Author Topic: Economical option for precision frequency reference?  (Read 117783 times)

0 Members and 1 Guest are viewing this topic.

Offline TSL

  • Regular Contributor
  • *
  • Posts: 240
  • Country: au
Re: Economical option for precision frequency reference?
« Reply #200 on: August 06, 2015, 10:57:59 pm »
TboltMon isn't the only program that can read TSIP, but it is mostly aimed at Thunderbolt owners. A better piece of software would be Trimble Studio and/or TSIP talker. You can get them from here...

ftp://ftp.trimble.com/pub/sct/embedded/bin/

regards

Tim
VK2XAX :: QF56if :: BMARC :: WIA :: AMSATVK
 

Offline motocoder

  • Supporter
  • ****
  • Posts: 769
  • Country: us
  • Electrical Engineer
Re: Economical option for precision frequency reference?
« Reply #201 on: August 07, 2015, 05:31:48 am »
I discovered that I can hook up Z38XX to either unit's diagnostic port, and see not just the OCXO data but also the GPS data. Given the GPS board is only in the REF1 unit, this MUST mean that there is some sort of communications link between REF0 and REF1 on the J5 interface cable.

I mention this because there was speculation on the time-nuts mailing list that the interface was very simple voltage-level signalling and the 1PPS signal only. That's clearly wrong, as there would be no way for REF0 to know what satellites REF1 was tracking were that the case.
 

Offline Shock

  • Super Contributor
  • ***
  • Posts: 3390
  • Country: au
Re: Economical option for precision frequency reference?
« Reply #202 on: August 07, 2015, 01:56:51 pm »
I discovered that I can hook up Z38XX to either unit's diagnostic port, and see not just the OCXO data but also the GPS data. Given the GPS board is only in the REF1 unit, this MUST mean that there is some sort of communications link between REF0 and REF1 on the J5 interface cable.

I mention this because there was speculation on the time-nuts mailing list that the interface was very simple voltage-level signalling and the 1PPS signal only. That's clearly wrong, as there would be no way for REF0 to know what satellites REF1 was tracking were that the case.

There is a lot happening on J5.

https://www.febo.com/pipermail/time-nuts/2014-November/087970.html
Soldering/Rework: Pace ADS200, Pace MBT350
Multimeters: Fluke 87V, 117, 27/FM       >>> WANTED STUFF <<<
Oszilloskopen: Lecroy 9314, Phillips PM3065, Tektronix 2215a, 314
 

Offline motocoder

  • Supporter
  • ****
  • Posts: 769
  • Country: us
  • Electrical Engineer
Re: Economical option for precision frequency reference?
« Reply #203 on: August 07, 2015, 02:24:28 pm »
I discovered that I can hook up Z38XX to either unit's diagnostic port, and see not just the OCXO data but also the GPS data. Given the GPS board is only in the REF1 unit, this MUST mean that there is some sort of communications link between REF0 and REF1 on the J5 interface cable.

I mention this because there was speculation on the time-nuts mailing list that the interface was very simple voltage-level signalling and the 1PPS signal only. That's clearly wrong, as there would be no way for REF0 to know what satellites REF1 was tracking were that the case.

There is a lot happening on J5.

https://www.febo.com/pipermail/time-nuts/2014-November/087970.html

Thanks. There's still a missing piece of the puzzle there, however. As it mentioned in that post, there must be some way that REF0 is able to send commands to the GPS receiver in REF1, possibly proxying through the micro on REF1.
 

Offline edpalmer42

  • Super Contributor
  • ***
  • Posts: 1820
  • Country: ca
Re: Economical option for precision frequency reference?
« Reply #204 on: August 07, 2015, 06:30:27 pm »
I discovered that I can hook up Z38XX to either unit's diagnostic port, and see not just the OCXO data but also the GPS data. Given the GPS board is only in the REF1 unit, this MUST mean that there is some sort of communications link between REF0 and REF1 on the J5 interface cable.

