-
#1000 Reply
Posted by
Bryan
on 02 Nov, 2018 21:37
-
Thanks all. I always thought the ribbon cables of that type were limited to about 25-50ma per track line. Even if they were all connected, would think it would be pushing it. I suspect the power input location is the same as the Symmetricom, at the diode/inductor/cap or whatever the component is. Will have to probe around later tonight.
But then again, I don't see any other power connector locations so I assume the mfg used the ribbon cable as well as a plug in play in the field.
EDIT:
Looks like a ZIF cable track can handle about 500ma, so should be no worries. All the power rails are tied together.
-
#1001 Reply
Posted by
BFX
on 03 Nov, 2018 15:25
-
Hi there,
Is here somewhere description about LEDs on Symmetricom GPSDO board? I didn't found that.
-
#1002 Reply
Posted by
texaspyro
on 04 Nov, 2018 00:29
-
Here is an example of why you want to look at GPSDO data over extended periods of time. A 1000 second run would miss that big honkin' burst of crapitude.
-
#1003 Reply
Posted by
BFX
on 09 Nov, 2018 18:49
-
Hi there,
Is here somewhere description about LEDs on Symmetricom GPSDO board? I didn't found that.
Nobody?
-
#1004 Reply
Posted by
Pipelie
on 12 Nov, 2018 06:49
-
Hello, I'm new in time-nuts, and just starting to play with this GPSDO, here is what I got.
ALM ACT status
RED GRN Indicative a fault was detected.
such as power supply out of tolerance
ORG GRN Antenna didn’t connect
OFF GRN Normal
OFF GRN flash(1Hz) Locking
OFF GRN flash(2Hz) GPS LOCK
-
#1005 Reply
Posted by
Astralix
on 22 Jan, 2019 12:26
-
Hi! First of all, let me thank you all for your great work on decoding these GPSDO modules. I read through the entire thread and copied out many sections that provide useful sequences for commands and output examples. And I apologize for the long post, but it recovers some bits and pieces from this rather long thread and may give you a hint of my way through what I tried and (mis-)understood.
I have bought myself one of these units, when this thread was very young. And I got a Trimble unit that was probably very old. It has firmware 1.0.0.2-01 installed and it misses some of the commands I collected from this thread. I connected a USB-V24 Adapter to the un-populated 4-pin header and have a 57600bps console showing the "UCCM >" prompt (not the "UCCM-P >").
-------------------------------------------------------------------------------
57964-05 serial number 85805368 firmware ver 1.0.0.2-01 W-CDMA mode
-------------------------------------------------------------------------------
Reference Status __________________________ Reference Outputs _______________
XX Ref 8KHz 0: [LOS]
XX Ref 8KHz 1: [LOS] TFOM 9 FFOM 1
XX Ref 8KHz 2: [LOS] UCCM A Status[ALARM]
XX Ref 8KHz 3: [LOS]
>> GPS: [phase:+1.3E+01, settling]
ACQUISITION ...................................................[GPS 1PPS Valid]
Tracking: 3 ____ Not Tracking: 3 ________ Time ____________________________
PRN El Az C/N PRN El Az GPS 10:37:13 22 Jan 2019
7 66 77 39 13 20 268
5 53 294 44 30 66 183 ANT DLY 50 ns
2 25 240 34 9 35 87 Position ________________________
MODE Hold
LAT N 50:xx:x.xxx
LON E 8:xx:xx.xxx
HGT +180.00 m (MSL)
ELEV MASK 20 deg
-------------------------------------------------------------------------------
The unit appears to work, but I always thought, the 10MHz output is a bit weak. Only my Racal Dana 1998 counter was willing to sync to that, but as it did sync, I never investigated that low output. Now with re-arranging my instruments on the bench, I stumbled over this low signal output again and connected it to my spectrum analyzer. Correctly terminated, the 10MHz output gives a clean sine wave around -36dBm level. And that explains why all my Marconi, Rohde & Schwarz and HP equipment did not recognize this as a 10MHz reference. Only the Racal Dana counter is happy with anything 10MHz as low as -39dBm. But that also explains, why a longer cable of probably lower quality now prohibits a proper sync...
Reading this absolutely informative thread, I encountered the MACT button and wow! Pressing it, the module throws +7dBm of signal and I can even use a simple splitter / combiner to feed the signal to 4 instruments! Unfortunately this works only for a while. After a time the MACT function resets itself automatically...
