New pics of RFTG-m-Rb, and some comparison details
My RFTG-m-Rb arrived yesterday, so of course it's in pieces on the bench
today. Photos (along with those of the -XO) are at
http://www.febo.com/time-freq/hardware/Lucent_GPSDO/
In looking at the two units, both use the same main and "mezzanine"
board, but there are component differences.
On the main board, the Rb unit lacks U200 and U202 through U205, but
does have a 10MHz output at the front left corner that's missing on the
XO version. The programmable device, U104, is identical in both units,
even though the part number includes "XO". Both units are marked
"104366-001 XO-Ver.02" and have slightly different date codes, both
beginning with 97. Both boards have the same assembly number, but the
XO board is suffixed -001 and the Rb board is suffixed -002. Both are
series AA.
On the mezzanine board, U302, a Burr Brown ADS7807U (presumably an ADC,
though I haven't looked it up yet) is present on the XO version, but
missing on the Rb version. And, programmable device U103 has a
different number. On the XO version it is "105681-001 03/02 Ver 9.0"
while on the Rb version it is "105089-001 Ver .04 9731". There's
another programmable device (U202), but it has identical part and
version numbers on both units.
The only other apparent difference between the two versions of the
mezzanine is that the Rb board lacks the mounting hardware and
connectors for installing the GPS, and of course lacks the GPS antenna
connector. As with the main board, both boards have the same assembly
number but the XO version is suffixed -003 while the Rb version is
suffixed -004. Both are revision J.
If the Rb isn't being disciplined at all, I'm not sure why they bother
including the mezzanine board, unless it's simply to allow for a PPS
output on the RS-485 connector of the Rb unit (ie, it simply passes
through the PPS from the XO unit).
John
On Dec 30, 2006, at 2:24 PM, John Ackermann N8UR wrote:
My RFTG-m-Rb arrived yesterday, so of course it's in pieces on the
bench
today. Photos (along with those of the -XO) are at
http://www.febo.com/time-freq/hardware/Lucent_GPSDO/
I did some reverse engineering on my RFTG-m-RB last night. The
information below is what I've found so far.
Upon power-up the unit will sit for a while, sometimes 10 minutes or
more, presumably allowing the
Rb to lock and settle. During this time there is a 10MHz square wave
output from "TP" but no output from the "10 MHz REF" nor the 15 MHz
connector. The FAULT and NOGPS lights are illuminated.
After the warmup period the FAULT light extinguishes, the ON light
illuminates, and the 10 and 15 MHz connectors become active. The
10MHz REF OUT is squarewave, the 15 MHz is sine.
Note that this behavior is fundamentally identical to the older
stainless box units containing three boards and an FRS-C that have
made their way around Ebay for a year or two. Could the RFTG have
been the replacement for the stainless box units?
Connectors
J5 INTERFACE:
J5-1: Comes high (+5) after the warmup period is complete and the
reference outputs are enabled.
J5-2: Unknown, maybe N/C
J5-3: Unknown but appears connected to something. GPS IN?
J5-4: GPS TX DATA (GPS header, pad 8)
J5-5: Unknown but appears connected to something. GPS IN?
J5-6: Unknown, maybe N/C
J5-7: PPS output from internal source (not GPS)
J5-8: Unknown, maybe N/C
J5-9: GPS PPS OUT (GPS header, pad 6)
Stimuli to GPS pin 8 (TX DATA) appears on J5-4.
Stimuli to GPS pin 6 (PPS OUT) appears on J5-9.
Since GPS TX and PPS show up on J5, I am betting that GPS RX is also
available on this connector. See below.
RS-485/1PPS connector:
J6-1: diff. output - 1 PPS+ from internal source
J6-2: ?
J6-3: ?
J6-4: ? believed to be diff. serial RXD input.
J6-5: diff. output - TXD serial -
J6-6: diff. output - 1 PPS- from internal source
J6-7: ?
J6-8: ? believed to be diff serial RXD input
J6-9: diff output - TXD serial +
Data on J6 TXD+/- is 9600 8N1 and is of the form:
:000000010100000000E004E6
These updates come once per second and a couple of the nybbles change
when it moves from warmup mode to online. The last byte is a CRC
check of some sort as it changes by exactly one as previous nybbles
change from 0 to 1 or 1 to 0.
I didn't bother with a differential-to-RS232 converter, I just ran
J6-9 into a PC serial port RXD.
J6-4 and J6-8 may be the serial input to the 80188 processor.
U101 (the 75174 differential receiver on the mezzanine board) has
some of its inputs connected to J5-4, J5-9, J6-4, and J6-8. I have
not yet traced the 75175 driver next to it.