I mention this because there was speculation on the time-nuts mailing list that the interface was very simple voltage-level signalling and the 1PPS signal only. That's clearly wrong, as there would be no way for REF0 to know what satellites REF1 was tracking were that the case.

There is a lot happening on J5.

https://www.febo.com/pipermail/time-nuts/2014-November/087970.html

Thanks. There's still a missing piece of the puzzle there, however. As it mentioned in that post, there must be some way that REF0 is able to send commands to the GPS receiver in REF1, possibly proxying through the micro on REF1.

There's some current discussion on Time-Nuts on this topic.  Check out the archive:

https://www.febo.com/pipermail/time-nuts/2015-August/093303.html

This behaviour sounds identical to what I saw with another member of the Z38 family.  The Z3817A was made for Motorola.  It listens to a GPS and can report anything it hears, but it has no ability to control the GPS.  It disciplines its OCXO to the incoming PPS.  It looks like they copied this behaviour for the Lucent units.

The Z3817A happens to be the best GPSDO I've got.  When I measure the PPS output, the standard deviation of the period is < 0.1 ns with a max-min range of < 1 ns.

Ed
 

Offline motocoder

  • Supporter
  • ****
  • Posts: 769
  • Country: us
  • Electrical Engineer
Re: Economical option for precision frequency reference?
« Reply #205 on: August 07, 2015, 08:13:50 pm »
I discovered that I can hook up Z38XX to either unit's diagnostic port, and see not just the OCXO data but also the GPS data. Given the GPS board is only in the REF1 unit, this MUST mean that there is some sort of communications link between REF0 and REF1 on the J5 interface cable.

I mention this because there was speculation on the time-nuts mailing list that the interface was very simple voltage-level signalling and the 1PPS signal only. That's clearly wrong, as there would be no way for REF0 to know what satellites REF1 was tracking were that the case.

There is a lot happening on J5.

https://www.febo.com/pipermail/time-nuts/2014-November/087970.html

Thanks. There's still a missing piece of the puzzle there, however. As it mentioned in that post, there must be some way that REF0 is able to send commands to the GPS receiver in REF1, possibly proxying through the micro on REF1.

There's some current discussion on Time-Nuts on this topic.  Check out the archive:

https://www.febo.com/pipermail/time-nuts/2015-August/093303.html

This behaviour sounds identical to what I saw with another member of the Z38 family.  The Z3817A was made for Motorola.  It listens to a GPS and can report anything it hears, but it has no ability to control the GPS.  It disciplines its OCXO to the incoming PPS.  It looks like they copied this behaviour for the Lucent units.

The Z3817A happens to be the best GPSDO I've got.  When I measure the PPS output, the standard deviation of the period is < 0.1 ns with a max-min range of < 1 ns.

Ed

Now that I have two GPSDO - the Lucent and the Trimble - I've been doing some experiments to compare the two. I also have the (Morion) OCXO in my 53131A frequency counter.

If I hook the Lucent to channel one on my scope, and the Trimble to channel two, trigger off channel one and set infinite persistance, I can get some data on how much the relative phases of the two units drift. It has been steadily improving since I plugged the Trimble unit in. Right now there is as much as 17ns drift between the two, which I think is pretty good. I suspect the bulk of this is the Trimble, probably because it has only been plugged in for a few days. I say this because if I hook up the two GPSDO to my frequency counter, using its internal OCXO as the reference, I can see that the Trimble unit has quite a bit more drift than the Lucent. The Lucent drift is < 0.0003 Hz, whereas the Trimble is closer to 0.001 Hz. Obviously the OCXO in my counter is also not perfect, but I think as a relative measurement this is a valid technique for a rough comparison of the two GPSDO.