The board is continuously in ALARM and initially, before I was knowing what I was doing, it gave some hints that it was one half of a couple. It claimed about missing LINK and MASTER and such, so I guess it was a slave board. Now, after trying a lot of commands, it looks being configured relatively close to a stand-alone unit:
UCCM >stat
- UCCM Slot STATE-
1-1. #Now ACTIVE STATUS ---------------- [Alarm]
1-2. #Before ACTIVE STATUS ------------- [OCXO Warm]
2-1. #Reference Clock Operation -------- [Not Used]
2-2. #Current Reference Type ----------- [NONE]
2-3. #Current Select Reference --------- [X1PPS]
2-4. #Current Reference Status --------- [Good Accuracy & Stable]
#GPS STATUS ----------------------- [Available]
#Priority Level ------------------- [LINK > GPS]
#ALARM STATUS
#H/W FAIL [ LINK ]
#OPERATION ALARM ------------------ [LINK ]
3-1. #PLL STATUS ----------------------- [ENABLE]
3-2. #Current PLL MODE ----------------- [ENABLE]
But even the status is "Good Accuracy & Stable" the 10MHz output is muted and ALARM is lit.
While the GPS:INIT:... commands are supported and do work, some diagnostics and setup commands do not work or have a different syntax:
UCCM >DIAG:LOOP?
Undefined header
UCCM >DIAG:ROSC:EFC:DATA?
Undefined header
UCCM >:PTIM:LEAP:ACC?
Undefined header
UCCM >PTIM:LEAP:ACC?
Undefined header
UCCM >DIAG:GPS:UTC 0
Undefined header
UCCM >PTIM:TCOD?
Undefined header
UCCM >:PTIM:TCOD?
Undefined header
As I thought, the unit might have been configured to sync to some other unit, I tried the configuration commands:
UCCM >sync:ref:enable?
Command complete
UCCM >ref:type?
MASTER
Command complete
UCCM >
I could not find any hint for that MASTER here in this thread, but it would explain, why it is in ALARM even it states, that the GPS and OCXO are Good and Accurate...
But trying to re-configure it for running stand-alone runs into a dead-end for me:
UCCM >ref:type?
MASTER
Command complete
UCCM >REFerence:TYPE GPS
Parameter error
UCCM >REFerence:TYPE LINK
Undefined header
UCCM >REFerence:TYPE GPS,LINK
Parameter error
UCCM >REFerence:TYPE MASTER
Parameter error
UCCM >REFerence:TYPE master
Parameter error
UCCM >REF:TYPE X1PPS
Parameter error
UCCM >stat
- UCCM Slot STATE-
1-1. #Now ACTIVE STATUS ---------------- [Alarm]
1-2. #Before ACTIVE STATUS ------------- [OCXO Warm]
2-1. #Reference Clock Operation -------- [Not Used]
2-2. #Current Reference Type ----------- [NONE]
2-3. #Current Select Reference --------- [X1PPS]
2-4. #Current Reference Status --------- [Good Accuracy & Stable]
#GPS STATUS ----------------------- [Available]
#Priority Level ------------------- [LINK > GPS]
#ALARM STATUS
#H/W FAIL [ LINK ]
#OPERATION ALARM ------------------ [LINK ]
3-1. #PLL STATUS ----------------------- [ENABLE]
3-2. #Current PLL MODE ----------------- [ENABLE]
Command complete
UCCM >
On the other hand, there are a set of commands that do work, but do not change anything...
