[time-nuts] Bye-Bye Crystals

Tue Mar 14 22:17:44 UTC 2017

Hi,

Oh yes. I remember a certain design from a certain vendor. It was an 
crystal oscillator, sort of. Available in many diverse frequencies 
(hint). It was really a crystal oscillator and a PLL that could be 
programmed relatively inexpensively. Solves many problems, so it's a 
fine product, but it is a no-go for gigabit serial links, such as 
fibre-channel, gigabit ethernet and the like, as the PLL caused some 
serious systematics. This systematics along with the noise was scaled up 
as the step-up PLL just did a wide-band step-up (as it should). The 
combination causes excessive bit-errors rendering the link quite 
unusefull. With the tools at my hand, a sub-sampling scope with some 
cool histogram capabilities I used a resistive divider and a coax delay 
so that I could measure the trigger point and also the 1 cycle jitter, 
which is essentially what dominates for this kind of setup. I could see 
the systematic jitter clearly this way, and I could also get numbers for 
the random jitter. The vendors rep said "you can't measure that!" where 
I insisted I could. We ended up using other products for that purpose. I 
ended up measuring many oscillators to approve their jitter properties.

So well yes, you learn the hard way what those 4-leggers do when you 
have a bit of requirements. Later I dug up the patent for the process, 
which was focused more on the production of one standard product and 
late setting the frequency for customer needs. For it's purpose a great 
concept, not for all cases thought.

I know of MEMS issues. One MEMS I measured was tossed into a lock-loop, 
but the noise of it made the scope-view completely smeared out, despite 
being locked on average. I only showed the scope, pointiing out it was 
locked and that was the end of that discussion.

There is nothing wrong with these devices for many purposes, but expect 
that it may not solve everything, so measure and learn what 
characteristics is important for that particular application.

