[time-nuts] Ultra Stable Rb

Wed Apr 8 22:56:40 UTC 2020

On 4/8/2020 3:33 PM, Dana Whitlow wrote:
> Bob, what I' was getting at was: what do you do differently to make a
> stable Rb versus one
> that drifts a lot?  Never mind price issues.

My understanding from working on the HP10816 Rb standard
is that aging (as opposed to temperature drift) is due
to the vagaries of Rb atoms moving around and blocking
the light causing light shift.  The lamp also has
wearout mechanisms having to do with the Rb dissolving
into the glass and possibly making it opaque etc.  If
the cell is not "flooded", then how unflooded it is changes
as Rb gets absorbed by the glass.  The glass used is the
next best thing to Fused Quartz, in order to make it
as impervious as possible to Rb.

> 
> Could one buy, say, a PRS-10, extract the physics package from it, then
> engineer a stable
> Rb with that as a core?

No, the aging is a function of the physics package, especially
the glassware.  The electronics would have to be fairly poorly
designed to contribute to aging.

Note:  thermal drift (as opposed to aging) is a different
discussion altogether.  There are a gazillion temperature
effects, both in the physics and the electronics that need
to be addressed.  I think some designs simply use a TCXO
type of technique to address all of these at once.

Rick

> 
> Dana
> 
> 
> On Wed, Apr 8, 2020 at 4:19 PM Bob kb8tq <kb8tq at n1k.org> wrote:
> 
>> Hi
>>
>> The GPS Rb’s are “couple million dollar” sort of devices. Once GPS is up
>> and running,
>> the order volume drops off. The idea was to branch out into the broader
>> military market.
>> The design of the FRK did not change in any way as the price “morphed”.
>> They simply
>> had been making a pretty healthy margin on the product.  What happened to
>> the RFS
>> in the years after I left … no idea.
>>
>> Bob
>>
>>> On Apr 8, 2020, at 4:32 PM, djl <djl at montana.com> wrote:
>>>
>>> So it's Chinese engineering? Find something good that works, start
>> removing parts until it doesn't, put in the part last removed, and sell it?
>>>
>>> My granddad worked for a guy in LA called Madman Muntz who made tv's
>> that way in the late 1940's. They worked. Sorta. For a while. If the signal
>> was strong enough.
>>> Voila! $X-$400
>>>
>>> On 2020-04-08 13:51, Bob kb8tq wrote:
>>>> Hi
>>>> At the time EG&G had done the GPS Rb’s but not done any other
>>>> military parts. Some research showed that indeed the FRK went into
>>>> a variety of systems and the price was $X. (It varied a bit with
>> quantity)
>>>> Push the numbers around and look at this and that. The decision was made
>>>> that indeed if you could sell a few hundred to maybe a thousand a year
>> at
>>>> $X, it was a good thing. A design was done and (as noted earlier) it was
>>>> a good little device.
>>>> The fun part came with that $X pricing. Out comes a request for some few
>>>> hundred pieces to this or that organization. Bid $X, order goes down to
>> the
>>>> competition for X-$300. Next request for a few hundred, bid X-$400,
>> order
>>>> goes down to the other guy for X-$500. This step by step process goes on
>>>> for a year …. same result again and again.
>>>> At the end of that time period it was far less clear just *why* one
>> does up
>>>> an FRK like part ….
>>>> Bob
>>>>> On Apr 8, 2020, at 2:32 PM, djl <djl at montana.com> wrote:
>>>>> What a tease!!!!! OK, very well WHY???
>>>>> Don
>>>>> On 2020-04-08 08:04, Bob kb8tq wrote:
>>>>>> Hi
>>>>>> A few of the long running FRK’s ( in a very similar package …
>>>>>> hmmm ….. wonder why …. yes, I was there way back then
>>>>>> and know very well why :) ) have crazy good long term aging.
>>>>>> That said, I don’t think that I’ve seen a FRK quite this good.
>>>>>> Thanks for sharing !!!
>>>>>> Bob
>>>>>>> On Apr 8, 2020, at 9:45 AM, martyn at ptsyst.com wrote:
>>>>>>> Hi Guys,
>>>>>>> Just though you'd be interested in my prototype rubidium frequency
>> standard
>>>>>>> I made in the 1990's.
>>>>>>> http://www.ptsyst.com/RFS10-FrequencyDrift.pdf
>>>>>>> I have measured its frequency at random intervals for the past 18
>> years.
>>>>>>> Its never been adjusted and is just free running.
>>>>>>> It was turned off in 2005 and sent to a customer in Japan for a few
>> weeks,
>>>>>>> then returned and turned back on.
>>>>>>> For the past 18 years its stayed within plus/minus 3 x 10E-11.
>>>>>>> The overall linear drift is something like 1.85 x 10E-13 per month.
>>>>>>> This is not an advert.  There's no way any of our production units
>> are as
>>>>>>> good as this one, well I assume so as I've never measured any for 18
>> years
>>>>>>> continuous!
>>>>>>> Its now over 25 years old, have hardly ever been turned off.  Any
>> day I
>>>>>>> expect it to fail, but it keeps on running!!
>>>>>>> Regards
>>>>>>> Martyn
>>>>>>> _______________________________________________
>>>>>>> time-nuts mailing list -- time-nuts at lists.febo.com
>>>>>>> To unsubscribe, go to
>> http://lists.febo.com/mailman/listinfo/time-nuts_lists.febo.com
>>>>>>> and follow the instructions there.
>>>>>> _______________________________________________
>>>>>> time-nuts mailing list -- time-nuts at lists.febo.com
>>>>>> To unsubscribe, go to
>>>>>> http://lists.febo.com/mailman/listinfo/time-nuts_lists.febo.com
>>>>>> and follow the instructions there.
>>>>> --
>>>>> Dr. Don Latham  AJ7LL
>>>>> PO Box 404, Frenchtown, MT, 59834
>>>>> VOX: 406-626-4304
>>>>> _______________________________________________
>>>>> time-nuts mailing list -- time-nuts at lists.febo.com
>>>>> To unsubscribe, go to
>> http://lists.febo.com/mailman/listinfo/time-nuts_lists.febo.com
>>>>> and follow the instructions there.
>>>> _______________________________________________
>>>> time-nuts mailing list -- time-nuts at lists.febo.com
>>>> To unsubscribe, go to
>>>> http://lists.febo.com/mailman/listinfo/time-nuts_lists.febo.com
>>>> and follow the instructions there.
>>>
>>> --
>>> Dr. Don Latham  AJ7LL
>>> PO Box 404, Frenchtown, MT, 59834
>>> VOX: 406-626-4304
>>>
>>>
>>> _______________________________________________
>>> time-nuts mailing list -- time-nuts at lists.febo.com
>>> To unsubscribe, go to
>> http://lists.febo.com/mailman/listinfo/time-nuts_lists.febo.com
>>> and follow the instructions there.
>>
>>
>> _______________________________________________
>> time-nuts mailing list -- time-nuts at lists.febo.com
>> To unsubscribe, go to
>> http://lists.febo.com/mailman/listinfo/time-nuts_lists.febo.com
>> and follow the instructions there.
>>
> _______________________________________________
> time-nuts mailing list -- time-nuts at lists.febo.com
> To unsubscribe, go to http://lists.febo.com/mailman/listinfo/time-nuts_lists.febo.com
> and follow the instructions there.
> 