UCCM >OUTP:TP:SEL PP1S
Command complete
UCCM >SYNC:REF:DISABLE LINK
Command complete
UCCM >SYNC:REF:ENABLE GPS
Command complete
UCCM >stat
- UCCM Slot STATE-
1-1. #Now ACTIVE STATUS ---------------- [Alarm]
1-2. #Before ACTIVE STATUS ------------- [OCXO Warm]
2-1. #Reference Clock Operation -------- [Not Used]
2-2. #Current Reference Type ----------- [NONE]
2-3. #Current Select Reference --------- [X1PPS]
2-4. #Current Reference Status --------- [Good Accuracy & Stable]
#GPS STATUS ----------------------- [Available]
#Priority Level ------------------- [LINK > GPS]
#ALARM STATUS
#H/W FAIL [ LINK ]
#OPERATION ALARM ------------------ [LINK ]
3-1. #PLL STATUS ----------------------- [ENABLE]
3-2. #Current PLL MODE ----------------- [ENABLE]
Command complete
So and to follow, what I have tried you can see the units log:
L47:19/01/20.15:12:17 Ref available
L48:19/01/20.15:13:22 Ref Change: GPS
L49:19/01/20.15:13:23 Output mode: Alarm
L50:19/01/20.15:13:23 Ref chang: External PPS
[b]L51:19/01/21.21:03:46 Reset board by preset command[/b]
L52:99/08/22.00:00:01 Output mode: Alarm
L53:99/08/22.00:00:01 Power up start
L54:99/08/22.00:00:01 A/S change: Slave(HIGH)
L55:99/08/22.00:00:01 REF_SEL change: LINK(HIGH)
L56:19/01/21.21:05:24 Ref available
L57:19/01/21.21:07:07 Ref Change: GPS
L58:19/01/21.21:07:08 Ref chang: External PPS
L59:19/01/22.02:01:31 Ref not available
L60:19/01/22.02:03:56 Ref available
L61:19/01/22.11:18:43 Ref not available
L62:19/01/22.11:21:29 Ref available
The Reset line was probably an SYST:PON I tried yesterday. Issuing this command did btw not reset the initial position and time, that I set manually before! But I did not try a power-cycle to check if it is stored in any kind of nv-memory.
Does someone know more about that LINK / GPS / MASTER thing so I can get this unit out of ALARM? I ordered a second unit as there where some "cheap" available from dhgate.com and I already bought a NOS Rohde & Schwarz DSR Analyzer unit, that, as this radio service is out of service for years now, should serve as a case for my self-built lab GPSDO... But with constant ALARM set...
73!
-
#1006 Reply
Posted by
cdev
on 22 Jan, 2019 17:34
-
With the caveat that I don't have any experience at all with the UCCM -
"Ext PPS"
It may be looking for an external PPS. Its ability to be a 'slave' as well as a 'master' might I would suspect open up a big can of additional variables, I would operate under the assumption that it needs to be reset to defaults, and then set into master mode, however thats done, but not do anything until you had info on how to do that and were fairly sure of it. Such information if it exists is likely earlier in this large thread. If you don't find it you should search the Time-nuts mailing list for it, and if you cant find it ask there. You also might want to ask the question by itself in a thread, how to get out of slave mode if after investigating you still think thats whats happening.
I would suspect your unit is likely not operating as it should and is resettable, first but sometimes GPSDOs age beyond their factory set ability to adjust themselves. It might still be adjustable by hand, but I would not assume thats the case. In any case its far better to have a GPSDO to get working right than none. Its probably something very simple thats wrong.
-
#1007 Reply
Posted by
ZigmundRat
on 22 Jan, 2019 18:56
-
Currently at work not at home so I can't look at a unit directly, but I agree that the 'X1PPS' looks suspicious. IIRC, I'm thinking it should be 'GPS 1PPS'.
Also, the 'Current Reference Type' is NONE which is also not correct. It should be LINK or GPS.
I see you tried 'REFerence:TYPE GPS' which it rejected. I noticed that ALL of the REF commands entered were rejected. Seems odd. Maybe try the shorter 'REF:TYPE GPS' and see what happens.
(Also thinking that maybe your line ending characters could be causing an issue. What terminal program/emulator are you using? If it keeps rejecting most of these commands, just for giggles, try changing the line ending to CR only, or LF only. I forget what it's supposed to be. But I've seen cases where CR NULL for example will cause the NEXT command to fail. The first will work.)
One other thing: You can try a 'SYST:PRES' and it should go back to defaults. It SHOULD come up that way. It will show LINK in the status messages but it will lock to GPS and produce 10MHz output.
Looks like the -36dBm is what you get when the output is disabled. Hitting the MACT button forces the output on so you get the normal level then.
I will have to see if one of my units is a UCCM (non-P version). I'll check back later.
-
#1008 Reply
Posted by
texaspyro
on 22 Jan, 2019 19:04
-
-
#1009 Reply
Posted by
Astralix
on 22 Jan, 2019 19:59
-
Thank you guys, for coming to rescue me
I tried the SYST:PRES command and it changed the behavior a bit. The unit immediately went to a new survey and in that it changed the roughly preset values of time and position to the real coordinates. And instead of running 10% in 24h like earlier times, it already was at 24% when I had to leave my QRL.
