Hi Ulrich,
I agree with your comments. I've seen the issues that can occur when
serial echo is set wrong.
However, the GPSTCXO has an additional issue. Serial echo can actually
cause the unit to reboot. This means that it loses whatever it's
learned about the oscillator and reverts to it's default values.
Performance suffers until it has relearned the oscillator's unique
characteristics.
Ed
On 4/15/2012 6:13 AM, Ulrich Bangert wrote:
Ed,
- If you use Z38XX to communicate with the GPSTCXO be sure to disable
serial echo as described in the manual. The GPSTCXO _really_ doesn't
like serial echo. It can cause the unit to reboot.
as the author of the mentioned Z38XX I would like to clarify this claim. If
serial echo is enabled then the GPSTCXO is the SOURCE of the serial echo and
echoes every single character sent by Z38XX. This worries Z38XX a lot
because it expects to put forward a question to the device and then to
receive an answer to that question. It does not expect to receive its own
question and for that reason it takes the echoed question as the answer
sometimes.
The problem is the "sometimes" because the next time the correct answer may
be read. Whenever you get the impression that the communication with a GPSDO
is kind of "unstable" with Z38XX you should give the serial echo the first
check. That applies to ALL GPSDOs and not only the GPSTCXO. The GPSTCXO was
the first device to notice this fact because the default for the serial echo
seems to be "On" which is a handy feature when you use a common terminal
program for communication.
Best regards
Ulrich Bangert
-----Ursprungliche Nachricht-----
Von: time-nuts-boun...@febo.com
[mailto:time-nuts-boun...@febo.com] Im Auftrag von Ed Palmer
Gesendet: Sonntag, 15. April 2012 07:18
An: Discussion of precise time and frequency measurement
Betreff: Re: [time-nuts] Opinion Of Attached - Possible HPSDR
GPSDO 'Engine'
I recently had a chance to look at the GPSTCXO. I was quite
impressed.
It's performance is a lot better than any bare GPS receiver,
but maybe
not as good as some GPSDOs.
Here are some measurements that I made on the 1 PPS of the
GPSTCXO and
other GPSDOs:
GPSTCXO .... Std. Dev.< 400 ps. min-max range< 4 ns.
Tbolt ................< 500 ps. ............. ~ 4 ns.
Z3801A ...............< 300 ps. ............. ~ 2 ns.
Z3817A ...............< 100 ps. .............< 1 ns.
These measurements were 1000 pulses or more measured with an
HP 5372A or
5370B for the Z3817A. The Z3817A requires an external 1 PPS
source to
lock to. I used a Navsync CW12-wi. Over longer periods of
time typical
GPS wander is present in all units.
I measured the 10 MHz output of the GPSTCXO with the Histogram Time
Interval function of my 5372A. Out of 200 million cycles,
every one had
a period of either 100.0 or 100.2 ns. There wasn't even one
outlier. I
have OCXOs that aren't nearly that good.
A few random points to remember about the GPSTCXO:
- There's no way to set the time zone offset - fixed at UTC.
- There's no way to set the elevation mask - fixed at 10 degrees.
- The Allen Deviation matched the graph in the manual which
is available
at the Jackson Labs site.
- On the evaluation board, the 1 PPS is wired to the "DCD
lead" on the
USB-emulated serial port.
- I can confirm that, as mentioned in the manual, performance
improves
substantially when you shield the unit from temperature fluctuations.
- If you use Z38XX to communicate with the GPSTCXO be sure to disable
serial echo as described in the manual. The GPSTCXO _really_ doesn't
like serial echo. It can cause the unit to reboot.
The GPSTCXO stands between a bare GPS timing receiver and a typical
GPSDO. It's performance is closer to a GPSDO than to a GPS receiver,
but it's price is closer to the bare receiver than to the
price of a new
GPSDO. It should be useful in many situations.
I have no relationship to Jackson Labs. They did make a GPSTCXO
available to me for this testing.
Ed
On 4/14/2012 7:10 PM, John Westmoreland wrote:
Hello Fellow Time-Nuts,
I am new to the group - and was asked by someone from HPSDR/TAPR to
get an opinion, good/bad, of the attached PDF which I hope doesn't
exceed the allowable byte size of posts.
Jackson Labs has given me a verbal on the attached as sub $300.00 QTY 1. We
are seeing if there is interest in a possible 'bulk' buy.
I will post more on the HPSDR Wiki as soon as I sort out my password - and will
post an e-mail regarding the other files associated with the GPSTCXO.
If there is sufficient interest we may do a spin of the HPSDR
Excalibur project or perhaps just do a new project with the attached device.
It should be easy to interface directly to the Alex Bus in
the OpenHPSDR Architecture.
Thanks and 73's,
John W.
San Jose, CA
AJ6BC (Ham Call Sign)
_______________________________________________
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.
