SRS sells the connector. You might want to make sure you have a tube of
ass-lube handy when you check the price... It's around three times the
distribution price. SRS really should have used a standard D-sub and a couple
of SMA connectors.
I considered laying out a PRBB clone, but didn't
The second LPRO circuit uses a feedback resistor across an inverter... I used
to make oscillators that way... I'm not sure how that works out for a squarer.
They call it a "self-biased" squarer.
My input gate is a 74HC86. It is normally configured as a buffer so a feedback
resistor would add
The next version of Lady Heather supports the Brandywine GPS-4 and compatible
units. If you are running something Linuxy or can built the current Windows
release I can send you the latest source code to compile.
___
time-nuts mailing list --
There is apparently a way to update the Tbolt firmware. Several years ago a
Chinese seller of Tbolts was selling units that he upgraded the firmware from
v2.xx to v3.xx He may have pulled a new firmware chip and dumped the image
and re-programmed the older ones externally or used JTAG,
Wenzel says an HC device tends to work better than an AC device in squarer
applications.
My calibrator board has a place for the feedback resistor so that I can
implement the second LPRO circuit (or add hysteresis to the squarer gate.
> I find it interesting that a simple
I am using the simplest possible sine to square wave converter on my HP5313xA
counter time interval calibrator... a capacitively coupled HCMOS gate (74HC86)
biased at VCC/2 with two 47K resistors as shown in the LPRO manual and Wenzel's
squarer page. I was not expecting anything good, but was
A couple of things to try...
Turn on the OSC plot (G O) and see what that looks like... usually like a lot
of noise. Try with the display filter turned on (like F D 60).
The "&" menu lets you set the disciplining parameters. It will also bring up a
display of all the parameters. You can
It's not anything causing a GPS / tracking outage... the sat count plot does
not drop to 0.
___
time-nuts mailing list -- time-nuts@febo.com
To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
and follow the instructions
I suspect that it is either temperature related (the funkiness starts around
when the temperature reaches a minimum) or related to the way the disciplining
parameters are hacked to get the extended time constant. Try setting up for
say a 10,000 second time constant and see how things change.
Broadcom has released a phone chip that supports L5 signals... claims 30 cm
accuracy.Maybe you will soon be able to use your phone to set your GPSDO
location better...
Also the new iphones now support Galileo in addition to GPS and Glonass.
X72 support in Lady Heather is a new feature not in the current v5.0 release.
If you are running under Linux or can build the v5.0 Windows code, I can send
you the latest version tp compile.
The X72 has a DDS synthesizer on it that lets you vary the output freq in steps
of 2.03E-12 parts.
GPS disciplined rubidium oscillators are generally not a good idea. Rubidiums
tend to be quite a bit more noisy than OCXOs. Their advantage is their long
term frequency stability. The GPS system in a GPSDO compensates for the OCXO
drift, so the only advantage of a GPS disciplined Rb is if
I used my prototype of a single channel of the calibrator's signal generator to
generate and feed a 1.0V p-p 10 MHz square wave into a real HP-59992 J06 that a
list member sent me from Australia for the cost of shipping... many thanks for
that. (Does anybody know how to remove the button tops
I connected a prototype of the HP-531xx calibrator signal circuits (input
squarer, freq doubler, divider, output buffer, level shifter) to a TAPR TICC to
see how stable the output was.
This prototype was built on perf board using DIP packaged 74AC logic so I was
not expecting good results.
And remember that commands sent to the Z3801A end in Line Feed, not Carriage
Return.
___
time-nuts mailing list -- time-nuts@febo.com
To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
and follow the instructions there.
My doubler circuit is just a 74HC86. One gate connects to the input 10 MHz
signal and is used as the squarer / buffer. I'm using the squarer from the
LPRO-101 manual / Wenzel web site. It can be configured as either version
shown in the LPRO manual, or a hybrid of the two.
Two gates are
I finally got in my BNC connectors and 2P6T rotary switches and have the
circuit board for the HP-531xx frequency counter calibrator just about
finished. This board lets you fully calibrate the HP-531xx counters including
the "fine" time interval calibration and the channel gains. The basic
It might be coming here...
http://www.theregister.co.uk/2017/09/06/biggest_solar_flare_in_years_heading_our_way/
You might want to break out your eclipse monitoring equipment...