I have found no outputs on the ALARM connector.
My mezzanine board looks identical to the one in John's pictures
except the GPS 10 pin header and the antenna bias circuitry is
unpopulated. There is +5 present on the header pads where the GPS
would go at at least GPS TXD and PPS make it out to J5.
I don't have an XO unit to experiment with but it looks almost
identical to the RB as John has noted. Like the Z3801 and 3816,
there must be a way for a technician to check the operation of the
GPS at deployment time at the cell site - and maybe to put it in
position-hold mode following a survey.
As the GPS TXD and PPS lines are tied right to J5 on the front panel,
I am betting that J5 provided a craft interface to the GPS. I
cannot, however, find a path for the GPS RXD from the front panel to
the 10 pin header. I watched GPS RXD during power up to see if the
processor on the mezzanine might be sending initialization or status
query commands and there were none. At this time I am wondering if
the mezz. board was only interested in 1 PPS OUT for XO disciplining
purposes, and TX/RX was carried right to the front panel for setup/
testing purposes.
Hope this helps. YMMV!
73 de JIM N8ECI
Has anyone used any of the following signav receivers for timing purposes.
They claim to actually correct the sawtooth error in hardware.
http://www.signav.com/index_files/PDF_Files/New_Brochures/SigNav%20TM3-01%20Brochure%20V1.2.pdf
http://www.signav.com/index_files/PDF_Files/New_Brochures/SigNav%20TM3-02%20Brochure%20V1.2.pdf
The receiver oscillator is disciplined to "eliminate" sawtooth correction.
The 2nd has a 10Mhz 1E-8 stability reference frequency output.
Bruce
From: Dr Bruce Griffiths bruce.griffiths@xtra.co.nz
Subject: Re: [time-nuts] Signav GPS timing receivers
Date: Sun, 31 Dec 2006 15:59:45 +1300
Message-ID: 459727A1.5040100@xtra.co.nz
Bruce,
Has anyone used any of the following signav receivers for timing purposes.
They claim to actually correct the sawtooth error in hardware.
http://www.signav.com/index_files/PDF_Files/New_Brochures/SigNav%20TM3-01%20Brochure%20V1.2.pdf
http://www.signav.com/index_files/PDF_Files/New_Brochures/SigNav%20TM3-02%20Brochure%20V1.2.pdf
The receiver oscillator is disciplined to "eliminate" sawtooth correction.
The 2nd has a 10Mhz 1E-8 stability reference frequency output.
Considering the options, I'd say that this is basically a modernized version of
the Allstar/Superstar receivers. Most probably it is the Zarlink GP4020/GP2015
chipset at work. The other posibility (uNav) does not neatly match frequencies
and the GP4020 has two extra divide counters which makes the clock output on
the TM3-02 very easy to acheive. The GP4020 also provides a handy SPI interface
so hooking up an external SPI DAC for EFC is trivial. The GP4020 has a refined
PPS output compared to the old GP2021, but it is an interesting "hack" ontop of
the old core, something which could be used to enhance older receivers if one
wishes. That way the 7 second beating pattern can be avoided. Naturally, that
7 second pattern is very predictable and could be included in negative sawtooth
output if you where going down that route.
The core clock is at 40 MHz in those, so a stepup from 10 MHz is trivial (and
indeed how it is done). Also delivering 10 MHz and 1 kHz outputs should not
meet any interesting issues at all.
It is interesting how they are targeting the M12+/M12M market. Looks like a
sweet deal. If you look down their webpages, they seem to have put alot of
their effort into the backend processing, including Kalman filtering, doppler
processing to dead-reckoning moving vehicles with too few satellites in view
etc. All quite doable on the old receiver cores as such, but takes bigger
footprint in software. Looks quite interesting.
Randy said earlier that he where going to sell them, but they don't seem to
have appeared on their web. What is your asking for one of these? I'd sure
like to run it against some other stuff here.
PS. My newly arrived Symmetricom 56k locked up this evening. There's now
another GPSDO in the house featuring both Rb (a TEMEX Rb) and OCXO (OSA 8666).
Cheers,
Magnus
On Sat, Dec 30, 2006 at 02:24:32PM -0500, John Ackermann N8UR wrote:
My RFTG-m-Rb arrived yesterday, so of course it's in pieces on the bench
today. Photos (along with those of the -XO) are at
http://www.febo.com/time-freq/hardware/Lucent_GPSDO/
My RFTG-m-XO arrived.
