[time-nuts] Lucent KS-24361, HP/Symmetricom Z3809A, Z3810A, Z3811A, Z3812...

Anthony Roby aroby at antamy.com
Thu Oct 23 22:39:18 EDT 2014

My curiosity got the better of me so I ordered these earlier this week and received them today.

I've powered both up and quickly measured the 10MHz output.  I don't yet have a GPS antenna feed that I can connect, so couldn't check that out.  And I need to look into why both of the units have the Fault and StdBy lights illuminated.  I was surprised how compact they are and they weight next to nothing.  And they are very nicely made.  I took the tops off both and took some photos (see http://goo.gl/87e8GG), but have not ventured into unscrewing everything to get to the bottom of the boards.  From the top, I didn't immediately spot anything extra on the board for the 10MHz out.  All the extras appear to be for the GPS, but the underside of the boards may tell a different story.


-----Original Message-----
From: time-nuts [mailto:time-nuts-bounces at febo.com] On Behalf Of Bob Stewart
Sent: Thursday, October 23, 2014 12:20 PM
To: Discussion of precise time and frequency measurement
Subject: Re: [time-nuts] Lucent KS-24361, HP/Symmetricom Z3809A, Z3810A, Z3811A, Z3812...

My units came in today.  What I got appears to be new-in-box.  It's probably the only thing I'll ever get with a blue Agilent sticker on the box.  =)  It has a yellow Symmetricom notice inside the box.
The circuit board appears to be the same on both units, but that says nothing about the firmware, of course.  The REF-1 has an Oncore receiver labeled TM-AB - whichever one that is, small parts to support it, and a TNC connector for the GPS receiver.

The REF-0  is missing everything related to the receiver, and has an SMA for the 10MHz output in the space where the REF-1 has the TNC along with a few extra small parts.  This is a shared space with both SMA and TNC pads, though they don't seem to share the same electrical path.  Since the SMA and TNC share the same physical space, even if the 10MHz is available somewhere, you'd have to do some surgery on the case before you could bring it out.  Probably by adding a hole in the case for the GPS antenna and using the pad space for the SMA.

It will be a day or two before I have the bits to apply power and connect an antenna.  So, that's what I know.  I'd probably just break something if I tried to find and bring out the 10MHz, so I'll have to leave that to someone else.  But, the appropriate signals need to get between the boards, so I wonder what's on the Interface pins?  Maybe just arbitration, 1PPS, and sawtooth comms?
In my case, I do need the 10MHz, so I'm just as happy to have bought both units at this point.  Maybe, down the road, someone will come up with the mods to convert a REF-1 into a REF-0, and vice versa, unless the firmware prevents that.

     From: GandalfG8--- via time-nuts <time-nuts at febo.com>
 To: time-nuts at febo.com
 Sent: Tuesday, October 21, 2014 5:59 AM
 Subject: Re: [time-nuts] Lucent KS-24361, HP/Symmetricom Z3809A, Z3810A, Z3811A, Z3812...
It seems from the auction revision table that this seller has been offering  these for some time, so perhaps another "hidden" gem:-), but it's  perhaps also worth noting that if this system functions on similar principles to earlier RFTG kit then the GPS conditioning is only applied to the unit actually  containing the GPS module, with the other unit intended as  a standby should the first one fail.
In other words, unless the system redundancy is really required most users would probably only need the GPS based unit, or would at least be  better off buying two of those for the same money that the "matched" pair  would cost.
The only advantage, as far as I'm aware anyway, of the non-GPS unit is that  it contains a 10MHz output.
However, Skip Withrow published modification details in January 2013 showing how straightforward it was to add the the 10MHz output, to the RFTGm-II-XO module, the PCB location for the socket was already available, so I would suspect it wouldn't be too difficult on these either.
In a message dated 20/10/2014 05:53:29 GMT Daylight Time, stewart.cobb at gmail.com writes:

Fellow  time-nuts,

This (long) post is a review of the HP/Symmetricom Z3810A  (or Z3810AS) GPSDO system built for Lucent circa 2000.  I wrote it  because I looked for more information before I bought one, and couldn't  find much.
It's relevant because (as of this writing), you can buy a full  system on the usual auction site for about $150 plus shipping.  For  those of you lamenting the dearth of cheap Thunderbolts, this looks like  one of the best deals going.  The description of these objects does  not include "GPSDO", so time-nuts may have missed it.  Search for one  of the part numbers in the subject line and you should find it.