Now in the evening it already had completed the survey but still is insisting on an ALARM and I have no real idea what to try next...
UCCM >syst:stat?
-------------------------------------------------------------------------------
57964-05 serial number 85805368 firmware ver 1.0.0.2-01 W-CDMA mode
-------------------------------------------------------------------------------
Reference Status __________________________ Reference Outputs _______________
XX Ref 8KHz 0: [LOS]
XX Ref 8KHz 1: [LOS] TFOM 9 FFOM 1
XX Ref 8KHz 2: [LOS] UCCM A Status[ALARM]
XX Ref 8KHz 3: [LOS]
>> GPS: [phase:+1.3E+01, settling]
ACQUISITION ...................................................[GPS 1PPS Valid]
Tracking: 4 ____ Not Tracking: 4 ________ Time ____________________________
PRN El Az C/N PRN El Az GPS 19:43:59 22 Jan 2019
14 20 250 34 26 12 287
25 63 95 35 2 30 52 ANT DLY 50 ns
29 82 205 34 12 28 98 Position ________________________
31 49 294 43 21 11 182 MODE Hold
LAT N 50:xx:x.xxx
LON E 8:xx:xx.xxx
HGT +239.45 m (MSL)
UCCM >stat
- UCCM Slot STATE-
1-1. #Now ACTIVE STATUS ---------------- [Alarm]
1-2. #Before ACTIVE STATUS ------------- [OCXO Warm]
2-1. #Reference Clock Operation -------- [Not Used]
2-2. #Current Reference Type ----------- [NONE]
2-3. #Current Select Reference --------- [X1PPS]
2-4. #Current Reference Status --------- [Good Accuracy & Stable]
#GPS STATUS ----------------------- [Available]
#Priority Level ------------------- [LINK > GPS]
#ALARM STATUS
#H/W FAIL [ LINK ]
#OPERATION ALARM ------------------ [LINK ]
3-1. #PLL STATUS ----------------------- [ENABLE]
3-2. #Current PLL MODE ----------------- [ENABLE]
I tried several command had no luck. I use linux and, as it does not alter chars you type, I prefer picocom for communication to the unit. But I can try to set cr-lf for some tests with line-endings. But actually I don't think that it is this issue...
I know lady heather and I have installed it. But besides printing nice data it doesn't get my module to enable the 10MHz output. But I try the latest version tonight or tomorrow morning. And I check the time-nuts forum. But would be great, if someone here might check again.
73!
-
#1010 Reply
Posted by
cdev
on 22 Jan, 2019 20:10
-
What kind of antenna are you using? Is there any chance that the amount of current its LNA is drawing is not registering as an antenna, i.e. is
too low? (or perhaps too high?)
If you have some means of trying another antenna do that, another idea to see if its amenable to a higher current, perhaps try a resistor to put the current drawn into an acceptable range. (say try 50 ma?) Its a long shot, but???
Thank you guys, for coming to rescue me
I tried the SYST:PRES command and it changed the behavior a bit. The unit immediately went to a new survey and in that it changed the roughly preset values of time and position to the real coordinates. And instead of running 10% in 24h like earlier times, it already was at 24% when I had to leave my QRL.
Now in the evening it already had completed the survey but still is insisting on an ALARM and I have no real idea what to try next...
UCCM >syst:stat?
...
-
#1011 Reply
Posted by
Astralix
on 22 Jan, 2019 22:24
-
Hi!
the GPS antenna is working, as you might have seen in my syst:stat? logs from above:
---------------------------------------------
Rcvr Status(1):
---------------------------------------------
Antenna Voltage: 5000.00mV , Antenna Current: 27.36mA
The GPS antenna is currently located indoors under the roof top. But as it has 4..6 sats constantly in sight and the survey is done in a few hours, it should work fine. There is also no alarm pending, that relates to this hardware. Reading through the thread I found several errors related to the OCXO, GPS, EFC or the logic devices. But none of these matches the description of what I see here.
The commands for hardware and operational error request report NONE.
So I am thinking around, what mode my board might be in... That some of the commands work and others throw a parameter error looks to me. like there is a basic mode of operation that such a board can work in. If it is MASTER it doesn't support LINK or GPS as the active reference, while in ?SLAVE? or whatever it might be called, the reference can be GPS or LINK. So is this the difference between the UCCM and UCCM-P unit? Is there a command to switch between these basic operational modes or is there a different firmware needed?