Seems like they use a version of the UT+ timing receiver
R5122U1155, which is an 8 channel version of the Motorola Oncore design
optimized for timing. This 2002 or so version seems to have been
retrofitted to a 1996/97 base design.
Apparently have an external antenna power supply on the mezzanine
board - presumably to allow the 80C188 to read antenna current (or maybe
it supplies 12 volts instead of 5).
--
Dave Emery N1PRE, die@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."
Ah, well. I have also received two Lucent KS-24019 L106B, aka
RFTG-m-XO boxes. The serial numbers indicate 95 and 96 construction.
The assembly numbers, 55761ASSY103902-001 (55761 is the CAGE number
for Efratom) have Rev E, S/N 522 and Rev H, S/N 2385.
The Efratom crystal(?) oscillators are 023005-001, dated 95 and 96.
Google has no hits for that part number.
The GPS receiver has part number 2643686MAB. This gets a few hits
from vendors that will issue a quote for anything, but will not
describe the part or its manufacturer.
I made a power cable after tracing the male DB-9. Both drew 600 MA
from a 12.5 volt battery, started out with No GPS and Fault lights,
then the No GPS light started blinking. There was no GPS antenna on
the TNC connector, so I didn't leave them on. The antenna jack
supplies +5 volts.
No signal appeared at the 15 MHz SMA connector, so I hope it is
disabled until the oscillator is locked. Both of their 10 MHz IN
SMA connectors still had plastic caps.
I bought these units, and two of their RB cousins, hoping to get
rid of the HP Z3801 units and the CS standard. They add $30 monthly
to the electric bill. Now it seems that I was seriously misled, but
that's the way it goes on the net. I may yet figure out how to
discipline the RB from the XO.
Anybody know how to divide by 2/3rds without a microprocessor? :-)
Happy New Year (no qualifications)
Bill Hawkins
-----Original Message-----
From: time-nuts-bounces@febo.com [mailto:time-nuts-bounces@febo.com] On
Behalf Of David I. Emery
Sent: Saturday, December 30, 2006 11:10 PM
To: Discussion of precise time and frequency measurement
Subject: Re: [time-nuts] New pics of RFTG-m-Rb, and some comparison
details
On Sat, Dec 30, 2006 at 02:24:32PM -0500, John Ackermann N8UR wrote:
My RFTG-m-Rb arrived yesterday, so of course it's in pieces on the
bench today. Photos (along with those of the -XO) are at
http://www.febo.com/time-freq/hardware/Lucent_GPSDO/
My RFTG-m-XO arrived.
Seems like they use a version of the UT+ timing receiver
R5122U1155, which is an 8 channel version of the Motorola Oncore design
optimized for timing. This 2002 or so version seems to have been
retrofitted to a 1996/97 base design.
Apparently have an external antenna power supply on the
mezzanine board - presumably to allow the 80C188 to read antenna current
(or maybe it supplies 12 volts instead of 5).
--
Dave Emery N1PRE, die@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."
time-nuts mailing list
time-nuts@febo.com
https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
Arrgh! Make that TWO 12.5 volt batteries.
-----Original Message-----
From: time-nuts-bounces@febo.com [mailto:time-nuts-bounces@febo.com] On
Behalf Of Bill Hawkins
Sent: Sunday, December 31, 2006 1:15 AM
To: 'Discussion of precise time and frequency measurement'
Subject: Re: [time-nuts] New pics of RFTG-m-Rb, and some comparison
details
Ah, well. I have also received two Lucent KS-24019 L106B, aka RFTG-m-XO
boxes. The serial numbers indicate 95 and 96 construction.
The assembly numbers, 55761ASSY103902-001 (55761 is the CAGE number for
Efratom) have Rev E, S/N 522 and Rev H, S/N 2385.
The Efratom crystal(?) oscillators are 023005-001, dated 95 and 96.
Google has no hits for that part number.
The GPS receiver has part number 2643686MAB. This gets a few hits from
vendors that will issue a quote for anything, but will not describe the
part or its manufacturer.
I made a power cable after tracing the male DB-9. Both drew 600 MA from
a 12.5 volt battery, started out with No GPS and Fault lights, then the
No GPS light started blinking. There was no GPS antenna on the TNC
connector, so I didn't leave them on. The antenna jack supplies +5
volts.
No signal appeared at the 15 MHz SMA connector, so I hope it is disabled
until the oscillator is locked. Both of their 10 MHz IN SMA connectors
still had plastic caps.
I bought these units, and two of their RB cousins, hoping to get rid of
the HP Z3801 units and the CS standard. They add $30 monthly to the
electric bill. Now it seems that I was seriously misled, but that's the
way it goes on the net. I may yet figure out how to discipline the RB
from the XO.
