[time-nuts] Troubleshooting an HP 58503A
medasaro at mit.edu
Sun Oct 14 18:03:29 EDT 2018
Thanks again for all the suggestions and advice I have gotten on
this project. It ran all night without loosing lock, so I am calling it
fixed. There is just one more order of business before I can use it, and
that is some way of "mounting" the GPS antenna on the roof in a
non-destructive manner, both to avoid damaging the (flat membrane) roof
and so that I can move it if necessary. See the attached photo of what I
have come up with. All the parts (minus the cable and antenna) came from
The concrete block is a generic 1 sqft paver stone which cost all
of $1.18. This is to provide a heavy base and avoid tipping. Attached to
that is an upside-down PVC toilet flange from the plumbing section. It
is attached with 6x stainless steel machine screws and wing-nuts that
pass all the way through the paver and are counter-sunk on the bottom so
that the heads won't damage the roof. Nylon spacers separate the toilet
flange and paver block by about 1/4 inch so that water can drain if
needed and to provide a space for the coax feed line to come out.
Pressed into the top of the paver is a 3" to 2" reducing adapter and
pressed into that is a 2" to 1-1/4" reducing adapter. The mast itself is
just a 2-foot section of 1-1/4" PVC pipe. The 58532A antenna is
apparently designed to mount over a 1-1/4" mast since its inner diameter
is 43mm and the outside of a 1-1/4" pipe is 42mm.
It is sitting on the back railing of the house for now but I
ordered a 40' length of LMR-250 cable with N-connectors on both ends
from http://www.usacoax.com/ and once that comes I can set it up for
real on the roof with the cable running into my lab. Over and out.
On 10/13/2018 09:21 PM, Matthew D'Asaro wrote:
> It just locked!
> On 10/13/2018 09:20 PM, Matthew D'Asaro wrote:
>> All -
>> I think I have it fixed! The problem was a broken wire inside my
>> genuine HP 58532A antenna. I was suspicious of the antenna because it
>> didn't seem to matter if I had it connected or not - the 58503A
>> behaved exactly the same. I confirmed my suspicions by measuring the
>> DC current it drew. Zero. Something was wrong so I took it apart and
>> measured the resistance between the center conductor and the board
>> and found it to be infinity. As it turns out the center conductor of
>> the micro-coax that connects the N-connector to the board had broken
>> right where it was soldered on. It took a bit to figure out how to
>> press the center conductor out of the connector, reattach the coax,
>> and press it back in. To press it out I first had to unsolder the
>> shield of the coax and then I could use the back of a drill-bit
>> placed in a vice to force the center conductor out of the insulator.
>> Once I had the cable reattached, I could press it back in by placing
>> it over the center contact of a male N-connector and then screwing
>> the two together. See the attached pictures.
>> Anyway, once I had the antenna repaired it found satellites to
>> track immediately - I didn't even have to set an initial date or
>> location. It is about 2% of the way through its survey now and
>> hopefully will have locked in by morning.
>> Thanks to all of you for your help!
>> -Matthew D'Asaro
>> On 10/13/2018 05:39 AM, Azelio Boriani wrote:
>>> Try a SYSTem:PREset first, before sending the GPS:INITial.
>>> On Sat, Oct 13, 2018 at 8:44 AM Matthew D'Asaro <medasaro at mit.edu>
>>>> All -
>>>> I have made substantial progress on my HP 58503A, but I am not
>>>> quite there yet. By watching the internal lights and the data on the
>>>> processor bus I could see that the CPU was starting to boot and then
>>>> restarting every ~15 seconds. Furthermore, the reset circuit didn't
>>>> to be working right. Sometimes when I power cycled it the front panel
>>>> lights would flash to indicate a self-test, other times they would
>>>> This seemed to indicate that the system was not being reset
>>>> correctly on
>>>> start-up. More telling, when I manually triggered the reset generator
>>>> chip, the system would not ever flash its front panel lights even
>>>> I could see that the CPU was being reset per the internal lights.
>>>> The problem, as it turned out, was more mechanical than
>>>> As shown in the attached photo, someone had put the case back together
>>>> with a screw that was too long! It dug into the board and cut two
>>>> traces, one of which was the reset line for the UART chip (and
>>>> some other chips as well - I didn't fully trace the line). Once I
>>>> repaired these two traces (red wires in my picture) it booted right up
>>>> and even passed its self-test!
>>>> However, the story does not end there. I now have it booting and
>>>> appearing to "work" but I can't get it to track satellites. I manually
>>>> set the location, date, and time with the following commands:
>>>> GPS:INIT:DATE (date)
>>>> GPS:INIT:POS (position)
>>>> GPS:INIT:TIME (time)
>>>> I then setup the unit (inside, but with the GPS antenna near a
>>>> window) and left it turned on overnight. However, no dice. The system
>>>> does not display errors but won't track any satellites. Here is the
>>>> results of a SYSTEM:STATUS? command:
>>>> scpi > system:status?
>>>> ------------------------------- Receiver Status
>>>> SYNCHRONIZATION ........................................... [ Outputs
>>>> Invalid ]
>>>> SmartClock Mode ___________________________ Reference Outputs
>>>> Locked TFOM 9
>>>> FFOM 3
>>>> Recovery 1PPS TI --
>>>> Holdover HOLD THR 1.000 us
>>>> >> Power-up: GPS acquisition Holdover Uncertainty ____________
>>>> Predict --
>>>> ACQUISITION .............................................. [ GPS 1PPS
>>>> Invalid ]
>>>> Tracking: 0 ____ Not Tracking: 14 _______ Time
>>>> PRN El Az PRN El Az UTC 05:08:22 [?] 11
>>>> Oct 2018
>>>> * 1 -- --- *10 -- --- GPS 1PPS Invalid: not
>>>> * 2 -- --- 16 35 50 ANT DLY 0 ns
>>>> * 3 -- --- *23 55 129 Position
>>>> * 5 -- --- 26 10 41 MODE Survey:
>>>> 0% complete
>>>> * 6 -- --- 27 21 96 Suspended: track <4 sats
>>>> 7 57 267 30 25 257 INIT LAT N
>>>> * 8 13 127 INIT LON W
>>>> 9 81 141 INIT HGT
>>>> m (MSL)
>>>> ELEV MASK 10 deg *attempting to track
>>>> HEALTH MONITOR
>>>> [ OK ]
>>>> Self Test: OK Int Pwr: OK Oven Pwr: OK OCXO: OK EFC: OK GPS
>>>> Rcv: OK
>>>> scpi >
>>>> Is there anything else I can try?
>>>> Thanks in advance,
>>>> Matthew D'Asaro
>>>> On 10/07/2018 12:26 AM, Mark Sims wrote:
>>>>> The 58503A needs a null-modem pinout serial cable to connect to a
>>>>> PC. The baud rate defaults to 9600:8:N:1
>>>>> time-nuts mailing list -- time-nuts at lists.febo.com
>>>>> To unsubscribe, go to
>>>>> and follow the instructions there.
>>>> time-nuts mailing list -- time-nuts at lists.febo.com
>>>> To unsubscribe, go to
>>>> and follow the instructions there.
>>> time-nuts mailing list -- time-nuts at lists.febo.com
>>> To unsubscribe, go to
>>> and follow the instructions there.
-------------- next part --------------
A non-text attachment was scrubbed...
Name: GPS Antenna Mount.JPG
Size: 358738 bytes
Desc: not available
More information about the time-nuts