Z3716A Serial Port Issues
Hi, this is my first post here.
I have a Z 3816A, at least that's what I think it is, it doesn't have
any markings on it to identify it. It has a serial port (DB-9) on the
back along with two 10 MHz, four 19.6608 MHz and four 1PPS outputs.
It used to work fine but it has sat unused for the last 15 years or
so. When powered up I get a 1 PPS signal out and eventually the GPS
Locked light comes on. I get some sort of data from the serial port
once a second but it is unreadable. I captured the output with my logic
analyzer but no matter what settings I try it can't decode the data
stream. It appears to be sending at 9600 baud as the shortest length
pulses in the data are about 104.8 uS long. I have tried getting
GPS.exe, Lady Heather and satstat.exe to talk to the unit. All of my
attempts to communicate with the unit have failed.
Of course I can't find a manual for it but I did find the HP 58503
manual which is supposedly similar. My understanding is that there is
no way to reset the serial port to the factory settings. Without the
serial port working I really have no way of knowing the status of the
unit. So, this is a long shot, does anyone have any ideas how I can get
this unit working or is it destined for the recycling bin?
Terry
If the lock light is coming on, that's 90%, as it's perfectly useable
without
the communications. And if you're getting the pps that's getting closer to
100%. Are you sure your computer port is configured correctly? Maybe
it got changed for some reason, blip on the power, etc. I've run into the
same problem with the three I have and have always found that something
changed, either I changed the Z3801A communications config myself(for some
dumb reason or other) or the config in the computer got changed by me,
by windows(imagine that??) or by another serial program.
John K0GCJ
On Sun, May 25, 2025 at 5:00 PM Terry via time-nuts <
time-nuts@lists.febo.com> wrote:
Hi, this is my first post here.
I have a Z 3816A, at least that's what I think it is, it doesn't have
any markings on it to identify it. It has a serial port (DB-9) on the
back along with two 10 MHz, four 19.6608 MHz and four 1PPS outputs.
It used to work fine but it has sat unused for the last 15 years or
so. When powered up I get a 1 PPS signal out and eventually the GPS
Locked light comes on. I get some sort of data from the serial port
once a second but it is unreadable. I captured the output with my logic
analyzer but no matter what settings I try it can't decode the data
stream. It appears to be sending at 9600 baud as the shortest length
pulses in the data are about 104.8 uS long. I have tried getting
GPS.exe, Lady Heather and satstat.exe to talk to the unit. All of my
attempts to communicate with the unit have failed.
Of course I can't find a manual for it but I did find the HP 58503
manual which is supposedly similar. My understanding is that there is
no way to reset the serial port to the factory settings. Without the
serial port working I really have no way of knowing the status of the
unit. So, this is a long shot, does anyone have any ideas how I can get
this unit working or is it destined for the recycling bin?
Terry
time-nuts mailing list -- time-nuts@lists.febo.com
To unsubscribe send an email to time-nuts-leave@lists.febo.com
It used to work fine but it has sat unused for the last 15 years or so.
Do you recall what serial cable / adapter / PC you used back then? Some
hp GPSDO use RS422 instead of RS232. Sometimes you can hack differential
422 to make it look like 232 but you have to get the polarity right.
That's good that you know it's transmitting data at 9600 baud. Now look
at the bits to see if it's 8N1 or 7O1 or 7E1.
So two things to check: polarity and parity.
/tvb
Mine are both set at 19,200 odd 7-1 but I can't remember if that's the way
they are normally set up in cell site useage. And I have null modems on
both.
On Sun, May 25, 2025 at 7:17 PM Tom Van Baak via time-nuts <
time-nuts@lists.febo.com> wrote:
It used to work fine but it has sat unused for the last 15 years or so.
Do you recall what serial cable / adapter / PC you used back then? Some
hp GPSDO use RS422 instead of RS232. Sometimes you can hack differential
422 to make it look like 232 but you have to get the polarity right.
That's good that you know it's transmitting data at 9600 baud. Now look
at the bits to see if it's 8N1 or 7O1 or 7E1.
So two things to check: polarity and parity.
/tvb
time-nuts mailing list -- time-nuts@lists.febo.com
To unsubscribe send an email to time-nuts-leave@lists.febo.com
Thanks for all the help. Still stuck but some progress has been made, I think. I am pretty sure that my unit is stuck sending TOD every second. I think that the receiver is set to 9600 8,N,1 but I'm not positive. I've tried all the various setting for the serial port on GPSCon for 9600 Baud but nothing decodes what the receiver is sending out.