Anybody know how to divide by 2/3rds without a microprocessor? :-)
Happy New Year (no qualifications)
Bill Hawkins
-----Original Message-----
From: time-nuts-bounces@febo.com [mailto:time-nuts-bounces@febo.com] On
Behalf Of David I. Emery
Sent: Saturday, December 30, 2006 11:10 PM
To: Discussion of precise time and frequency measurement
Subject: Re: [time-nuts] New pics of RFTG-m-Rb, and some comparison
details
On Sat, Dec 30, 2006 at 02:24:32PM -0500, John Ackermann N8UR wrote:
My RFTG-m-Rb arrived yesterday, so of course it's in pieces on the
bench today. Photos (along with those of the -XO) are at
http://www.febo.com/time-freq/hardware/Lucent_GPSDO/
My RFTG-m-XO arrived.
Seems like they use a version of the UT+ timing receiver
R5122U1155, which is an 8 channel version of the Motorola Oncore design
optimized for timing. This 2002 or so version seems to have been
retrofitted to a 1996/97 base design.
Apparently have an external antenna power supply on the
mezzanine board - presumably to allow the 80C188 to read antenna current
(or maybe it supplies 12 volts instead of 5).
--
Dave Emery N1PRE, die@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."
time-nuts mailing list
time-nuts@febo.com
https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
time-nuts mailing list
time-nuts@febo.com
https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
Bill Hawkins wrote:
Arrgh! Make that TWO 12.5 volt batteries.
-----Original Message-----
From: time-nuts-bounces@febo.com [mailto:time-nuts-bounces@febo.com] On
Behalf Of Bill Hawkins
Sent: Sunday, December 31, 2006 1:15 AM
To: 'Discussion of precise time and frequency measurement'
Subject: Re: [time-nuts] New pics of RFTG-m-Rb, and some comparison
details
Ah, well. I have also received two Lucent KS-24019 L106B, aka RFTG-m-XO
boxes. The serial numbers indicate 95 and 96 construction.
The assembly numbers, 55761ASSY103902-001 (55761 is the CAGE number for
Efratom) have Rev E, S/N 522 and Rev H, S/N 2385.
The Efratom crystal(?) oscillators are 023005-001, dated 95 and 96.
Google has no hits for that part number.
The GPS receiver has part number 2643686MAB. This gets a few hits from
vendors that will issue a quote for anything, but will not describe the
part or its manufacturer.
I made a power cable after tracing the male DB-9. Both drew 600 MA from
a 12.5 volt battery, started out with No GPS and Fault lights, then the
No GPS light started blinking. There was no GPS antenna on the TNC
connector, so I didn't leave them on. The antenna jack supplies +5
volts.
No signal appeared at the 15 MHz SMA connector, so I hope it is disabled
until the oscillator is locked. Both of their 10 MHz IN SMA connectors
still had plastic caps.
I bought these units, and two of their RB cousins, hoping to get rid of
the HP Z3801 units and the CS standard. They add $30 monthly to the
electric bill. Now it seems that I was seriously misled, but that's the
way it goes on the net. I may yet figure out how to discipline the RB
from the XO.
Anybody know how to divide by 2/3rds without a microprocessor? :-)
Happy New Year (no qualifications)
Bill Hawkins
-----Original Message-----
From: time-nuts-bounces@febo.com [mailto:time-nuts-bounces@febo.com] On
Behalf Of David I. Emery
Sent: Saturday, December 30, 2006 11:10 PM
To: Discussion of precise time and frequency measurement
Subject: Re: [time-nuts] New pics of RFTG-m-Rb, and some comparison
details
On Sat, Dec 30, 2006 at 02:24:32PM -0500, John Ackermann N8UR wrote:
My RFTG-m-Rb arrived yesterday, so of course it's in pieces on the
bench today. Photos (along with those of the -XO) are at
http://www.febo.com/time-freq/hardware/Lucent_GPSDO/
My RFTG-m-XO arrived.
Seems like they use a version of the UT+ timing receiver
R5122U1155, which is an 8 channel version of the Motorola Oncore design
optimized for timing. This 2002 or so version seems to have been
retrofitted to a 1996/97 base design.
Apparently have an external antenna power supply on the
mezzanine board - presumably to allow the 80C188 to read antenna current
(or maybe it supplies 12 volts instead of 5).
Bill
If you mean multiply the 15MHz output by 2/3 to generate 10MHz, simply
use asynchronous divide by 3 counter (2 fliplops) to produce a 5MHz 1/3
duty cycle (or 2/3) output then filter out the 10MHz 2nd harmonic
component with a bandpass filter. The 3rd Harmonic (15MHz) will
conveniently be very close to a null (only departure from exact 1/3 duty
cycle and clock feedthrough from the divider prevent a perfect null).