___
time-nuts mailing list -- time-nuts@febo.com
To unsubscribe, go to
One thing that I have used is Lady Heather's lat/lon scattergram (GI keyboard
command). Connect an antenna, clear the data queue (CC from the keyboard) and
let it run for say 24 hours. Lather, rinse, repeat with different antennas.
Compare the resulting scattergrams and see how "tight"
Real time nuts run phase stable cable (some well over $50 / foot) in climate
controlled ducts... which is all for nought unless you also climate control the
antenna.Which is all standard practice for precision geodesy. Try to keep
it all with a milli-Kelvin or two. Oh, and don't forget
Cheap RG-59 cable coax is more than sufficient for 50 .. 150+ feet (unless you
are doing geodetic level GPS work). It is recommended by several GPSDO makers.
The 50/75 ohm mismatch is not an issue. No need to waste money on fancy pants
artisanal luxury coax.
I connected a Tbolt to the 1PPS input of a Symmetricom SA22.c rubidium
oscillator and ran Heather's software discipling code on it. For some strange
reason, you do get better performance using a clean 1PPS rather than the
crappiest one you can find ;-)
Anyway some interesting results came
The data that RDR showed was from the SA22.c "w" command. I think he was just
using a terminal program.
Attached is a screen dump of Heather doing a software disciplining algorithm on
a SA22.c (using the dev board as an interface). The 1PPS reference that I am
using is the nosiest one that I
Sounds like a serial port driver problem. I started having some issues with
my Prolific adapters in Win XP after installing some EEPROM burner software.
It shows up as garbaged data after typically a day or two of continuos use.
It could also be related to another issue in v5.0 if the Tbolt
I got in a SA22.c rubidium from RDR electronics and one of those interface
boards that appeared briefly on Ebay last week (he had 6 for $7 each and they
went fast... there is another seller that has one for $80). The X72 code that
I added to Lady Heather works with the SA22. I'm still
The Z3801 rollover date depended upon the receiver firmware. It jumped back
in time when it rolled over. Doing the manual date entry fixed it. On the
Lucent boxes you need to be talking to the GPS side.
> Does anybody know when the Z3801A rolled over? Did it keep going
I'm going to buy enough boards to get them for a decent price... way lower than
you can have them built for. Will also probably do them with a kit of parts.
RDR Electronics has SA22's for not much more than the X72. The problem with
SA22's is the connector. It is a dual row 18 pin connector
I think these work like the Z801A where you can turn off the unit, disconnect
the antenna, power up, set the date (and time?), re-connect the antenna, and
it should should recover and remember the date correction.
___
time-nuts mailing list --
Yes, you're missing something... the SA22.c runs on +15V and +5V. The X72
manual says they have 5V versions and versions that can run on 10V .. 30V (but
the HEALTH message reports some values that appear to indicate it alarms or
shuts down at <11.47V or >23.9969).
But who knows? The X72 /
Here's a look at the first cut of the signal interface board for the X72. The
board does not bring out the LEDs. You can pick those off the connector board.
It can also be used with the SA22.c once a board that breaks out the SA22
connector pinouts to a 34 pin cable (8 pins are for the SA22.c
The Mark II (groan)... changed power connector to a 2.1mm barrel jack, added
a 15uF or so filter cap, added a PPS LED, added a separate ground test point,
added series damping resistors for the ACMOS and FXO outputs (can be omitted
to isolate these signals from the cable),
Image of the first cut of the X72 interface board attached...
This board can be used stand-alone or connected to another soon-to-be board
with RS-232, 4xBNC, and 2.1mm power connectors. Hopefully the second board
will also be able to connect to a
Cable has a DB-25 on the other end. X72 uses 26 pins. So logically it would
be pin 26 that is missing... 10 MHz sine wave output... nobody needs that
one now, do they?
> If you have it open, can you tell which is the pin that's missing?