I was thinking that the TOD serial string would be plain text. That's certainly not what I'm seeing. The logic analyzer seems to decode it but only garbage prints out on the PC's com port.
This is an example of what the logic analyzer outputs for the TOD(?) data coming from the receiver.
\0 \x05\ \x02 \x08 \x03 \0 \t \x01 \x05 \x05 \t \x01 \x08 \xFA \0 ' @ \r
On 5/25/25 6:14 PM, Walt wrote:
About halfway down this page is a paragraph about getting the serial port out of full-duplex mode and into SCPI mode. It might be helpful, as well as trying the mentioned GPSCon software. I'm not aware that you can configure the serial port on the Z models - I always thought they were fixed.
I downloaded the Jackson version of GPSCon. Thank you for the link, that was very helpful. According to the info I have found for the Z8316, the serial port is user configurable. It shipped at 9600 8,N,1.
On 5/25/25 6:27 PM, JC via time-nuts wrote:
Mine are both set at 19,200 odd 7-1 but I can't remember if that's the
way they are normally set up in cell site useage. And I have null
modems on both.
I am using a null modem box that I built. It is set up to make it easy to connect my logic analyzer to either the RX or TX lines.
On 5/25/25 6:13 PM, Tom Van Baak via time-nuts wrote:
It used to work fine but it has sat unused for the last 15 years or so.
Do you recall what serial cable / adapter / PC you used back then?
Some hp GPSDO use RS422 instead of RS232. Sometimes you can hack
differential 422 to make it look like 232 but you have to get the
polarity right.
That's good that you know it's transmitting data at 9600 baud. Now
look at the bits to see if it's 8N1 or 7O1 or 7E1.
So two things to check: polarity and parity.
Unfortunately, I don't remember what cables etc I used back then.
I've messed around with this enough now I am begging to doubt
everything. 🙁
With GPSCon I have tried all the various settings for 9600 baud. I
didn't see a setting for polarity but I will go back and look again. I
do know that RS-232 is inverted from TTL serial. I seem to get valid
data on my logic analyzer when I invert the RS-232 data, I suspect that
the analyzer is expecting TTL logic levels. To decode the commands that
GPSCon sends to the receiver I must also set the analyzer to inverted.
Since they are both the same polarity I hope that that isn't the problem.
On 5/25/25 5:32 PM, JC via time-nuts wrote:
If the lock light is coming on, that's 90%, as it's perfectly useable
without
the communications. And if you're getting the pps that's getting closer to
100%. Are you sure your computer port is configured correctly? Maybe
it got changed for some reason, blip on the power, etc. I've run into the
same problem with the three I have and have always found that something
changed, either I changed the Z3801A communications config myself(for some
dumb reason or other) or the config in the computer got changed by me,
by windows(imagine that??) or by another serial program.
John K0GCJ
It's good to know I might still be able to us the 10MHz reference
signal, that's the part I really want to use. As noted above I've tried
all the various setting for 9600 Baud without luck so far. It's no
doubt something dumb on my part but I sure would like to figure it out.
Terry W7AMI
Terry,
Thanks for the additional information and the analyzer trace files.
- Your file Z3816-Hex.txt (attached) shows a nice burst of 18 bytes
every second. This is somewhat similar to other hp SmartClock GPSDO. The
0x0D stands out as a line terminator (CR) at the end of each "packet".
Other hp OEM products use 0D0A (CRLF), so again similar, not identical.
If you arrange the bytes in your trace as rows of 18 columns you get this:
00 05 02 08 04 01 01 02 02 00 06 01 08 FA 00 60 40 0D
00 05 02 08 04 01 01 02 02 00 07 01 08 FA 00 60 40 0D
00 05 02 08 04 01 01 02 02 00 08 01 08 FA 00 60 40 0D
00 05 02 08 04 01 01 02 02 00 09 01 08 FA 00 60 40 0D
00 05 02 08 04 01 01 02 02 01 00 01 08 FA 00 60 40 0D
This is delightful and thanks for collecting more than one second of
data. Notice that the first 5 bytes are always the same (0 5 2 8 4). The
next 6 bytes are 1 1 2 2 0 6 and it's counting to 1 1 2 2 1 0. So this
is likely the time (hhmmss). The remaining 7 bytes are always the same
and likely include status codes similar to other hp GPSDO. Note there
are no checksum byte(s), a checksum would change every second.
captured 5 seconds of data in hex format starting at about 17:22:05
Perfect. So your logic trace and your UTC clock matches the GPSDO
timecode within a second. That's good confirmation it's a valid
timecode, locked to UTC, a working GPSDO. Note the hour is off by 6
hours, suggesting the GPSDO is set to some local time offset not UTC.