 

Offline TSL

  • Regular Contributor
  • *
  • Posts: 240
  • Country: au
Re: Economical option for precision frequency reference?
« Reply #206 on: August 07, 2015, 10:30:32 pm »

The Z3817A happens to be the best GPSDO I've got.  When I measure the PPS output, the standard deviation of the period is < 0.1 ns with a max-min range of < 1 ns.

Ed

OK... how were you measuring that ? What's your base reference ?

thanks

Tim
VK2XAX :: QF56if :: BMARC :: WIA :: AMSATVK
 

Offline edpalmer42

  • Super Contributor
  • ***
  • Posts: 1820
  • Country: ca
Re: Economical option for precision frequency reference?
« Reply #207 on: August 07, 2015, 10:43:32 pm »

The Z3817A happens to be the best GPSDO I've got.  When I measure the PPS output, the standard deviation of the period is < 0.1 ns with a max-min range of < 1 ns.

Ed

OK... how were you measuring that ? What's your base reference ?

thanks

Tim

I was using an HP 5370B (resolution of < 50 ps - yes picoseconds) referenced to my Efratom FRT Rb standard.  I made multiple runs of ~1000 readings per run.  Notice that I didn't say what the period was, only what the standard deviation was.  The 5370B reports that directly.  That means the absolute accuracy isn't relevant, just the resolution.  This test measures jitter.  The assumption with any GPSDO is that they've got the basics right and the frequency is correct.

Ed
 

Offline motocoder

  • Supporter
  • ****
  • Posts: 769
  • Country: us
  • Electrical Engineer
Re: Economical option for precision frequency reference?
« Reply #208 on: August 08, 2015, 12:15:29 am »
TboltMon isn't the only program that can read TSIP, but it is mostly aimed at Thunderbolt owners. A better piece of software would be Trimble Studio and/or TSIP talker. You can get them from here...

ftp://ftp.trimble.com/pub/sct/embedded/bin/

regards

Tim

Ok, TrimbleStudio does not appear to work with this device, despite it being labelled as "Trimble". The docs mention a protocol called UCCM. If I connect to the device with a serial terminal, at 57600/8/N/1, I get a UCCM-P prompt, and if I type help it lists the commands below. This appears similar, but not identical to the Lucent command set. For example, you do not enter a ":" character before running a command like "SYST:STAT?". Entering that command gives the following response (GPS coordinates removed):

UCCM-P >SYSTEM:STATUS?

-------------------------------------------------------------------------------
57964-80     serial number  60498671     firmware ver  2.0.1.6-01 LINK    mode
-------------------------------------------------------------------------------
Reference Status __________________________   Reference Outputs _______________
XX Ref 8KHz 0: [LOS]
XX Ref 8KHz 1: [LOS]                          TFOM     2            FFOM      0
XX Ref 8KHz 2: [LOS]                          UCCM A Status[ACTIVE]
XX Ref 8KHz 3: [LOS]
>> GPS: [phase:-8.6E-08]
ACQUISITION ...................................................[GPS 1PPS Valid]
Tracking: 5 ____   Not Tracking: 6 ________   Time ____________________________
PRN  El  Az  C/N   PRN  El  Az                GPS      00:13:33     08 Aug 2015
  2  14  73   35     5  45  62
 26  31 307   33    25  31 195                ANT DLY  10 ns
 29  85  88   47    16  10 327                Position ________________________
 20  61 162   43    18  10 205                MODE     Hold
 13  11 108   36    15   7 141
                    21  44 261                LAT      N  xx:xx:xx.xxx
                                              LON      W xxx:xx:x.xxx
                                              HGT              +xxx.xx m (MSL)