So  what is it?  It's a dual GPSDO built by HP as a reference (Redundant  Frequency and Time Generator, or RFTG) for a Lucent cell-phone base  station, built to Lucent's spec KS-24361. Internally, it's a close cousin  of a later-model Z3805A.  Externally, it looks to be almost a drop-in  replacement for the earlier RFTG system built to Lucent's spec  KS-24019.  That was a redundant system containing one rubidium (LPRO,  in the one I have) and one OCXO in two almost-identical boxes.  That  spec went through several revisions with slightly different nameplates and  presumably slightly different internals.  You can generally find one  or two examples on the auction site (search for RFTG or  KS-24019).

This system is similar, but the two boxes each contain a  Milliren
(MTI) 260-0624-C 5.000MHz DOCXO, and neither contains a  rubidium.  The Milliren DOXCO is the same one used in the later models  of the HP Z3805A / 58503A.  It's a very high-performance DOCXO, in the  same class as the legendary HP 10811, and better than the one in  most surplus Thunderbolts.  The 5 MHz output is multiplied up to 10  MHz in at least one unit, and 15 MHz in both units.  I don't have the  ability to measure phase noise on these outputs, but I'd be interested to  see the results if someone could.

Nomenclature:  The Z3810AS  (there always seems to be an "S" at the
end) is a system consisting of the  Z3811A (the unit containing a GPS receiver), the Z3812A (the unit with no  GPS receiver), and the Z3809A (a stupid little interconnect cable).  The GPS receiver inside the Z3811A is a Motorola device, presumably some  version of an OnCore.
Where the Z3811A has a TNC GPS antenna input, the  Z3812A has an SMA connector labeled "10MHz TP".  That is indeed a 10  MHz output.  It comes active as soon as power is applied to the unit,  and its frequency follows the warmup curve of the OCXO.  The two units  have identical PCBs (stuffed slightly differently), and I have no  doubt that someone can figure out how to add a 10 MHz output to the  Z3811A as well.

Operation:  From the outside, these units are  broadly similar to earlier units in the Lucent RFTG series. The (extremely  valuable) website run by Didier, KO4BB, has a lot of information on  those earlier units, much of which still applies here.  The purpose of  these units was to provide a reliable source of frequency and  timing information to the cell-site electronics.  The 15 MHz outputs  from both units were connected to a power combiner/splitter and directed  to various parts of the transmitter.  The units negotiate with each  other so that only one 15 MHz output is active at a time.  The  outputs labeled "RS422/1PPS" contained a 4800 baud (?) serial time code  as well as the PPS signal, which were sent to the control  computer.

Power is applied to the connector labeled "+24VDC" and "P1",  in exactly the same way as the earlier RFTG units. Apply +24V to pin  1 and the other side of the power supply (GND or RTN) to pin 2.  In these units, that power supply goes directly to an isolated  Lucent DC/DC converter brick labeled "IN: DC 18-36, 1.9A".  Presumably  you can run both units with a 4-amp supply.

Once you have applied  power, connect the Z3809A cable between the jacks labeled "INTERFACE J5" on  each unit.  The earlier RFTG units used a special cable between two  DE-9 connectors, and it mattered which end of the cable connected to which  unit.  The interconnect for these units is a high-density DE-15  connector (like a VGA plug).  The Z3809A cable is so short that the  two units need to be stacked one above the other, or the cable won't  reach.  It doesn't seem to matter which end of the cable goes to which  unit.  I don't know whether it's a straight-through cable, or whether  you could use a VGA cable as a substitute.

When you apply power, all  the LEDs on the front panel will flash.  The "NO GPS" light will  continue flashing until you connect a GPS antenna.
Once it sees a  satellite, the light will stop flashing and remain on.
The unit will  conduct a self-survey for several hours.  Eventually, if all is well,  the Z3812A ("REF 0" on its front panel) will show one green "ON" light and  the Z3811A ("REF 1") will show one yellow "STBY"
light.  This means  that the Z3812A is actually transmitting its 15MHz output, and the other  one is silently waiting to take over if it fails.