Maybe that one isn't needed and
I'm leaning more towards a two board solution now. A small 0.032" board that
breaks out the X72 connector to a standard 26 pin ribbon cable header and a
larger 0.062" board with the BNC's and RS-232 interface.
This should provide a more robust solution that offers better mounting options.
A few years ago I made a few tweaks to Lady Heather at the request of some
researchers working in Antarctica. I am pretty sure they were working at -90
lat.BTW, Heather can display coordinates in Polar Stereographic Projection
coordinates...
___
There may not be a reason for GPS to not work at the poles... but I have seen
GPS receivers that had hiccups at the equator and know of some that had
problems at the poles... seems to be caused by math degenerating at 0 and 90
degrees.
The equator problem manifested itself as a
Another possibility is some code that attempts to detect a Datum Starloc
receiver. These look like Thunderbolts, but have VERY buggy firmware. If
Heather sees a condition where all satellites are being reported at az/el = 0,
Heather sets a flag that the device is a Starloc and attempts to
Symmetricom sold some interface boards for the X72. They are 0.32-ish inch
thick PCB's with a dual sided edge tab pattern on them. The PCB edge tab
will insert into the connector on the X72. The interface board for my X72's
only breaks out 12 pins... a lot of Ebay X72's come with that 12
The US has artillery shells, mortar rounds, etc that can home in on GPS
spoofers and, uhh, "turn them off".
___
time-nuts mailing list -- time-nuts@febo.com
To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
and follow the
Looking at the log file, it looks like the receiver is sending garbage data.
I have some USB dongles with a (supposedly) Prolific PL23xx chip in them.
After, typically a day or two, they start sending corrupted data. I think the
problem showed up after installing some program that updated
I wonder if they have tested their system with a Thunderbolt-E? The -E is not
necessarily a drop-in replacement for the original.
There are a few differences between the original Tbolt messages and the -E
messages. Some messages are different or unsupported and I know of a couple
that
Lady Heather has pretty much all the infrastructure to do it including the
ability to echo receiver data to another port. It currently echoes data as it
comes in without even looking at it. You would need to add a few dozen lines
of code to do the echo on a packet-by-packet basis and patch
Answer to second question: For GMT time display set the time zone name to GMT
with a time zone offset of 0. (TZ keyboard command TZ GMT, command line
option /tx=GMT).
First question: Yes Heather can set your system clock (assuming the program
has access privileges to the change the clock.
At one time one of the Chinese sellers of Tbolts was upgrading their firmware
from 2.x to 3.0 I don't know how they did that... perhaps they extracted a
3.0 image via something like JTAG or pulling a chip and dumping it.
--
> Unfortunately the unit that you have does not
Lower drain means a smaller battery or backup source... and in today's world
of electronics smaller is better. But, past a certain point, it all boils
down to a "spec waving" contest ;-)
> What's the motivation for this, other than "because we can"? Aren't
existing RTC
A friend of mine is an engineer for one of the biggest manufacturers of clock
chips and has worked quite a bit on their clock chips and is quite familiar
with the issues of building consistent ultra low power oscillators in a
production product. Getting nanowatt (and now sub-nanowatt) level
It looks like it took three hours for the effects of the rollover glitch to
mostly settle out.
BTW, if you only use Lady Heather with a Thunderbolt, you can force the
rollover state from the command line or heather.cfg file by using the /ro
command line option. If you do that you won't have
Attached is a plot of Thunderbolt data before and after the event. 2 seconds
after rollover the Thunderbolt reported it was re-initializing the loop filter
and 4 seconds after the event it reported is was starting to phase lock the
1PPS. The DAC jumped 0.023V V which is around 75 mHz of
Does anybody have a manual for the X99?
BTW, a few days ago the Symmetricom web site was still active. It now
re-directs to Microsemi. It looks like what meagre information was available
is no longer available...
___
time-nuts mailing list --
I'm in the process of adding support to Lady Heather for the Symmetricom X72
(and SA22) rubidium oscillators. I have a most of the functionality working.
The X72 has a "health" message that dumps a couple dozen values. The values
are labeled with some rather cryptic names that give some hint
On macOS and Linux, a program that messes with the time needs root privileges.