Bruce
On Sun, Dec 31, 2006 at 01:15:12AM -0600, Bill Hawkins wrote:
Ah, well. I have also received two Lucent KS-24019 L106B, aka
RFTG-m-XO boxes. The serial numbers indicate 95 and 96 construction.
The assembly numbers, 55761ASSY103902-001 (55761 is the CAGE number
for Efratom) have Rev E, S/N 522 and Rev H, S/N 2385.
I have two XOs, a L106B Rev H similar to yours I have apart
around me as I type and a L109 (not yet here and disassembled, but L109
photos are on the net).
My L106B has a Symmetricom Technical Service label on it and
like yours is a Rev H board with the 023005-1 Efratom OCXO on it.
The newer units have different part number OCXOs which are
somewhat larger.
The Efratom crystal(?) oscillators are 023005-001, dated 95 and 96.
Google has no hits for that part number.
Yep..
The GPS receiver has part number 2643686MAB. This gets a few hits
from vendors that will issue a quote for anything, but will not
describe the part or its manufacturer.
I have determined that this almost certainly is a Motorola
Oncore UT (not UT+) receiver with flashable firmware. I think 6
channels but I haven't dug quite deep enough yet. I have no clear
notion of what firmware is loaded in the one on my H board yet. There
probably is a way of reading this out via the serial port stuff.
The L109 has a UT+ as in my previous message.
And shoe firmly in mouth, yes I checked - the GPS receivers are
INTERNAL Xtal oscillator versions and appear to be JUST used for the 1
PPS output. (As for my other shoe, I plan to take apart a 58540A to
see more about what it uses for GPS and whether it derives GPS timing
from the 10 mhz OCXO or an internal GPS crystal oscillator - I knew I
should have grabbed that Thunderbolt when I saw it cheap).
It appears the mezzanine board has 80C188 firmware dated 03/03
on my H board - same version number as the other people's, however.
This does imply an upgrade of some type in the 2001-2003 timeframe. I
have yet to read the PROM, but I suppose it will disassemble nicely...
Also studying available photos it appears the whole family is
basically the same design with very few visible differences in the main
board or the mezzanine processor/pll board. Apparently there are L106A
versions, the L106B (Rev H and J) and a L109 - all obviously the same
design with few variations but made between 1995 and 2002. They do
seem to have kept up to date with newer Oncore receivers, however.
I bought these units, and two of their RB cousins, hoping to get
rid of the HP Z3801 units and the CS standard. They add $30 monthly
to the electric bill. Now it seems that I was seriously misled, but
that's the way it goes on the net. I may yet figure out how to
discipline the RB from the XO.
Only time will tell whether they have respectable performance. I
was hoping for a couple of GPS disciplined rubidiums for a DSP based
satellite tracking project and may and may not have gotten sent down the
garden path. Looks like I might possibly have found what I need with a
little effort put in. But I don't see them as Cs replacements.
Anybody know how to divide by 2/3rds without a microprocessor? :-)
The big annoyance is finding a 10 mhz xtal filter to replace the
15 mhz one if one wants to use the output circuitry...
--
Dave Emery N1PRE, die@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."
On Sat, Dec 30, 2006 at 02:24:32PM -0500, John Ackermann N8UR wrote:
On the mezzanine board, U302, a Burr Brown ADS7807U (presumably an ADC,
though I haven't looked it up yet) is present on the XO version, but
missing on the Rb version. And, programmable device U103 has a
different number. On the XO version it is "105681-001 03/02 Ver 9.0"
while on the Rb version it is "105089-001 Ver .04 9731". There's
another programmable device (U202), but it has identical part and
version numbers on both units.
The ADS7807 on the mezzanine board is indeed a 16 bit A/D with
sampler and a 25 us sample rate.
There is also a Analog Devices AD660AR 16 bit integrated DAC and
precision reference and buffer amp with guaranteed monotonicity in 16
bits.
And next to it is a Ref02c precision 5 volt reference and a
AD603 programmable gain amplifier.
And further over is a Maxim MAX154AEWG 4 channel 8 bit ADC with
2.5 us sample time and bus interface for a micro build in.
And a LM324 Quad op amp. And a 4 channel Harris SP720AB
transient protector.
Other than that, I find a XiCOR 2K by 8 EEPROM, a CMOS 32K by 8
RAM, and various glue logic.
--
Dave Emery N1PRE, die@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."