crystal ladder filters
 

Just got the lot of 400+ crystals I won on ebay.
Turns out they are NOT HC49/u rocks, but HC6/U!
Isn't it ture that the larger crystals are better
for use in filters than the smaller one's (higher
Q, less loss, etc?)

Looks like some fun ahead experimenting with
this and maybe a homebrew spectrum analyzer in the
works to scan around 8.3866mhz (in my test oscillator
I'm getting 8.389855mhz, better let the oscillator and
the frequency counter warm up for a while).

Time to smash some surplus 9pin tube sockets up for
crystal socket pins ....

"Ken Scharf" wrote in message
. ..
Just got the lot of 400+ crystals I won on ebay.
Turns out they are NOT HC49/u rocks, but HC6/U!
Isn't it ture that the larger crystals are better
for use in filters than the smaller one's (higher
Q, less loss, etc?)

While Q comes with cubic feet (of space around inductors and capacitors) it
isn't the only factor when it comes to crystals. There, it's mostly surface
finish, cut and the amount of plateback used to place them on frequency. You
can have Q of several hundred K in the small HC-49, and less than 50K in the
larger holders. (Crystal itself may be larger or smaller, doesn't HAVE to go
with the can).

Have to measure a representative sample to determine the average, I haven't
seen a crystal made in recent years with really high Q. Quality control
folks have learned how to take all the quality out and make them cheaper.

After WW2, some manufacturers bought up all the FT-243 crystals they could
find. Chopped the wafers into tiny pieces and made really good crystals for
many years.

W4ZCB

"Ken Scharf" wrote in message
. ..
Just got the lot of 400+ crystals I won on ebay.
Turns out they are NOT HC49/u rocks, but HC6/U!
Isn't it ture that the larger crystals are better
for use in filters than the smaller one's (higher
Q, less loss, etc?)

While Q comes with cubic feet (of space around inductors and capacitors) it
isn't the only factor when it comes to crystals. There, it's mostly surface
finish, cut and the amount of plateback used to place them on frequency. You
can have Q of several hundred K in the small HC-49, and less than 50K in the
larger holders. (Crystal itself may be larger or smaller, doesn't HAVE to go
with the can).

Have to measure a representative sample to determine the average, I haven't
seen a crystal made in recent years with really high Q. Quality control
folks have learned how to take all the quality out and make them cheaper.

After WW2, some manufacturers bought up all the FT-243 crystals they could
find. Chopped the wafers into tiny pieces and made really good crystals for
many years.

W4ZCB

"Ken Scharf" wrote in message
. ..

Looks like some fun ahead experimenting with
this and maybe a homebrew spectrum analyzer in the
works to scan around 8.3866mhz (in my test oscillator
I'm getting 8.389855mhz, better let the oscillator and
the frequency counter warm up for a while).

Maybe they're designed for a circuit with parallel C, not series type..

You're probably intending to make a sweep generator and view the bandpass on
a scope, right? This site has a suitable sweep generator
http://www.intio.or.jp/jf10zl/

Time to smash some surplus 9pin tube sockets up for
crystal socket pins ....

But you're going to solder the xtals directly to boards for the actual
filters, right? You could make a board pretty easily, all the xtals having
the same pin separation. Lotta drilling.

Inspired by you I bought a bunch of HC-49s on eBay. They will plug into a
DIP socket, which is what I'm using in my measurement jig. Buffered
oscillator running off 9 V, no pulling when I hook up the counter.
http://tinyurl.com/3dpdn

73,

"PM"

"Ken Scharf" wrote in message
. ..

Looks like some fun ahead experimenting with
this and maybe a homebrew spectrum analyzer in the
works to scan around 8.3866mhz (in my test oscillator
I'm getting 8.389855mhz, better let the oscillator and
the frequency counter warm up for a while).

Maybe they're designed for a circuit with parallel C, not series type..

You're probably intending to make a sweep generator and view the bandpass on
a scope, right? This site has a suitable sweep generator
http://www.intio.or.jp/jf10zl/

Time to smash some surplus 9pin tube sockets up for
crystal socket pins ....

But you're going to solder the xtals directly to boards for the actual
filters, right? You could make a board pretty easily, all the xtals having
the same pin separation. Lotta drilling.

Inspired by you I bought a bunch of HC-49s on eBay. They will plug into a
DIP socket, which is what I'm using in my measurement jig. Buffered
oscillator running off 9 V, no pulling when I hook up the counter.
http://tinyurl.com/3dpdn

73,

"PM"

Ken Scharf ) writes:
Just got the lot of 400+ crystals I won on ebay.
Turns out they are NOT HC49/u rocks, but HC6/U!
Isn't it ture that the larger crystals are better
for use in filters than the smaller one's (higher
Q, less loss, etc?)

