Hi

As long as we are cataloging all this stuff:

The run of the mill OCXO has a hermetic case on it. You get in and out of the 
case via pins with glass to metal seals. 
One of the common things to make the pins out of is Kovar. It’s great stuff, 
matches the TC of glass fine, easy to work 
with. It has one issue …. it’s not solderable. Kovar oxide *is* solderable. 
Unfortunately to form a proper oxide layer on the
stuff, you need to get it quite hot. 

So why does this matter in context? 

Well, if you happen to strip the oxide off of a pin (it’s not at all easy to 
do) or if  the pin was not properly oxidized, it’s  not 
going to solder. The solder will wet to it, the joint will look just fine. 
Exactly how it fakes you out is a bit of a mystery. Do a 
few thermo cycles, give the solder a tug and it pops off the pin. 

If you are rummaging around inside a flakey OCXO and the crystal is suspect …. 
give the leads a tug. They probably 
are Kovar. 

I do have empirical data on that last one, including SEM cross sections and all 
sorts of fun stuff ….

Bob


> On Sep 19, 2016, at 7:11 PM, Magnus Danielson <mag...@rubidium.dyndns.org> 
> wrote:
> 
> Dave,
> 
> While this is a bit of a tangent out of normal time-nuts, it still has some 
> relevance in terms handling of old equipment. Lets not get to crazy on this 
> tangent.
> 
> We had a similar issue many years ago. The BGA mounted chips would break away 
> from the boards and rattle around after transport. Quite annoying for new 
> gear.
> 
> The investigation took quite some time, but in the end, it turned out that 
> the gold layer had got too tick, and as you solder gold, it dissolves into 
> the solder, and if you now have too high ratio of gold, you get brittle 
> layers of gold-tin alloy which is much more fragile than a solder joint is. 
> As the PCB fab got their gold layer thickness in control, the problem went 
> away.
> 
> One method used in hand-soldering of gold-layered boards is to solder it, 
> then clean it right off and then solder again. This way most of the gold gets 
> removed and a clean contact achieved.
> 
> Gold does have benefits for the connector parts among other things.
> 
> Cheers,
> Magnus
> 
> On 09/20/2016 12:24 AM, Dave M wrote:
>> Adrian,
>> Your connector has probably experienced a problem common to gold plated
>> contacts and component leads.  Years ago, I worked in the avionics
>> repair and calibration facility at the Navy base in Jacksonville, FL.  I
>> saw this quite frequently in HP and Tektronix instruments that had been
>> in service for a number of years.
>> The issue is that the gold plating peels away from the base metal of the
>> contact or component lead, exposing the base metal to contaminants. Our
>> microminiature soldering lab began instructing all technicians to
>> completely remove the affected components and tin the leads with
>> solder.  Then, wick away the solder to completely rid the amalgam of the
>> gold, then resolder the component in place.
>> This was the only permanent solution to the problem; it might occur on a
>> different component, but not on the one that was properly repaired.
>> 
>> More often than not, the problem was exhibited as an intermittent
>> failure, consistant with board flexure or pressure.  Some of the
>> problems were very hard to find, but we had to keep investigating until
>> we had isolated the real point of failure.
>> 
>> Cheers,
>> Dave M
>> 
>> 
>> Adrian wrote:
>>>> Christopher,
>>>> 
>>>> my 4060 has been locking exactly 0.5 Hz off of the correct frequency.
>>>> The 1000B has had a contact problem on the edge connector.
>>>> Over the years, an insulating black layer had built up between the
>>>> solder and the gold plating of the contacts.
>>>> I had to re-solder each of them carefully twice to get it working
>>>> reliably.
>>>> 
>>>> Regards,
>>>> Adrian
>>>> 
>>>> Christopher Hoover schrieb:
>>>>> I've been debugging a FTS 4050 with the 5000M module.
>>>>> 
>>>>> It looks like the problem is the FTS 1200-100 OCXO.   It seems to
>>>>> have aged to the point that it is very low in frequency.   (Chuck
>>>>> Harris reported a similar situation a long time ago.)
>>>>> 
>>>>> I have it out of the 5000M and determined that I need to give it
>>>>> +8.8V on the control to get it within 50uHz of 5 MHz.
>>>>> 
>>>>> The power draw after a short warm up is < 2.5W (on this hot day);
>>>>> it looks like the oven is behaving.
>>>>> 
>>>>> Has anyone figured out how  to adjust the nominal frequency of an
>>>>> FTS 1200 -100 to recover the EFC range?
>>>>> 
>>>>> Also, anyone have a schematic for any model of the FTS 1200?
>>>>> 
>>>>> Thanks,
>>>>> -christopher
>>>>> 73 de AI6KG
>> 
>> _______________________________________________
>> time-nuts mailing list -- time-nuts@febo.com
>> To unsubscribe, go to
>> https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
>> and follow the instructions there.
> _______________________________________________
> time-nuts mailing list -- time-nuts@febo.com
> To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
> and follow the instructions there.

_______________________________________________
time-nuts mailing list -- time-nuts@febo.com
To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
and follow the instructions there.

Reply via email to