Again, there are no errors reported that match errors of other people in this thread and that clearly relate to dead hardware or ovens or such. What might help are commands, that report more information about the oven control voltage or mechanisms.
But reading "SETTLING" as the status of the PLL might be a hint, that this "MASTER" unit still tries to get the 10MHz right and therefore tries to get the 10MHz from a SLAVE unit, that is not connected...
For now I have started ladyheather and let it run over night. And I definitely will try to bring my GPS antenna outside to get a better reception...
I still need to figure out how to interpret the values presented from LH. The graph shows the DAC very low, but the DAC value is -1.96% what I would see as pretty in tolerance.
73!
-
#1012 Reply
Posted by
texaspyro
on 22 Jan, 2019 23:45
-
Some firmware might show an alarm if either the LINK or GPS is missing. The alarm may be a non-issue and is just a warning that the LINK data is missing.
These units show the DAC value as +/-100%. A value near 0 says the tuning voltage is well-centered.
-
#1013 Reply
Posted by
ZigmundRat
on 23 Jan, 2019 01:37
-
My units are all UCCM-P and Symmetricom. I have no Trimble units and all are UCCM-P. That said, I did some digging and found that your unit with the 'W-CDMA mode' in the SYST:STAT? output first line has been mentioned here on EEVBLOG before. Usually associated with a problem
User Jacobo mentioned problems with his, but never came back:
https://www.eevblog.com/forum/projects/a-look-at-my-symmetricom-gpsdo-(ocxo-furuno-receiver)/msg1061671/#msg1061671User lachesis also had problem with this same unit, and also never came back:
https://www.eevblog.com/forum/projects/a-look-at-my-symmetricom-gpsdo-(ocxo-furuno-receiver)/msg1204764/#msg1204764Another thread with indications that these units in W-CDMA mode can work properly, although the one in question was damaged, then repaired:
https://www.eevblog.com/forum/projects/trimble-65256-5v-10mhz-as-fitted-in-trimble-gpsdo-and-some-bg7tbl-output-level/But these messages are informative because they show similar messages from the UCCM. These units and yours, have the 'W-CDMA mode' tag. But they are basically the same firmware as my UCCM-P (1.0.0.2-01).
It's always possible that there is a hardware fault which is forcing the mode switch. If that is the case then all this software twiddling will not help. So please take a careful look at the board and see if there are any damaged or missing components. I suspect the "[phase:+1.3E+01, settling]' may be an indication of this. That is WAY off from anything normal. But with it trying to use external PPS, that could be correct. Who knows?
So let's start with some displays.
Please post the output from:
ALARm:HARDware?
ALARm:OPERation?
OUTPut:STATe?
OUTPut:MODE?
SYNC:REFERENCE:ENABLE?
REFerence:TYPE?
DIAGnostic:LOOP?
All of these are copy/paste from a console session and hopefully should work for you too. Then let's take a stab at getting it to switch to GPS. Try these commands:
SYNChronization:REFerence:DISABLE LINK
SYNChronization:REFerence:ENABLE GPS
REFerence:TYPE GPS
And then let's see the STATUS and SYST:STAT? output.
You may have a more 'rare' kind of unit with the W-CDMA mode, the 4x 8khz Link entries in the Ref area of the SYST:STAT? output, and the X1PPS reference. Absent a hardware issue, I'd still be surprised if we couldn't force it to GPS. But we also know there's at least four different versions of these, so there could also be customizations that interfere.
-
#1014 Reply
Posted by
cdev
on 23 Jan, 2019 02:05
-
Maybe take a bunch of high resolution well lighted photos of the board and then inspect them very closely on the computer?
A number of people got their boards in a fairly dinged up shape. Some crack in a component or missing component might lead to an alarm whose cause is otherwise non-obvious.
-
#1015 Reply
Posted by
Astralix
on 23 Jan, 2019 09:26
-
Currently at work I cannot take any photos of the unit, but it is already on my list to unmount it from its case and measure some details.
For the requested data:
UCCM >ALARm:HARDware?
NONE
Command complete
UCCM >ALARm:OPERation?
NONE
Command complete
UCCM >OUTPut:STATe?
ALARM
Command complete
UCCM >OUTPut:MODE?
