[time-nuts] AN/URQ13 reference AT cut crystal?

Thu Feb 18 23:40:31 UTC 2021

Hi,

A lot of fascinating steps. It would be real fun if one would do a
coarse in which one would actually build a handful of crystals oneself,
to learn the basics, and measure them up. It would be a fun
summer-coarse to do.

I have no expectance that it would be spectacular performance, just
being to build something which would result in an oscillating oscillator
would be quite cool.

Cheers,
Magnus

On 2021-02-18 22:07, Bob kb8tq wrote:
> Hi
>
>> On Feb 18, 2021, at 9:17 AM, Lux, Jim <jim at luxfamily.com> wrote:
>>
>> On 2/18/21 3:53 AM, Poul-Henning Kamp wrote:
>>> --------
>>> Bob kb8tq writes:
>>>
>>>> Turning an “idea” into a production capable part involves making many
>>>> batches of test samples. Think in the thousands of batches and hundreds
>>>> of parts in each batch. You have a “search” process at the blank chopping
>>>> level. You also have a search at the resonator fabrication level. Getting the
>>>> chopping part right is only a small part of the whole process….
>>> I realize this used to be a manual process, but today I would expect
>>> that you could automate a lot, of not most of it, if you wanted to ?
>>
>> That was my first thought, and then I thought through all the steps. I think it would be challenging to automate (and that brings up your question below about "is it worth it?")
>>
>> First, are you starting with natural or grown quartz? ( While grown quartz is used for most crystals, isn't there some performance benefit from natural crystals?)
> Finding natural mono crystalline quartz is a challenge. Finding big pieces of it is 
> more of a challenge. (sand is easier to find that big chunks). 
>
> Finding un-twinned crystals ( it comes in left and right hand orientations) is impossible. You
> are left working out a strategy on a per chunk basis to avoid twinning and inclusions.
> Once you are done, it’s slow growth quartz so it may be better in some respects.
>
>> I assume there is some sort of process to create "bars" of quartz from the raw boules or crystals.
> Quartz is grown from a seed. You dump raw quartz bits along with a chemical 
> soup into a pressure vessel. ( think in terms of a 16” naval gun barrel ….). You 
> seal it up, pressurize it and heat it up. At high enough temperatures the silicon
> dioxide feeder material dissolves into solution. You try *not* to dissolve your seed
> blanks at the same time :)
>
> The temperatures of the vessel ( = there are multiple zones) are varied over time.
> The pressure could also be varied as well. As you drop the temperature the crystals
> grow on the seed bars. The faster they grow, the more likely there are to be minor 
> defects in the resulting crystal. 
>
> The seed gets cut out of the bar before it really goes anywhere. 
>
>> I believe you'd need to do some analysis to determine the crystal axes and then the bars go into a series of saw steps - those might be automateable, in terms of sawing angles - is the sawing with a wire, with a diamond blade, or these days, perhaps abrasive water jet? 
> For volume production, a multi pass wire saw or a blade pack saw are pretty common. 
> For one off “lab” stuff a diamond saw might be used. In between the “grow it” and 
> “cut it” stages there is an x-ray process and some method of mounting the material 
> in a holder. 
>
>
>> That would give you a bunch of slabs with the cut with the right angles - I'm sure the machine they use today probably dumps them in a hopper, and I've watched enough food packaging shows to know that you could get those fairly large slabs onto a conveyor.
> For an OCXO crystal, the “slab” is going to be < 25mm square, generally much less. 
>
>> Then you'd have to cut your desired crystal shapes out of the slab (whether round, rectangular, or some other shape) - another sawing or grinding step, I assume.
> The most common shape for a precision part is a round blank. You stack the angle corrected
> squares together and put them on a centerless grinder. … oh, angle correction … Well you 
> can’t cut blanks accurately enough with a sawing operation, you have to correct the angle 
> after you have lapped them flat … oh, yes, you have to take the parts that come off of the
> saw over to a lapping machine and get the sides parallel, generally to a fraction of a light 
> band. 
>
>> And then mount in a holder automatically.
> Before you put it in a holder, you need to contour the blank. There might be another angle
> correct stage in there. Once you go through contour, you get to polish and then to chemical
> etch (to remove the debris). After that you baseplate the parts. Without that, you have no
> way to get an electrode on the “bottom” ( think TO-5 / TO-8 / HC-40) of the blank. 
>
> After baseplate the blank goes into the mount. If it’s a 4 point mount, that’s a manual 
> process. Next up is epoxy and baking to cure the epoxy. After that finish plate and seal. 
>
> For an OCXO part, you do as many of those steps (past baseplate) without breaking vacuum
> as you possibly can. You also do a bit of cleaning and bake before the final seal. 
>
>> All the individual steps are sort of "mass production" but I think today, there's significant (manual) setup time for the machine between steps (kind of like making tiny screws on a Swiss Screw Machine - some time for setup, then feed in bar stock and tiny screws or spacers come out of the machine)
>>
>> https://www.youtube.com/watch?v=mhwUHgWzzKs
>>
>>
>>
>>
>>> It would still be a lot of work, and very expensive, but like
>>> biochemist trying out hundred of thousand compounds from their
>>> "libraries", robots really lower the cost.
>> Lot of work, I think, underestimates the magnitude of the task. It would be interesting to compare the processes used for creating high performance crystals (e.g. for a USO, where they start 1000 blanks to get a dozen or so oscillators) and those used for mass production of crystals for things like kitchen timers and microcontrollers.
> The real contrast is to watch crystals. You photo fabricate multiple crystals from
> one blank. You also may not do much of a seal on them.  Indeed there are 
> crystals that are not destined for a watch that get processed that way as well. 
>
> Fun !!!!!
>
> Bob
>
>>   I have heard that for USOs, there's a couple people who have the "knack" for installing the crystal in the holder in a way that minimizes the stresses, etc.   That is *really hard* to automate.
>>
>> This is, of course, where SiLabs has a thing - they make hundreds (if not thousands) of MEMs oscillators at once with lithography, so the piece parts are very inexpensive - but they're performance limited by the material.
>>
>>
>>> The real question must therefore be, if anybody reasonably expects
>>> there to be any superior "new" cuts to find in the first place ?
>>>
>>> What properties would you program a quartz-crystal-prototyping robot to search for ?
>>>
>>> Which parameter(s) of current crystal-cuts are "their weak point" ?
>>>
>>
>> _______________________________________________
>> 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.