Commands
*IDN?
ALARm:HARDware?
ALARm:OPERation?
DIAGnostic:OUTPut ON|OFF
OUTPut:REFerence?
OUTPut:ACTive:ENABle
OUTPut:ACTive:DISable
OUTPut:ACTive:HOLDover:DURation:THReshold <seconds>
OUTPut:ACTive:HOLDover:DURation:THReshold?
OUTPut:INACTive
OUTPut:INACTive?
OUTPut:STATe?
SYNChronization:HOLDover:DURation:STATus:THReshold <seconds>
SYSTem:PRESet
SYNChronization:TFOMerit?
LED:GPSLock?
SYNChronization:FFOMerit?
GPS:POSition N or S,<deg>,<min>,<sec>,E or W,<deg>,<min>,<sec>,<height>
GPS:POSition?
GPS:POSition:HOLD:LAST?
GPS:REFerence:ADELay <numeric value>
GPS:REFerence:ADELay?
GPS:SATellite:TRACking:COUNt?
GPS:SATellite:TRACking?
DIAGnostic:ROSCillator:EFControl:RELative?
SYNChronization:TINTerval?
DIAGnostic:LOG:READ:ALL?
DIAGnostic:LOG:CLEar
SYSTem:PON
OUTPut:MODE?
SYSTem:STATus?
SYSTem:COMMunication:SERial1:BAUD 9600|19200|38400|57600
SYSTem:COMMunication:SERial1:BAUD?
SYSTem:COMMunication:SERial1:PRESet
SYSTem:COMMunication:SERial2:BAUD 9600|19200|38400|57600
SYSTem:COMMunication:SERial2:BAUD?
SYSTem:COMMunication:SERial2:PRESet
OUTPut:STANby:THReshold <seconds>
changeSN
SYNChronization:REFerence:ENABLE LINK|GPS
SYNChronization:REFerence:DISABLE LINK|GPS
SYNChronization:REFerence:ENABLE?
STATus
POSSTATus
TOD EN|DI
TIME:STRing?
REFerence:TYPE GPS|LINK
REFerence:TYPE?
PULLINRANGE 0|1|2|...|254|255
PULLINRNAGE?
DIAGnostic:LOOP?
DIAGnostic:ROSCillator:EFControl:DATA GPS|<value>
DIAGnostic:ROSCillator:EFControl:DATA?
OUTut:TP:SELection PP1S|PP2S
OUTut:TP:SELection?
GPSystem:SATellite:TRACking:EMANgle <degrees>
GPSystem:SATellite:TRACking:EMANgle?
DIAGnostic:TCODe:STATus:AMASk
DIAGnostic:TCODe:STATus:OMASk
DIAGnostic:TCODe:ERRor:AMASk
DIAGnostic:TCODe:ERRor:OMASk
DIAGnostic:HOLDover:DELay
DIAGnostic:HOLDover:DELay?
GPS:SATellite:TRACking:IGNore <PRN>, ...,<PRN>
GPS:SATellite:TRACking:IGNore?
GPS:SATellite:TRACking:INCLude <PRN>, ...,<PRN>
GPS:SATellite:TRACking:INCLude?
GPS:SATellite:TRACking:<select>:ALL
? or HELP

 

Offline TSL

  • Regular Contributor
  • *
  • Posts: 240
  • Country: au
Re: Economical option for precision frequency reference?
« Reply #209 on: August 08, 2015, 12:53:32 am »

I was using an HP 5370B (resolution of < 50 ps - yes picoseconds) referenced to my Efratom FRT Rb standard.


Very nice, wish I had one of those in my lab :)


VK2XAX :: QF56if :: BMARC :: WIA :: AMSATVK
 

Offline TSL

  • Regular Contributor
  • *
  • Posts: 240
  • Country: au
VK2XAX :: QF56if :: BMARC :: WIA :: AMSATVK
 

Offline edpalmer42

  • Super Contributor
  • ***
  • Posts: 1820
  • Country: ca
Re: Economical option for precision frequency reference?
« Reply #211 on: August 08, 2015, 01:40:59 am »

I was using an HP 5370B (resolution of < 50 ps - yes picoseconds) referenced to my Efratom FRT Rb standard.