W-CDMA
Command complete
UCCM >SYNC:REFERENCE:ENABLE?
GPS, LINK
Command complete
UCCM >REFerence:TYPE?
MASTER
Command complete
UCCM >DIAGnostic:LOOP?
Undefined header
DIAG:LOOP? is one of the commands not working.
For the commands:
UCCM >SYNChronization:REFerence:DISABLE LINK
Command complete
UCCM >SYNChronization:REFerence:ENABLE GPS
Command complete
UCCM >REFerence:TYPE GPS
Parameter error
Here I already tried all possible reference types that match and keywords in the text, the STAT or the SYST:STAT? and none worked.
I also found that working units with firmware 1.0.0.2-01 show LINK mode in the head-line of SYST:STAT? and not W-CDMA mode. As this correlates to the OUTP:MODE? I tried OUTP:MODE LINK but it also doesn't work and a command error sounds like this is a read-only parameter.
I also can confirm, that my W-CDMA unit behaves exactly the same as described in the first link, besides that after some SYST:REST and SYST:PON and such, the log has become much cleaner...
UCCM >diag:log:read:all?
Log status: 20 entries
L00:19/01/22.13:05:31 Ref not available
L01:19/01/22.15:01:22 Ref available
L02:19/01/22.15:05:02 Ref not available
L03:19/01/22.15:23:52 Ref available
L04:19/01/22.15:25:17 Ref not available
L05:19/01/22.16:03:14 Ref available
L06:19/01/22.16:09:40 Ref not available
L07:19/01/22.16:11:29 Ref available
L08:19/01/22.16:13:03 Ref not available
L09:19/01/22.17:28:20 Ref available
L10:19/01/22.17:34:34 Ref not available
L11:19/01/22.17:36:07 Ref available
L12:19/01/22.17:39:05 Ref not available
L13:19/01/22.17:47:38 Ref available
L14:19/01/22.17:49:54 Ref not available
L15:19/01/22.17:51:16 Ref available
L16:19/01/22.17:52:31 Ref not available
L17:19/01/22.17:57:02 Ref available
L18:19/01/23.01:57:38 Ref not available
L19:19/01/23.01:58:59 Ref available
Command complete
UCCM >stat
- UCCM Slot STATE-
1-1. #Now ACTIVE STATUS ---------------- [Alarm]
1-2. #Before ACTIVE STATUS ------------- [OCXO Warm]
2-1. #Reference Clock Operation -------- [Not Used]
2-2. #Current Reference Type ----------- [NONE]
2-3. #Current Select Reference --------- [X1PPS]
2-4. #Current Reference Status --------- [Good Accuracy & Stable]
#GPS STATUS ----------------------- [Available]
#Priority Level ------------------- [LINK > GPS]
#ALARM STATUS
#H/W FAIL [ LINK ]
#OPERATION ALARM ------------------ [LINK ]
3-1. #PLL STATUS ----------------------- [ENABLE]
3-2. #Current PLL MODE ----------------- [ENABLE]
From the photo of the second link I can see, that this unit is a more modern version as it seems to be a 57964-B PCB where mine is labeled 57964-A and for the hardware check. I really looked closely to the board and it had no dents or bruises. It also came well packed with lots of bubble-wrap. However I did not exactly watch out for missing parts and therefore did not inspect empty solder pads if they where never placed parts or ripped-off ones. So I do that inspection again.
Another thing is, that on the 50-pin FFC connector only pins are described, that could easily be found. So GND and Vcc are very easy, and where there is a TX, there should be a RX too
But reading my ALARMs there should be inputs that are not that easy to find, as they don't "talk". I wonder if we could bring these W-CDMA boards to life, if we short the 10MHz square wave or the 1PPS output to any of the possible inputs. So we fake the external reference?
Und if it expects an external 8kHz signal... There should be another 4 inputs that we never found before...
So I ordered a bunch of FFC cables and FFC adapter boards to be able to connect all pins. But delivery from China for small parts is precise in timing like shipping next day! (2 weeks +/- 18 weeks delivery time) I have enough equipment to carefully test the input pins that are still unknown.
Ah, and another thing I just found is, that while "?" does give a "Command error", "help" does give a "Parameter error" but I had no success in finding any working help parameter:
UCCM >?