Try running Heather via the sudo command. Then issue the TS keyboard
command. If you hear a beep after a few seconds, that verifies that Heather
can set the system time. Automatic time sets don't beep...
Note
I am in the process of adding X72 support to Lady Heather. The X72
documentation (and the protocol used) is rather horrendous. The available
docs are rather old and crusty. Does anybody have anything newer than the
stuff on Symmetricoms web site... particularly for the latest firmware that
And once you get it working... Lady Heather speaks to Acron Zeits! You should
configure the clock to UTC mode or lie to Heather about the time zone offset...
Heather wants the input device to send UTC or GPS time. The satellite map
shows a single "satellite" near the horizon at the azimuth
Does anybody know why it is taking so long to commission the last batch of 4
sats launched at the end of last year? It took them 6 months to enable the
first two. The other two are still unavailable.
___
time-nuts mailing list -- time-nuts@febo.com
I did a simple test run. driving the 1PPS input with a Tbolt.. over 2 hours
the reported PRS case temperature varied 0.8 C. The TT time tag varied 60 ns
(75 ns / degree). The FC frequency control word varied the frequency by
0.231 ppt (0.300 ppt / degree). All my high res counters are
I am still getting to know the device. Heather can calculate a least squares
trend line and FFT of any plot (and now also a histogram) but I have not tried
to calculate the temperature sensitivity yet.
Heather's temperature control PID will work with any device that reports a
temperature.
I recently got in a SRS PRS-10 rubidium oscillator and added support to Lady
Heather to control and monitor it.Attached is a plot of it locking to a
Thunderbolt 1PPS. The plots shown are the TT time tag, the FC frequency
control value, and the case temperature. Air conditioning was
Back in the day, I talked with him several times and worked on some issues
with their original GPS board firmware.
Lady Heather actually started out as a program to control their first OEM
boards ($800 each, qty 1... unheard of at the time). It was written around the
time of the first
Lady Heather has the date marked on her calendar (with a note to wear a clean
pocket protector).
Speaking of eventful dates... Trimble Thunderbolt dates rollover on 30 Jul
2017.Heather's automatic date rollover compensation should fix the date
after 15 seconds... hopefully
I told Heather to display the time as Unix seconds and to do a screen dump at
the Magic Moment... which was duly captured for posterity... Some poor souls
are easily amused by such things...
OK, there was a 1 millisecond delay for the dump to trigger. Receiver was a
I recently taught Lady Heather how to calculate real time histograms of the
various plots.
While characterizing the performance of a UCT-8663 DOCXO installed in a
HP-53132A counter (using the Gerry Sweeny board), I enabled the histogram
feature and got a nice plot showing the TAPR TICC
I've also had to replace one. It had a lot of problems acquiring sats. I
believe the problem is that the TCXO on the board has drifted out of range. I
seem to remember seeing a post somewhere about that and there is supposed to be
a fix. Also, isn't the receiver a UT+ 8 channel unit? Or
I did a (rather crude - breadboard) experiment of using an analog vs XOR freq
doubler driving a flip flop. I did not see much difference between the $0.10
XOR doubler vs the $6 analog doubler.
Two $0.10 2P4T PCB mount slide switches would replace the relays. The total
parts cost for the
I think it will take some testing to see if using an analog frequency doubler
or XOR digital doubler will provide the best performance. My gut says the
analog doubler ($6) will work better than the ($0.20) XOR doubler. I hope the
XOR doubler works well.
It would be even better if the
More like $18 for the simple splitter and $38 for the 180 phase shift splitter.
Also, does anybody know if the phase shift splitter shifts the phase on both
outputs or on only one output. Different HP docs say different things.
> The original parts were nothing special!
I think the way the fine cal works by checking the the intervals between four
different edges that a lot of asymmetries in the signals are nulled out in the
software.
How good are 1:2 180 degree phase shifters at exactly shifting by 180 degrees?
At what cost?
Also coax and RF relays cost
Yes, they do show up... but usually for big-ish bucks. I want to build a
small affordable replacement that anybody with a 531xx can have.