Very nice, wish I had one of those in my lab :)

Actually, that's not my best time interval measurement tool.

https://www.eevblog.com/forum/testgear/wavecrest-dts-2077-teardown/msg279391/#msg279391

Ed
 

Offline motocoder

  • Supporter
  • ****
  • Posts: 769
  • Country: us
  • Electrical Engineer
Re: Economical option for precision frequency reference?
« Reply #212 on: August 08, 2015, 01:55:08 am »


Commands
*IDN?


This is SCPI the *IDN? is a dead giveaway!

http://www.ivifoundation.org/scpi/default.aspx

http://www.ivifoundation.org/docs/scpi-99.pdf

Nope, it's someone's half-ass copy of SCPI. If it was SCPI, you'd commands with either a ':' or a "*".

I think this thing can return all the same sort of useful information you can get with the other units, but it just seems to be some proprietary command set that isn't going to be compatible with any software out there. That's really ashame, as it otherwise seems like a decent unit. Many of the commands are the same, but without the ':" prefix on them, the software isn't going to work.
« Last Edit: August 08, 2015, 02:04:46 am by motocoder »
 

Offline TSL

  • Regular Contributor
  • *
  • Posts: 240
  • Country: au
Re: Economical option for precision frequency reference?
« Reply #213 on: August 08, 2015, 02:08:06 am »

I was using an HP 5370B (resolution of < 50 ps - yes picoseconds) referenced to my Efratom FRT Rb standard.


Very nice, wish I had one of those in my lab :)

Actually, that's not my best time interval measurement tool.

https://www.eevblog.com/forum/testgear/wavecrest-dts-2077-teardown/msg279391/#msg279391

Ed


Cricky - that's nice too :-+ and a quick search shows the 2079's selling for reasonable $ - ggrrr too many toys, not enough budget ! :)



VK2XAX :: QF56if :: BMARC :: WIA :: AMSATVK
 

Offline TSL

  • Regular Contributor
  • *
  • Posts: 240
  • Country: au
Re: Economical option for precision frequency reference?
« Reply #214 on: August 08, 2015, 02:12:15 am »

Nope, it's someone's half-ass copy of SCPI. If it was SCPI, you'd commands with either a ':' or a "*".

I think this thing can return all the same sort of useful information you can get with the other units, but it just seems to be some proprietary command set that isn't going to be compatible with any software out there. That's really ashame, as it otherwise seems like a decent unit. Many of the commands are the same, but without the ':" prefix on them, the software isn't going to work.

You're right! - I should have looked closer, didn't even notice the : was missing.

And that's bloody annoying -  :wtf: why only go 99% of the way ?
VK2XAX :: QF56if :: BMARC :: WIA :: AMSATVK
 

Offline Vgkid

  • Super Contributor
  • ***
  • Posts: 2600
  • Country: us
Re: Economical option for precision frequency reference?
« Reply #215 on: August 08, 2015, 03:16:20 am »

I was using an HP 5370B (resolution of < 50 ps - yes picoseconds) referenced to my Efratom FRT Rb standard.


Very nice, wish I had one of those in my lab :)
I assume that you figured out how to use it, can you show us what it does.

Actually, that's not my best time interval measurement tool.

https://www.eevblog.com/forum/testgear/wavecrest-dts-2077-teardown/msg279391/#msg279391

Ed
If you own any North Hills Electronics gear, message me. L&N Fan
 

Offline edpalmer42

  • Super Contributor
  • ***
  • Posts: 1820
  • Country: ca
Re: Economical option for precision frequency reference?
« Reply #216 on: August 08, 2015, 07:28:22 am »

I was using an HP 5370B (resolution of < 50 ps - yes picoseconds) referenced to my Efratom FRT Rb standard.


Very nice, wish I had one of those in my lab :)


Actually, that's not my best time interval measurement tool.

https://www.eevblog.com/forum/testgear/wavecrest-dts-2077-teardown/msg279391/#msg279391

Ed
I assume that you figured out how to use it, can you show us what it does.