Command error
UCCM >help
Parameter error
UCCM >help outp
Parameter error
UCCM >
Undefined header
UCCM >help outp:mode
Parameter error
UCCM >help
Parameter error
UCCM >help
Parameter error
UCCM >help?
Command error
UCCM >help reference
Parameter error
UCCM >help syst
Parameter error
UCCM >help syst:stat
Parameter error
UCCM >help syst:stat?
Parameter error
UCCM >
I had hopes that this could reveal the missing parameters for mode or link settings...
-
#1016 Reply
Posted by
metrologist
on 23 Jan, 2019 16:06
-
Some firmware might show an alarm if either the LINK or GPS is missing. The alarm may be a non-issue and is just a warning that the LINK data is missing.
These units show the DAC value as +/-100%. A value near 0 says the tuning voltage is well-centered.
That'd be good news for me. Really, they go negative dac %?
-
#1017 Reply
Posted by
texaspyro
on 23 Jan, 2019 16:36
-
Really, they go negative dac %?
A DAC range of +/-100% is fairly common on SCPI based GPSDOs. I don't know if the UCCM devices report negative values.
-
#1018 Reply
Posted by
cdev
on 24 Jan, 2019 00:46
-
Have you tried asking the manufacturer or looking on their web site for documentation of this model and its user interface and features?
-
#1019 Reply
Posted by
ZigmundRat
on 24 Jan, 2019 04:32
-
It's an OEM module built for Samsung CDMA/LTE/WiMAX systems by more than one supplier (Symmetricom and Trimble, maybe others). UCCM = "Universal Core Clock Module" - It's used in several products. The UCCM-P is a "Plus" version. Two different modules are in the attached pics. You can see the UCCM mounted board side up - look for the ribbon cable. Unfortunately knowing this and which modules doesn't take us very far even with user manuals, alarm manuals, etc. You can dig around and find out a fair amount of info but it's unlikely we'll ever find out anything about the software commands or the real details of operation. These are internal components of modules within a chassis in a frame in a larger system and not exposed to the user. I can find that the TL1 command "RTRV-GPS-STS;" returns the GPS status (locked, holdover, time of day, position, etc.) but I'll not find how it's talking to the UCCM module. And so we get to play detective
BTW in revisiting some of these docs, i found some interesting verbiage: "If GPS signals, which have been provided continually, are not provided due to a fault, the UCCM-P uses its previously learned ability to carry out the holdover function...". I wonder if that means it learns the aging rate and is adaptive when in holdover? But that's another topic.
So thanks for running those commands. It seems things are a bit closer, but it's still not switching to the right source. Let's keep poking at it for a bit. It also might be worth a PM to the guy that seemed to have it working (davebb I think it was) even though they don't seem to have been active lately. Post an update with some clear good res pics of the board too, maybe we'll spy something out of place.
-
#1020 Reply
Posted by
cdev
on 24 Jan, 2019 05:08
-
One guy in the Ukraine posted information about the connector pinout on his UCCM board, its earlier in the thread. There were multiple UARTs -
Its interesting to see how the GPSDO manages the GPS and training process/learning, etc.
-
#1021 Reply
Posted by
Astralix
on 24 Jan, 2019 10:33
-
Most test equipment of my hobby lab are former dead bricks, bought as parts or sold as "for spare parts or repair". Racal Dana, HP, Agilent, Marconi, Rohde & Schwarz. To repair or restore these units, I always contacted the original manufacturer, as long es it still exists, and in 80% I found someone, who gave me important hints or even schematics. For my radio test set I got overwhelming 500MB of pdf that include every single part, schematic, PCB prints, service steps and tests...
So I started the journey and tried contacting some brands I found on these GPS modules. Let's see what comes back.
@cdev, for the plug there is a problem. As I wrote before it is easy to find outputs that are active. 10MHz, 1PPS / 2PPS, TX lines of UARTs that periodically send data, that is no magic. And if you have a TX signal, the matching RX could not be too many pins away. But my W-CDMA module seems to rely on a second board that must either communicate with my board or send its clock to mine. And there are also 4 inputs of a magic 8kHz that my module seems to be sorry about:
57964-05 serial number 85805368 firmware ver 1.0.0.2-01 W-CDMA mode
-------------------------------------------------------------------------------
Reference Status __________________________ Reference Outputs _______________
XX Ref 8KHz 0: [LOS]
XX Ref 8KHz 1: [LOS] TFOM 9 FFOM 1
XX Ref 8KHz 2: [LOS] UCCM A Status[ALARM]
XX Ref 8KHz 3: [LOS]
I have no time to investigate that before Friday evening or the weekend. But I could imagine, that feeding back the 1PPS signal from my board into the right pin on the 50p connector could make a difference.