- Your file TEK00002.gif (attached) captured one packet. You were
worried that it was only 18 ms long, because other hp GPSDO have 21 byte
(T1 format) and 25 byte (T2 format) timecodes. But we know yours is 18
bytes at 9600 baud, and 1/9600 * 10 * 18 = 18.75 ms, just like your
'scope trace shows.
So far the good news is your wiring, baud rate, polarity, and parity
appear to be correct. I don't think you could get 5 seconds of 18 bytes
like this if there was a problem with communication.
I think what you have is an OEM variant of the hp SmartClock series.
It's not outputting the standard T1 or T2 time code message format, but
something functionally similar. The bytes at the end may be status
codes. The bytes at the beginning could be a date or day number.
- You collected that data earlier today 5/27/2025. Now that it's a new
UTC day, can you look one more time and see if the first 5 bytes are
different? If the 0528 prefix becomes 0529, then it's likely month and day.
If so the next question is why did it say 0528 when you collected the
data on 05/27? One answer may be that the local time offset is in fact
not -6h but 24h - 6h = +18h. You can verify this theory by using the
TZONE SCPI command.
- You should be able to disable the timecode. That will allow you to
send SCPI commands. While in timecode mode the GPSDO will not respond to
any SCPI command except the one command that turns timecode off. So
send -- :PTIM:TCOD:CONT 0 -- and see if the timecode stops. If the old
favorites like *IDN? and SYST:STAT? work you're all set.
The general advice when exploring SCPI commands on a used GPSDO is to
issue only query commands at first. This may give you a clue to its
origin and configuration. Once you've done that you can try any SCPI
commands you like.
The standard T1 or T2 timecodes, local vs UTC, and TCOD trick are
documented in various SmartClock GPSDO manuals, such as:
http://www.leapsecond.com/museum/z3801a/097-z3801-01-iss-1.pdf
http://www.leapsecond.com/museum/hp58540a/097-58540-01-iss-1.pdf
After all this, if you run out of things to do, turn continuous timecode
back on and log the raw serial data using a terminal program for a day.
Do things like pull the GPS antenna for an hour, put it back on, etc.
This might help identify the status bytes.
/tvb
On 5/27/2025 7:29 AM, Terry wrote:
Thanks for all the help. Still stuck but some progress has been made, I think. I am pretty sure that my unit is stuck sending TOD every second. I think that the receiver is set to 9600 8,N,1 but I'm not positive. I've tried all the various setting for the serial port on GPSCon for 9600 Baud but nothing decodes what the receiver is sending out.
I was thinking that the TOD serial string would be plain text. That's certainly not what I'm seeing. The logic analyzer seems to decode it but only garbage prints out on the PC's com port.
This is an example of what the logic analyzer outputs for the TOD(?) data coming from the receiver.
\0 \x05\ \x02 \x08 \x03 \0 \t \x01 \x05 \x05 \t \x01 \x08 \xFA \0 ' @ \r
...
At least one version of the Z38xx has a second serial port that outputs time code only, and doesn't accept commands. Are you maybe looking at that port?
John
On May 28, 2025, 12:04 AM, at 12:04 AM, Terry via time-nuts time-nuts@lists.febo.com wrote:
Thanks for all the help. Still stuck but some progress has been made,
I think. I am pretty sure that my unit is stuck sending TOD every
second. I think that the receiver is set to 9600 8,N,1 but I'm not
positive. I've tried all the various setting for the serial port on
GPSCon for 9600 Baud but nothing decodes what the receiver is sending
out.
I was thinking that the TOD serial string would be plain text. That's
certainly not what I'm seeing. The logic analyzer seems to decode it
but only garbage prints out on the PC's com port.
This is an example of what the logic analyzer outputs for the TOD(?)
data coming from the receiver.
\0 \x05\ \x02 \x08 \x03 \0 \t \x01 \x05 \x05 \t \x01 \x08 \xFA \0 ' @
\r
On 5/25/25 6:14 PM, Walt wrote:
About halfway down this page is a paragraph about getting the serial
port out of full-duplex mode and into SCPI mode. It might be helpful,
as well as trying the mentioned GPSCon software. I'm not aware that
you can configure the serial port on the Z models - I always thought
they were fixed.
I downloaded the Jackson version of GPSCon. Thank you for the link,
that was very helpful. According to the info I have found for the
Z8316, the serial port is user configurable. It shipped at 9600 8,N,1.