It's a Time Interval Counter.  It measures frequency, time from A to B, rise time, fall time, period, pulse width, etc.  But it does it *really* well!  :)  I've been struggling to extract its full potential because it needs an input with very sharp rise and fall times.  Forget 10 MHz, even 100 MHz isn't really enough, 1 GHz is okay!  I built a circuit to square up a 5 or 10 MHz signal so that it can be properly measured, but it needs a little more work.

One of the particularly neat features is called a "strobing voltmeter".  It allows you to sample the input at time intervals as small as 10 ps and report the results.  In other words, it's digitizing the signal at 100 Gs/s.   :o  I threw together a proof-of-concept program to test this.  The attached picture is the output from my squaring circuit.  Due to limitations in the strobing voltmeter, it took me quite a while to figure out how to get this picture.  Initially I didn't think I would be able to do it.  The horizontal scale is in units of 100 ps.  The vertical scale is in volts.  The voltage is small because the maximum voltage on the input is only 1.7 volts.  I really don't want to blow the input on this thing!

Ed

 

Offline Vgkid

  • Super Contributor
  • ***
  • Posts: 2600
  • Country: us
Re: Economical option for precision frequency reference?
« Reply #217 on: August 08, 2015, 08:13:06 am »
So for the display, what can you get out if it(resolution/digits). Yes, i read the manual, i replied in your orig thread.
If you own any North Hills Electronics gear, message me. L&N Fan
 

Offline bingo600

  • Super Contributor
  • ***
  • Posts: 1508
  • Country: dk
Re: Economical option for precision frequency reference?
« Reply #218 on: August 08, 2015, 09:57:17 am »

One of the particularly neat features is called a "strobing voltmeter".  It allows you to sample the input at time intervals as small as 10 ps and report the results.  In other words, it's digitizing the signal at 100 Gs/s.   :o  I threw together a proof-of-concept program to test this.  The attached picture is the output from my squaring circuit.  Due to limitations in the strobing voltmeter, it took me quite a while to figure out how to get this picture.  Initially I didn't think I would be able to do it.  The horizontal scale is in units of 100 ps.  The vertical scale is in volts.  The voltage is small because the maximum voltage on the input is only 1.7 volts.  I really don't want to blow the input on this thing!

Ed

Hi Ed

I have a 2079 , any chance of a copy of your proof program ?
The strobo graph seems really usefull

/Bingo
 

Offline edpalmer42

  • Super Contributor
  • ***
  • Posts: 1820
  • Country: ca
Re: Economical option for precision frequency reference?
« Reply #219 on: August 08, 2015, 05:12:49 pm »
So for the display, what can you get out if it(resolution/digits). Yes, i read the manual, i replied in your orig thread.

I just measured a time interval from A to B.  The reported result was 753,191,898.407 ps as the average of 1000 results.  It also displays Std. Dev. (113.2 ps) and pk (+- 199.9 ps).  The measurement took a couple of seconds because the input frequency was only 500 Hz.

Ed
 

Offline edpalmer42

  • Super Contributor
  • ***
  • Posts: 1820
  • Country: ca
Re: Economical option for precision frequency reference?
« Reply #220 on: August 08, 2015, 05:35:47 pm »
Hi Ed

I have a 2079 , any chance of a copy of your proof program ?
The strobo graph seems really usefull

/Bingo

Sent via PM
 

Offline motocoder

  • Supporter
  • ****
  • Posts: 769
  • Country: us
  • Electrical Engineer
Re: Economical option for precision frequency reference?
« Reply #221 on: August 08, 2015, 06:10:45 pm »
Some more info on the Trimble/Wun-Hung-Lo hybrid. I powered it down to drill some ventilation holes in the case. Upon powering it back up, I see that the elevation mask setting has been reset.

