[time-nuts] Some additional RTFG learnings

Tue Feb 20 20:32:02 UTC 2007

I traced out where most of the interface connections went to.  I don't 
remember exactly what I found out, but I remember it was rather straight 
foward so I'm leaning towards it being  a simple cross over.  What I'm not 
sure about is the logic that selects the XO or the RB to be "ON".  But I'm 
not so sure that's important.

What is curious is that on the daughter board, the is another jumper W202 
right next to the U203 eeprom.  My guess is that U203 controls the GPS stuff 
and the U103 has the disciplining data for the particluar oscillator.

BTW you can get the RB unit to go "ON" with "NO GPS" turned off  you just 
take an Oncore TX and PPS to the interface...no XO unit needed.

----- Original Message ----- 
From: "John Ackermann N8UR" <jra at febo.com>
To: "Discussion of precise time and frequency measurement" 
<time-nuts at febo.com>
Sent: Tuesday, February 20, 2007 14:42
Subject: Re: [time-nuts] Some additional RTFG learnings

>I believe that there is more to the interface cable than a simple
> crossover -- I wonder if there are any pins that are jumped on one or
> the other of the connectors.
>
> By the way, in the experiment I reported on earlier, the XOs seem to be
> running undisciplined as well, and are about 1 Hz off nominal frequency.
>
> There's definitely something else needed to make these things work 
> properly.
>
> John
> ----
>
> TheInfamousFlavio at hotmail.com said the following on 02/20/2007 02:33 PM:
>> I got similar results in that the rb oscillators are not GPS disciplined
>> after a 72 hr run.
>>
>> I started off by manually offsetting the frequency of a warmed up rb unit 
>> by
>> doing a manual adjustment to the rb oscillator via the manual adjustment 
>> pot
>> so that it would be approximately 0.0050Hz off of my GPSDO.  Then I let 
>> it
>> run for 72hrs and no change to the oscillator.
>>
>> A few interesting things I discovered:
>>
>> 1. there is the unit is putting voltage to the electronic cfield 
>> connector
>> of the rb oscillator.  I discoved this when I tried adjusting the 
>> oscillator
>> when it was outside of the case.  When I put it back in the case an 
>> entirely
>> different frequency displayed then when I put it back in the case.  When 
>> I
>> manually adjusted the cfield pot on the rb oscillator, the unit did NOT
>> compinsate for the manual change...which lead me to believe that there is 
>> no
>> disciplining happening on the RB unit side. Although there might be
>> potential for it to happen because of the  voltage to the cfield pin ...
>> perhap some sort of initialization command is needed.
>>
>> 2. Both the XO and the RB unit can be in operational mode 
>> simulataneously.
>> On the Interface xover cable, I simply disconnected pins 1 and 5 (the
>> outside pins of the top row).  A simple +seems to control the A/B 
>> behavior
>> of the XO and RB units.  So both units display "NO GPS" off and "ON" lit.
>>
>> 3. There a little BCA shorting wires a few places on both RB and XO 
>> units.
>> The RB unit has BC jumped at each of these locations and the XO unit has 
>> AC
>> jumped.  W201 determines from which points the unit takes the signal from 
>> to
>> and from the oscillator.  W202 and W203 determine the to and from for the
>> GPS receivers TX and PPS signal at the interface.
>>
>> 4. You can xover the interface with only pins 2 and 6 on the XO to pins 4
>> and 9 on the RB. This simply send the GPS receiver's TX and PPS in the XO 
>> to
>> the RB unit. By doing this both units will have 'NO GPS' off and 'ON' on.
>>
>> 5. Grounding the CPURESET doesn't seem to make a difference.
>>
>> Does anyone have an idea of what PLDENB might mean?
>>
>>
>>
>>
>>
>> ----- Original Message ----- 
>> From: "John Ackermann N8UR" <jra at febo.com>
>> To: "Discussion of precise time and frequency measurement"
>> <time-nuts at febo.com>
>> Sent: Tuesday, February 20, 2007 10:00
>> Subject: Re: [time-nuts] Some additional RTFG learnings
>>
>>
>>> I started data gathering of XO vs. XO about three minutes after yanking
>>> the interconnects as described below (but after the units had been
>>> running in GPS locked mode for well over a day).
>>>
>>> There is definitely a stabilization period required after the XO becomes
>>> operational; on this measurement, the one second tau is around 4x10e-9!
>>> I'll try to get an idea how long it takes to stabilize and report back.
>>>
>>> John
>>> ----
>>>
>>> John Ackermann N8UR wrote:
>>>> Hi --
>>>>
>>>> I'm doing some stability comparisons of the RTFG-m-RB and RTFG-m-XO
>>>> units.
>>>>
>>>> I fired up two RTFG pairs (thanks to Jim Miller for lending me his
>>>> units) with the 10 MHz and cross-over interface cables in place.  The
>>>> units fired up normally.
>>>>
>>>> I did a 24 hour frequency stability run of the RB units measured 
>>>> against
>>>> each other, and all I learned in that time is that either (a) the RBs
>>>> are not actually GPS disciplined, or (b) the loop time constant is
>>>> longer than a 24 hour data collection will show.  After doing the XO
>>>> run, I will rerun the RBs for a longer period to see if there is any
>>>> sign of discipline.
>>>>
>>>> Then, I disconnected the RB units so I could activate the XOs to run 
>>>> the
>>>> same test.
>>>>
>>>> I thought the behaviour on doing that was worth noting:
>>>>
>>>> 1.  Disconnect the 10 MHz reference cable.  RB stays in "ON" mode, XO
>>>> goes to "FAULT."
>>>>
>>>> 2.  Disconnect crossover interface cable.  RB stays in "ON" mode, XO
>>>> stays in "FAULT" and "NO GPS" comes on.  But there is signal at the RF
>>>> OUT connector.
>>>>
>>>> 3.  In about 30 seconds, XO "FAULT" and "NO GPS" go off, "ON" comes on.
>>>>
>>>> So, the XO is definitely testing for the presence of the 10 MHz
>>>> reference input signal, and becomes unhappy when it goes away.
>>>>
>>>> John
>>>>
>>>> _______________________________________________
>>>> time-nuts mailing list
>>>> time-nuts at febo.com
>>>> https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
>>>>
>>>>
>>> _______________________________________________
>>> time-nuts mailing list
>>> time-nuts at febo.com
>>> https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
>>>
>>
>>
>> _______________________________________________
>> time-nuts mailing list
>> time-nuts at febo.com
>> https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
>>
>>
>
>
> _______________________________________________
> time-nuts mailing list
> time-nuts at febo.com
> https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
> 