Most time-nuts  want to see more than a pretty green light.  The old RFTG series  allowed you to hook up a PC to the "RS422/PPS" port and peek under the hood  with a diagnostic program.  The program is available on the KO4BB  website.  It is written for an old version of Windows, and I had no  luck getting it to run under Windows 7.  It does run under WINE (the  Windows emulator for Linux) on Ubuntu 12.04 LTS.
To use it, you need to  make an adapter cable to connect the oddball
RS-422 pinout to a  conventional PC RS-232 pinout.  The adapter cable looks like  this:

RFTG          PC

DE-9P  DE-9S

7 <----------> 5

8  <----------> 3

9 <----------> 2

(According to the  official specs, this is cheating, because you're connecting the negative  side of the differential RS-422 signals to the RS-232, and ignoring the  positive side of the differential signals.
However, it's a standard hack,  and it's worked every time I've tried

With that adapter, you  can see the periodic timetag reports from the unit.  The RFTG program  will interpret these timetags when it starts up in "normal mode".  However, when I try to use any of the diagnostic features built into the  program, it crashes WINE.  The timetag output was required for  compatibility, but I suspect that HP didn't bother to implement the Lucent  diagnostics.

Instead, they added a connector which is not on the  previous RFTG series.  That connector is labeled, logically enough,  "J8-DIAGNOSTIC".
It too is wired with RS-422, so you need to use the same  adapter cable as before.  Once you do, you'll find that this connector  speaks the usual HP SCPI command set (Hooray!).  I used the official  SATSTAT program (again under WINE on 12.04 LTS), but I'm sure that  other programs written for this command set will work as well.  The  default SATSTAT serial port settings of 9600-8-N-1 worked for  me.

After about 24 hours, with a poorly-sited indoor GPS antenna,  my system has converged to TFOM=3, FFOM=0 (the best possible  numbers), and a "predicted 24-hour holdover uncertainty" of 5.2  microseconds, which is not too shabby.  It found the correct day and  year without any assistance, so if it has a "GPS week number rollover"  problem, it's still in the future.  I don't currently have the ability  to compare the 10 MHz output to anything else.  Again, if someone  else can, I'd be interested to see the results.

Additional Notes:  The parts on the boards all have date codes of 1998 or 1999.  The  Motorola GPS receiver has a firmware label that reads "02/04/00".  The  SCPI error logs inside the HP units were virgin when I first got  them.  They had 84 and 94 power cycles, respectively.
Before the GPS  receiver acquired time, the error log timestamps read
"2000-05-09  00:00:00", which I interpret as a firmware release date.

The PCB has an  interesting feature.  Next to each soldered-in pin of the Milliren  OCXO is a single-pin socket soldered into the board.  I'm guessing  this was used in manufacturing, to temporarily install a Milliren and  confirm that the system worked before permanently soldering it in.  (At production prices, the Milliren would have cost far more than the rest  of the PCB.)  You might be able to use this in reverse, if you have a  set of Millirens to test from another source.

The Z3809A interconnect  cable has three of the 15 pins on each end clipped a bit shorter than the  rest.  Not so short that they won't eventually make contact, but short  enough to make contact later than the rest.  Don't know why, but it's  clearly deliberate.  A lot of hot-plug connectors are built that way,  including USB connectors.  I have no idea what the pinout of the  interconnect is.

The redundant system slaves both DOCXOs to the same  GPS reference.
Inside the GPS loop bandwidth, the two oscillators will have  almost the same frequency and will differ only by phase noise and  short-term stability.  This is almost a perfect setup for  experimenting with certain kinds of time-nut measurements, assuming someone  can figure out how to get 10MHz out of the Z3811A unit.  If you then  command both units into holdover, you could measure longer-term stability  as well.

The units are described as "new in factory sealed box".  After an archeological investigation of the various strata of labels and  tape on the boxes, I would say that's probably accurate.  My set seems  to have been shipped from the Agilent factory in Korea to Symmetricom  in Sunnyvale, CA sometime in August, 2000, shortly after it was  built, and remained untouched until I opened it.  I'm guessing it was  built and saved as part of a spares program for Lucent, and kept  until Lucent decided they didn't need spares any more.

I have no  connection with the current seller of these units (or any other sellers,  for that matter) except as a satisfied customer.  I think I'll order  another set as a spare, before the feeding frenzy hits.

Request for  help:  Both the SatStat and RFTG programs run under WINE on stock  Ubuntu 12.04 LTS (32-bit) without any tricks or  special configuration.  Neither seems to run under WINE on Ubuntu  14.04 LTS (64-bit). I am a WINE novice.  Any hints from WINE experts  would be appreciated.  Also, I've been able to run TimeLab under WINE,  but I can't connect it to my USB-to-488 interface, so I can't take  data.  If anyone can tell me how to set that up, I'd be extremely  grateful.

time-nuts  mailing list -- time-nuts at febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
and follow the  instructions there.

time-nuts mailing list -- time-nuts at febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
and follow the instructions there.

time-nuts mailing list -- time-nuts at febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
and follow the instructions there.

More information about the time-nuts mailing list