A thing that makes me bother is also, that my unit calls its own mode MASTER. I would assume that it should be the clock, providing the reference. However it doesn't seem to trust itself, so it is switching to the backup, the X1PPS of a secondary system. As the OCXO control pin is not smashed against a limit, the reference status is [Good Accuracy & Stable] and the GPS status is [Available] there is probably not much missing...
The only thing that makes me thinking is the fact, that out of 9 satellites visible often only a few, sometimes only one is tracked. Also I have a log, that reads "Ref not available" a few times a day, but within 2...3 minutes it comes back with "Ref available". So the GPS seems to be at it's lower limit.
Even all above is just "guess-work", it sounds logical to me, that the LINK boards work out of the box, while the W-CDMA boards do complain about a missing link. A LINK board always just sends it reference, but the MASTER board selects the source to go with. And so the MASTER will select a LINK if it's own reference is not trusted and it will constantly complain about a LINK missing. So hopefully my soon to arrive second board is a LINK, not a MASTER...
-
#1022 Reply
Posted by
texaspyro
on 25 Jan, 2019 04:48
-
-
#1023 Reply
Posted by
ZigmundRat
on 25 Jan, 2019 05:49
-
Thanks for the tip @texaspyro. That's certainly worth a try. According to the doc I found (below) pin39 goes to U10 pin 67.
The 'Master' state seems to be normal. Every status display I've seen from these shows this (at least when it's not in ALARM or otherwise not ready). I think this is to be expected since these units are supposed to provide the reference clock/timing for the system. I didn't see anything about redundant or 'slave' units in the Samsung cell docs.
I am sure that the 8khz inputs are not required for operation. 8khz *is* frequently found as a backplane clock in telco systems so its presence kind of makes sense. But we know that none of the other UCCM versions require it and happily come up fine on GPS alone.
Exercising my Google-Fu a bit more I came up with
another document (in Polish, by SP2IQW) which describes the fine pitch connector in yet more detail and provides some tracing of some of those signals. Also in this document we can see that this same model Trimble 57964-05 with the W-CDMA mode can operate off GPS just as well as the other UCCM modules we've encountered. The original forum thread is
at this site but you'll have to register to see the links.
When we look at the displays from your unit, we see that it's not getting the reference input required. This is why the TFOM and the 'phase' is so high. If that phase indication is consistent with other units, thats the difference between the internal 1PPS and the GPS 1PPS. With the GPS 1PPS being indicated as OK, we're back to the current reference being stuck on X1PPS.
My feeling is that there is 1) a hardware fault or 2) the configuration is corrupted and we don't know how to reset it.
Perhaps some other hidden commands can be discovered. Forcing holdover and recovery, restarting the survey and checking its progress, and other commands have been discovered by some educated guessing, but I suspect there's more to be found. Apparently my EPROM programmer will read the on board S29AL008D flash chip, given the right adapter. I may just have a go at it.
-
#1024 Reply
Posted by
Astralix
on 26 Jan, 2019 12:47
-
@texaspyro, @ZigmundRat
Thanks guys, that was a good hint, but it didn't solve all issues. Toggling this pin 39 does switch something, as the log shows some RefSwitch [LINK] and RefSwitch [GPS] but the unit later still switches to X1PPS, ignoring it's own reference and ALARM still active. As my unit is a revision A PCB, the LEDs are on the bottom. So it's hard to see what they do. But pulling pin 39 low seems to change the behavior of the green LED. It looks like it is off or dark green.
The ALARM LED is still constantly on.
As my unit still has the FFC connector installed and is mounted inside a small case, I did not try any other pins, to check if they change the things. I have ordered a bunch of adapter boards an foils from china and will wait until they arrive. Till then I installed the SV1AFN reference which is by far precise enough to repair the amateur radios waiting on my bench. And when it comes to the ICOM IC1271, the parts from china will have arrived.
I take a short break from here and will be back with the FCC and breakout board attached and hopefully a second Trimble unit of a more modern PCB or software. Till then, big thanks to all who supported me!
73!