My design is currently leaning towards a board with the clock generator and a
5V reference for the gain calibration (they spec 5V +/- 1mV). I was
To do a full and proper calibration of the HP-53131/53132/53181 counters you
need a J06 HP-59992A time interval calibrator... lots of luck finding one of
those. Looking at the manual, it is actually a fairly simple device which can
be greatly simplified for the task of calibrating a 531XX
The option described in the utilities menu changes from US format 9. to
Euro format 9, (i.e. swapping commas and decimal points). What I am
looking for is how to change the display from 10.0 to 10.000,000,000
mode (enabling commas after the decimal point).
>
I recently got in a 53132A with what I think is fairly early firmware (it's
tied up doing some measurements right now so I can't check the version number).
All the photos of the 53132A that I have seen show it with commas separating
groups of 3 digits after the decimal point. Mine does not
Plus, when I was aligning mine, I found several errors and omissions in the
HP manuals... I can't find my notes now. I do seem to remember some of the
dip switch settings were wrong. Also some inconsistencies in the HP- 8082A (?)
signal generator setup. Anyway... it's a pain in the ass
The K04BB device is a great little relatively inexpensive and compact
Thunderbolt monitor.
Another option is to use a Raspberry PI and the 7" color LCD touchscreen along
with the latest Lady Heather code. I've added touchscreen support and some
optimizations to the screen code for better
This is a very common problem... basically the fan blows directly down the
chassis onto the front panel area where a large dust bunny builds its nest.
The switches are gold plated leaf springs that slide on a gold plated circuit
board. They are open to the environment and collect all sort of
I think all the toggle switch does is switch the TX and RX data lines on the
computer between the serial ports on the RFTG-m... so it would need to be a
2PDT ot 4PDT switch (depening upon how you interface it). Some of the Lucent
docs I've seen talk about using the switch and others talk
No, it's not a bug. It's the way it is intended to work. Whenever Heather
sees a command to open/change the serial port, the currently open port (if any)
is closed and the new one is opened.
Heather lets you change the serial port or read a config file on the fly from
the keyboard.
Heather's configuration priority is to process: hard coded defaults, then the
config file, and finally the command line options. This lets you set your
preferred settings in the config file and then override your config file
options from the command line.
> When I
Heather only requires TXD, RXD, and GND. If you want to use the temperature
control feature RTS and DTR.
Most Linux distros have decent USB serial port drivers built in. I tend to use
no-name Chinese PL-2303 based USB dongles... because I have them. They have
worked fine on all the Linuxy
I'm using code from a Fortran program called solid.f that I converted to C
using F2C and modified to use more accurate sun / moon positions. The solid.f
program is based on an edited version of the dehanttideinelMJD.f source code
provided by Professor V. Dehant. This code is an
Lady Heather can run under Linux, macOS, FreeBSD, and Windoze. It runs well
on the PI (2 or 3) and the soon to be released version 6 has support for the
touchscreen and several new devices. I can send anybody interested in testing
the new version the latest source code to build. Contact
I have not measured mine, but I saw a spec of 2A max for each side mentioned
somewhere (would drop considerably after warmup). I am using a 24V 100W supply
that came with my Z3811/Z3812 Lucent boxes. It was already wired up with the
proper connectors for the boxes.
--
> I
I just added the ability to calculate solid earth tides and the vertical
gravity offset due to the sun and moon to Lady Heather. The lat/lon offset is
typically around +/- 60 cm per day.Vertical offset is around +/- 180mm.
Depending upon the day and where you are, the swings are not
GPS receiver antenna alarms tend to be "informational" only. They monitor
the current that the antenna is drawing. Better ones report open, short, or
OK. Others provide just an OK / fault. A few have no antenna monitoring at
all. The RFTG-m reports the voltages at each end of a 100
I use an HP / Symmetricom 58517A 8 port splitter (the 58536A splitter is a 4
port version and the 58535A is a two port version). They have an amplifier
built in. It is powered by any/all of the the connected GPS devices and feeds
power to the antenna. It work very well... I have it driving 7
My 5065A has one of those tick-monsters in it. You can hear the damn thing
two states away (three at night). Luckily it powers up disabled, but once you
enable it you can't stop it. I've only enabled it once... that was enough...
lesson learned.