On 5/25/25 6:27 PM, JC via time-nuts wrote:
Mine are both set at 19,200 odd 7-1 but I can't remember if that's
the
way they are normally set up in cell site useage. And I have null
modems on both.
I am using a null modem box that I built. It is set up to make it easy
to connect my logic analyzer to either the RX or TX lines.
On 5/25/25 6:13 PM, Tom Van Baak via time-nuts wrote:
It used to work fine but it has sat unused for the last 15 years or
so.
Do you recall what serial cable / adapter / PC you used back then?
Some hp GPSDO use RS422 instead of RS232. Sometimes you can hack
differential 422 to make it look like 232 but you have to get the
polarity right.
That's good that you know it's transmitting data at 9600 baud. Now
look at the bits to see if it's 8N1 or 7O1 or 7E1.
So two things to check: polarity and parity.
Unfortunately, I don't remember what cables etc I used back then.
I've messed around with this enough now I am begging to doubt
everything. 🙁
With GPSCon I have tried all the various settings for 9600 baud. I
didn't see a setting for polarity but I will go back and look again. I
do know that RS-232 is inverted from TTL serial. I seem to get valid
data on my logic analyzer when I invert the RS-232 data, I suspect that
the analyzer is expecting TTL logic levels. To decode the commands that
GPSCon sends to the receiver I must also set the analyzer to inverted.
Since they are both the same polarity I hope that that isn't the
problem.
On 5/25/25 5:32 PM, JC via time-nuts wrote:
If the lock light is coming on, that's 90%, as it's perfectly useable
without
the communications. And if you're getting the pps that's getting
closer to
100%. Are you sure your computer port is configured correctly? Maybe
it got changed for some reason, blip on the power, etc. I've run into
the
same problem with the three I have and have always found that
something
changed, either I changed the Z3801A communications config myself(for
some
dumb reason or other) or the config in the computer got changed by
me,
by windows(imagine that??) or by another serial program.
John K0GCJ
It's good to know I might still be able to us the 10MHz reference
signal, that's the part I really want to use. As noted above I've
tried
all the various setting for 9600 Baud without luck so far. It's no
doubt something dumb on my part but I sure would like to figure it out.
Terry W7AMI
time-nuts mailing list -- time-nuts@lists.febo.com
To unsubscribe send an email to time-nuts-leave@lists.febo.com
Terry,
I have both the Z3801 and the Z3805A. This is what I have recorded on the back of the units:
Z3801: 19200,7,1,1 (ODD)
Z3805A: 9600,N,8,1
My Z3801 has been converted to RS232 from 422. There is some info out there regarding the conversion.
I assume you have a gender changer as well. The gender of the Z38xx connector is inconsistent with the type of unit (DCE vs DTE). The Z38xx manual shows a DTE-to-DCE connection to a computer, and calls the computer DCE. Computers are typically DTE. Classic 56K modems are DCE, and use female sockets, and a STRAIGHT cable, but the Z3801/05 are DTE and female sockets. Confusing... Is it possible the gender is dictated by the RS-422?
IMHO, the Z38 should be DCE. The GPS satellite network is akin to the PTSN your 56K modem connected to. I've attached a pdf that I put together to get my head straight on these connections.
Hope this helps.
Bob
-----Original Message-----
From: JC via time-nuts time-nuts@lists.febo.com
Sent: Sunday, May 25, 2025 8:28 PM
To: Discussion of precise time and frequency measurement time-nuts@lists.febo.com
Cc: JC johnk0gcj@gmail.com
Subject: [time-nuts] Re: Z3716A Serial Port Issues
Importance: High
Mine are both set at 19,200 odd 7-1 but I can't remember if that's the way they are normally set up in cell site useage. And I have null modems on both.
On Sun, May 25, 2025 at 7:1 PM Tom Van Baak via time-nuts < time-nuts@lists.febo.com> wrote:
It used to work fine but it has sat unused for the last 15 years or so.
Do you recall what serial cable / adapter / PC you used back then?
Some hp GPSDO use RS422 instead of RS232. Sometimes you can hack
differential
422 to make it look like 232 but you have to get the polarity right.
That's good that you know it's transmitting data at 9600 baud. Now
look at the bits to see if it's 8N1 or 7O1 or 7E1.
So two things to check: polarity and parity.
/tvb
time-nuts mailing list -- time-nuts@lists.febo.com To unsubscribe send
an email to time-nuts-leave@lists.febo.com
time-nuts mailing list -- time-nuts@lists.febo.com To unsubscribe send an email to time-nuts-leave@lists.febo.com