whether it
involves some kind of messages passed back and forth with the XO CPU or
even the UT/UT+ receiver itself.   It may be necessary to disassemble the
firmware to find out...

-- 
   Dave Emery N1PRE,  die at dieconsulting.com  DIE Consulting, Weston, Mass 02493
"An empty zombie mind with a forlorn barely readable weatherbeaten
'For Rent' sign still vainly flapping outside on the weed encrusted pole - in 
celebration of what could have been, but wasn't and is not to be now either."





More information about the Time-nuts_lists.febo.com mailing list