> I hear they are quite
Of all the GPS devices that Lady Heather supports, only three send the time
code before the 1PPS. The last byte of the time code message arrives the
indicated number of milliseconds from the PPS pulse:
Z38xx and related devices (including Lucent KS): -965 ms
Lucent RFTG-m: -663 ms
Trimble
Another thing to watch out for on processor ADCs is their performance near the
supply rails... the AVR ADCs are particularly entertaining below around 300 mV
(with a 5V Vref).
___
time-nuts mailing list -- time-nuts@febo.com
To unsubscribe, go to
If you run it on a system with an accurately set clock, Lady Heather can
measure the offset (and standard deviation and ADEVs) between the time in the
receiver time code message and when the last byte of the time code message
comes in. It also calculates a histogram of the message offsets.
e
effects of the sawtooth error on the control loop. The Thunderbolt locks the
GPS receiver clocks to the 10 MHz OCXO and does not have any sawtooth error to
compensate for.
---
> Mark Sims, can you comment on the SawT parameter, I assumed being reported by
> the M12 GPS, d
Be careful when using one unit's location to set a different model's
location... particularly the altitude. Some devices report altitude in MSL,
others in AGL... and different units may use different models for the
ellipsoid. You are always better off using coordinates generated by the
So does Lady Heather... the "S" menu control things like self-surveys, entering
fixed position coordinates, and the receiver operating mode. SN will put it
in Navigation (3D) mode. SH will put it in position Hold mode (timing mode).
---
> TAC32 program has a quick setting
I was once tasked with building some building sway monitoring systems. The
People With the Bucks were rather tight lipped about why they wanted to monitor
building sway, but I think this was closely related to the reason... and yes,
tall buildings wobble like a weeble (but don't fall down)
If somebody on the list bought that "for parts" Brandywine GPS-4 off of Ebay,
I can tell you it has something wrong in the oscillator or EFC circuit. It
reports an OCXO failure alarm and DAC voltage at the lower limit.
The oscillator is unmarked. It is around 2"x2"x1.25" I believe it has a
While playing with the RFTG-m GPSO I tried using a couple of RS-422 to RS-232
converters. These converters can be externally powered or "self-powered" by
the RS-232 signals.
Both units worked fine as long as the RFTG was only sending the 22 character
ASCII time code message once per second.
I have finally managed to decode the RFTG-m voltages message and have all the
values except antenna current... that is apparently hidden away in one of the
other receiver messages.
As far as calculating sat positions for the receiver... the main issue is
getting the current almanac and
If you can build the source code I can send you the latest version. Linux is
easy to do. Not many people seem to be able to handle the Windows build but if
you are familiar with Visual Studio (particularly command line builds) it is
easy. Contact me off list for the code.
I have figured
I have Lady Heather working fairly well with the RFTGm's.I used a serial
port monitor program to capture the traffic in and out of the serial port and
used the Lucent control program to set and read various parameters. By
analyzing the captured traffic and comparing the results to what the
I have my RFTG connected and have the Lucent software talking to it. I also
have a (crappy) serial port monitor program (Microsoft portmon) running and
sniffing the traffic. It appears that the control requests and responses are
in what amounts to TSIP format. No idea yet what the contents
I recently got in a 53132A with the standard oscillator. I now have Lady
Heather working with it (along with any counter that can stream readings out a
serial port). Attached is a plot of the performance over a 15 hour period.
Counter set to time interval mode, start=HP-5071A 1PPS,
I also got one in. Unfortunately it talks some proprietary, undocumented
command set. I was hoping to be able to sniff what the Lucent code is doing
and eventually add support to Lady Heather.
How did you connect up to the device to use the Lucent code? The documentation
talks about using
The 1PPS pulse is very narrow. You will miss it if it comes in while parsing a
time message. You need to set up an interrupt handler to that triggers on the
rising edge of the pulse and sets a flag (which must be declared "volatile").
Check that flag in your loop() instead of waiting for a
201 - 300 of 1485 matches
Mail list logo