Cheers,
Magnus

On 03/14/2017 04:02 PM, Bob Camp wrote:
> Hi
>
> Some (but not all) of the resonant structures in the MEMS parts are effectively multi
> resonator / multi peak structures. Because of this the phase noise has multiple major
> bumps in it as you get into the region of all the peaks. Thats not going to give you
> great close in phase noise or ADEV. Since the manufacturers are often a bit unclear
> on “what’s inside” you need be a bit careful as you sort through the different parts out
> there. Even after sorting, you still run the risk of an “improved” design suddenly
> replacing the one you decided on.
>
> So much fun !!!
>
> Bob
>
>> On Mar 14, 2017, at 8:39 AM, Magnus Danielson <magnus at rubidium.dyndns.org> wrote:
>>
>> Hi,
>>
>> Some claims that MEMS will kill crystals. It will surely eat a good market share, but I think there is applications where MEMS is not mature enough compared to crystals.
>>
>> Another aspect is that various forms of synthesis technologies now exists, so that a high frequency CMOS oscillator is locked and divided down. Works sufficiently well for a whole bunch of applications.
>>
>> Again, your milage may vary and there is applications where you need the real deal or the right stuff.
>>
>> Cheers,
>> Magnus
>>
>>
>> On 03/14/2017 01:06 PM, Bob Camp wrote:
>>> Hi
>>>
>>>> On Mar 14, 2017, at 3:19 AM, Chris Albertson <albertson.chris at gmail.com> wrote:
>>>>
>>>> I think what he means is that the typical device sold today has four
>>>> terminals not two.  It looks like a crystal because it is inside a
>>>> little silver can but has four lead wires Power, ground and "output"
>>>> and the fourth lead might not be used.  It is an "XO" not an "X".
>>>>
>>>> But I argue that every one of these device has a crystal inside.  So
>>>> they still make crystals, just you don't see them
>>>
>>> These days, they may well have a MEMS resonator in them. No quartz and
>>> no crystal. Good luck on the close in noise if that’s what they are doing ….
>>>
>>> Bob
>>>
>>>>
>>>> On Mon, Mar 13, 2017 at 12:07 AM, Bryan _ <bpl521 at outlook.com> wrote:
>>>>> sorry, what do you mean by "complete oscillator" have outnumbered loose crystals?
>>>>>
>>>>>
>>>>> -=Bryan=-
>>>>>
>>>>>
>>>>> ________________________________
>>>>> From: time-nuts <time-nuts-bounces at febo.com> on behalf of Richard (Rick) Karlquist <richard at karlquist.com>
>>>>> Sent: March 12, 2017 4:38 PM
>>>>> To: Discussion of precise time and frequency measurement
>>>>> Subject: Re: [time-nuts] Bye-Bye Crystals
>>>>>
>>>>> I got a job in 1975 to design Konel's first synthesized radio, which
>>>>> was to obsolete their crystal controlled radios.  That's over 40 years
>>>>> ago.  The other trend (not mentioned) is that since 20 years ago or
>>>>> so, complete oscillator sales have vastly outnumbered sales of loose
>>>>> crystals.
>>>>>
>>>>> Rick N6RK
>>>>>
>>>>> On 3/11/2017 8:51 PM, Bob Camp wrote:
>>>>>> Hi
>>>>>>
>>>>>> International’s main business  was re-channeling non-synthesized radios and replacing
>>>>>> broken crystals in various pieces of com gear. It’s been a *lot* of years since the last of the
>>>>>> non-synthesized radios came out. The business probably has been dropping off pretty steadily
>>>>>> for many years …
>>>>>>
>>>>>> Bob
>>>>>>
>>>>>>> On Mar 11, 2017, at 10:39 PM, jimlux <jimlux at earthlink.net> wrote:
>>>>>>>
>>>>>>> On 3/11/17 4:30 PM, Scott McGrath wrote:
>>>>>>>> From the tone of the letter it sounds like the bank cancelled line of credit,
>>>>>>>
>>>>>>> Or, he wants to retire and nobody wants to carry it on.  His dad started it in 1950, the son picked it up in 1970.  It's 47 years later.
>>>>>>>
>>>>>>>>
>>>>>>>> Which is stupid given that much of their line is military which is getting a huge boost in spending
>>>>>>>
>>>>>>> Plenty of other crystal and oscillator manufacturers around.
>>>>>>>
>>>>>>> There's also a change in what kinds of crystals are needed.   I suspect most things being built and designed today use the crystal as a "master oscillator" that is used to drive some sort of synthesis chain. The need for "I have to have a 12.345,324 Hz crystal" is going away.
>>>>>>>
>>>>>>>
>>>>>>>>
>>>>>>>>> On Mar 11, 2017, at 4:56 PM, Bryan _ <bpl521 at outlook.com> wrote:
>>>>>>>>>
>>>>>>>>> Disappearing or manufacturing just moving overseas?. Video at the bottom is interesting, classic.
>>>>>>>>>
>>>>>>>>>
>>>>>>>>> http://hackaday.com/2017/03/11/so-long-and-thanks-for-all-the-crystals/
>>>>> [https://www.bing.com/th?id=OVF.LElrlkkbByR3K%2f6qfaeHjg&pid=Api]<http://hackaday.com/2017/03/11/so-long-and-thanks-for-all-the-crystals/>
>>>>>
>>>>> So Long, and Thanks for all the Crystals<http://hackaday.com/2017/03/11/so-long-and-thanks-for-all-the-crystals/>
>>>>> hackaday.com
>>>>> There was a time when anyone involved with radio transmitting -- ham operators, CB'ers, scanner enthusiasts, or remote control model fans -- had a collection of ...
>>>>>
>>>>>
>>>>>
>>>>>>>>>
>>>>>>>>>
>>>>>>>>>
>>>>>>>>> -=Bryan=-
>>>>>>>>> _______________________________________________
>>>>>>>>> time-nuts mailing list -- time-nuts at febo.com
>>>>>>>>> To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
>>>>> time-nuts Info Page - American Febo Enterprises<https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts>
>>>>> www.febo.com
>>>>> time-nuts is a low volume, high SNR list for the discussion of precise time and frequency measurement and related topics. To see the collection of prior postings to ...
>>>>>
>>>>>
>>>>>
>>>>>>>>> and follow the instructions there.
>>>>>>>> _______________________________________________
>>>>>>>> time-nuts mailing list -- time-nuts at febo.com
>>>>>>>> To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
>>>>> time-nuts Info Page - American Febo Enterprises<https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts>
>>>>> www.febo.com
>>>>> time-nuts is a low volume, high SNR list for the discussion of precise time and frequency measurement and related topics. To see the collection of prior postings to ...
>>>>>
>>>>>
>>>>>
>>>>>>>> and follow the instructions there.
>>>>>>>>
>>>>>>>
>>>>>>> _______________________________________________
>>>>>>> time-nuts mailing list -- time-nuts at febo.com
>>>>>>> To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
>>>>> time-nuts Info Page - American Febo Enterprises<https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts>
>>>>> www.febo.com
>>>>> time-nuts is a low volume, high SNR list for the discussion of precise time and frequency measurement and related topics. To see the collection of prior postings to ...
>>>>>
>>>>>
>>>>>
>>>>>>> and follow the instructions there.
>>>>>>
>>>>>> _______________________________________________
>>>>>> time-nuts mailing list -- time-nuts at febo.com
>>>>>> To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
>>>>> time-nuts Info Page - American Febo Enterprises<https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts>
>>>>> www.febo.com
>>>>> time-nuts is a low volume, high SNR list for the discussion of precise time and frequency measurement and related topics. To see the collection of prior postings to ...
>>>>>
>>>>>
>>>>>
>>>>>> and follow the instructions there.
>>>>>>
>>>>>>
>>>>> _______________________________________________
>>>>> time-nuts mailing list -- time-nuts at 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 at febo.com
>>>>> To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
>>>>> and follow the instructions there.
>>>>
>>>>
>>>>
>>>> --
>>>>
>>>> Chris Albertson
>>>> Redondo Beach, California
>>>> _______________________________________________
>>>> time-nuts mailing list -- time-nuts at 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 at 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 at 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 at febo.com
> To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
> and follow the instructions there.
>