So apparently they don't write settings to non-volatile RAM.  :-DD
 

Offline edpalmer42

  • Super Contributor
  • ***
  • Posts: 1820
  • Country: ca
Re: Economical option for precision frequency reference?
« Reply #222 on: August 08, 2015, 06:37:48 pm »
Some more info on the Trimble/Wun-Hung-Lo hybrid. I powered it down to drill some ventilation holes in the case. Upon powering it back up, I see that the elevation mask setting has been reset.

So apparently they don't write settings to non-volatile RAM.  :-DD

Maybe there is no non-volatile RAM in these units.  It appears that they came from a telco environment where they would be powered off of continuously float-charged lead-acid batteries.  They would never be powered down so, no need for non-volatile storage.  With a bit of reverse-engineering you should be able to add a battery backup to the GPS receiver itself.  It's designed to support that.

FYI, I have a similar Trimble GPSDO that I obtained from a source other than our favorite auction site.  It's command prompt is UCCM> rather than UCCM-P>.  The OCXO is a Trimble 65256 with a 2006 date code instead of a 73090 unit with 2009 or 2010 date code.  The performance seems similar to the RFTG.  In other words, not as good as the Trimble Thunderbolt or other members of the Z38 family.  I unsoldered the OCXO and found that, by itself, the performance (other than aging) is better than when it's in the circuit.

Ed
 

Offline motocoder

  • Supporter
  • ****
  • Posts: 769
  • Country: us
  • Electrical Engineer
Re: Economical option for precision frequency reference?
« Reply #223 on: August 08, 2015, 06:48:11 pm »
Some more info on the Trimble/Wun-Hung-Lo hybrid. I powered it down to drill some ventilation holes in the case. Upon powering it back up, I see that the elevation mask setting has been reset.

So apparently they don't write settings to non-volatile RAM.  :-DD

Maybe there is no non-volatile RAM in these units.  It appears that they came from a telco environment where they would be powered off of continuously float-charged lead-acid batteries.  They would never be powered down so, no need for non-volatile storage.  With a bit of reverse-engineering you should be able to add a battery backup to the GPS receiver itself.  It's designed to support that.

FYI, I have a similar Trimble GPSDO that I obtained from a source other than our favorite auction site.  It's command prompt is UCCM> rather than UCCM-P>.  The OCXO is a Trimble 65256 with a 2006 date code instead of a 73090 unit with 2009 or 2010 date code.  The performance seems similar to the RFTG.  In other words, not as good as the Trimble Thunderbolt or other members of the Z38 family.  I unsoldered the OCXO and found that, by itself, the performance (other than aging) is better than when it's in the circuit.

Ed

Well, it does provide a command to set the serial number, and if you set that, it does persist past a power cycle. So there must be some non-volatile storage there.

And even if their wasn't, that's shoddy design IMHO. Even if you expect it to be battery backed up indefinitely, if you ever have to do maintenance on the equipment, or if there is some AOG that removes power for an extended period of time, when the power is restored your technician will need to re-configure everything. It's hard to find a microcontroller these days without some built-in EEPROM, so I think most engineers would back up the settings in EEPROM.

But drawing some conclusions from the firmware and the non-functional software they provided with this unit, I think this is classic Wun-Hung-Lo engineering, so I am not surprised that they didn't bother to add the few extra lines of code to store the GPS settings in non-volatile RAM and restore them on power-up.
 

Offline motocoder

  • Supporter
  • ****
  • Posts: 769
  • Country: us
  • Electrical Engineer
Re: Economical option for precision frequency reference?
« Reply #224 on: August 08, 2015, 06:50:49 pm »
By the way, Ed - the Lucent seems to be almost an order of magnitude more stable than this thing so far. It's interesting to watch the Lucent OCXO age. The EFC voltage has been following a general upward trend since I powered it on. I guess at some point, this will level off, and at that point the OCXO will be considered broken in?
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf