[time-nuts] build your own crystals.

[Lux, Jim]

On 2/18/21 4:38 PM, Bob kb8tq wrote:
> Hi
>
> Ok
>
>> On Feb 18, 2021, at 6:40 PM, Magnus Danielson <magnus at rubidium.se> wrote:
>>
>> 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.
> A crystal plant isn’t going to let you in to do this. The risk of breaking gear /
> stopping production isn’t going to work for them.
>
> So what do you need to come up with for the group?:
>
> You need some quartz bars. Forget about trying to grow them yourself unless
> you have the ability to deal with battleship guns ….
>
> Next you will need an x-ray setup for the angles you wish to deal with. For AT
> cuts that may not be to bad. For SC’s … good luck at < $500K.
>
> For a summer camp sort of setup, go with a diamond saw, dice up the blank slowly,
> but with cheap gear.
>
> Now you need lapping equipment. Normally this is a device made from
> cast iron and with “couple of meters in each direction” sort of dimensions.
> Not much of an alternative out there so a fairly unique thing to find.
>
> Rounding the blanks can be done a couple of ways, a fairly normal lathe
> might be adapted to do this.
>
> For the contour process, I’d just go with drum processing. I’ll take you a
> couple of months, but there isn’t a lot of fancy gear involved.
>
> Chemical etch is easy, but you need to do it.
> (note that “polish” got left out …. )
>
> Now you need to put the base plate on the blank after cleaning it. It’s
> a thin film deposition process. If you are set up to make semiconductor
> wafers, you likely have the gear to do it. Same sort of deal, masks,
> mounting fixtures, thickness monitoring, gold evaporation at very
> high vacuum.
>
> Next up drop it in a solder seal base and solder it into an enclosure.
> (note that finish plate got left out).
>
> You now have a working device.
>
> Bob
>
If you want RF resonators to build in a "simple lab", it's probably 
easier to do SAW devices on a blank you buy. Then it's a single layer of 
aluminum that you photo etch.  Not hugely different than PC boards or a 
thick film hybrid.

Getting small feature sizes might be a challenge - when we did it where 
i used to work, they did photo reduction onto a photoresist covered 
substrate. I wonder if someone has a laser rig that could programmed to 
"draw" the pattern on the resist at the right scale.  As I recall, 50 
MHz has a wavelength of about 60 microns, and your transducer fingers 
are half wavelength apart.  You're probably not going to scribe those 
with an X-acto knife by hand. But it's not particularly exotic.

We bought the substrates already plated, but that is something you can 
conceivably do in a small lab - vacuum system and evaporation rig.

Then you have to bond wires onto the aluminum. But that's a "available 
from surplus" thing - and if you are equipped for doing hybrids you're 
all set.