I have no idea. But remember other factors come into play.
Traditionally, only a handful of crystals were comodity items.
You could buy surplus, or you could get crystals for a handful
of frequencies that were premade, but generally you had to have them
specifically made for the application. I imagine (but don't know)
that such crystals tended to be better quality than a commodity crystal
from today. If nothing else, you were paying more for them.

And HC-6 holders tend to suggest something older, and hence maybe
dating from a day when more effort was put into making the crystals.
But that impression is not about the size of the holder, but about
the era of the crystal.

Of course, I've come across mass produced HC6 crystals so even then
it may be no indicator at all.

Michael VE2BVW

Looks like some fun ahead experimenting with
this and maybe a homebrew spectrum analyzer in the
works to scan around 8.3866mhz (in my test oscillator
I'm getting 8.389855mhz, better let the oscillator and
the frequency counter warm up for a while).

Time to smash some surplus 9pin tube sockets up for
crystal socket pins ....

Ken Scharf ) writes:
Just got the lot of 400+ crystals I won on ebay.
Turns out they are NOT HC49/u rocks, but HC6/U!
Isn't it ture that the larger crystals are better
for use in filters than the smaller one's (higher
Q, less loss, etc?)

I have no idea. But remember other factors come into play.
Traditionally, only a handful of crystals were comodity items.
You could buy surplus, or you could get crystals for a handful
of frequencies that were premade, but generally you had to have them
specifically made for the application. I imagine (but don't know)
that such crystals tended to be better quality than a commodity crystal
from today. If nothing else, you were paying more for them.

And HC-6 holders tend to suggest something older, and hence maybe
dating from a day when more effort was put into making the crystals.
But that impression is not about the size of the holder, but about
the era of the crystal.

Of course, I've come across mass produced HC6 crystals so even then
it may be no indicator at all.

Michael VE2BVW

Looks like some fun ahead experimenting with
this and maybe a homebrew spectrum analyzer in the
works to scan around 8.3866mhz (in my test oscillator
I'm getting 8.389855mhz, better let the oscillator and
the frequency counter warm up for a while).

Time to smash some surplus 9pin tube sockets up for
crystal socket pins ....

Paul_Morphy wrote:
"Ken Scharf" wrote in message
. ..


Looks like some fun ahead experimenting with
this and maybe a homebrew spectrum analyzer in the
works to scan around 8.3866mhz (in my test oscillator
I'm getting 8.389855mhz, better let the oscillator and
the frequency counter warm up for a while).


Maybe they're designed for a circuit with parallel C, not series type..

You're probably intending to make a sweep generator and view the bandpass on
a scope, right? This site has a suitable sweep generator
http://www.intio.or.jp/jf10zl/


Time to smash some surplus 9pin tube sockets up for
crystal socket pins ....


But you're going to solder the xtals directly to boards for the actual
filters, right? You could make a board pretty easily, all the xtals having
the same pin separation. Lotta drilling.

Inspired by you I bought a bunch of HC-49s on eBay. They will plug into a
DIP socket, which is what I'm using in my measurement jig. Buffered
oscillator running off 9 V, no pulling when I hook up the counter.
http://tinyurl.com/3dpdn

73,

"PM"


Yes I built a similar oscillator, a bit cruder though
I clipped my scope probe to the output and connected the
probe to the counter. I noticed that the frequency drifts somewhat
when the crystal is first plugged into the test oscillator. I
guess I need to let the frequency counter warm up for about a
half hour before taking any measurements, the oscillator may
also need a warmup period, and perhaps even the crystal.
If you handle the crystal while inserting it into the socket
the heat from your hands probably affects the crystal, though
I noticed that if I continue to hold the crystal with
two fingers things stablize more quickly (is my hand acting
as a crystal oven?) Total drift is under 250hz.

Paul_Morphy wrote:
"Ken Scharf" wrote in message
. ..


Looks like some fun ahead experimenting with
this and maybe a homebrew spectrum analyzer in the
works to scan around 8.3866mhz (in my test oscillator
I'm getting 8.389855mhz, better let the oscillator and
the frequency counter warm up for a while).


Maybe they're designed for a circuit with parallel C, not series type..

You're probably intending to make a sweep generator and view the bandpass on
a scope, right? This site has a suitable sweep generator
http://www.intio.or.jp/jf10zl/


Time to smash some surplus 9pin tube sockets up for
crystal socket pins ....


But you're going to solder the xtals directly to boards for the actual
filters, right? You could make a board pretty easily, all the xtals having
the same pin separation. Lotta drilling.

Inspired by you I bought a bunch of HC-49s on eBay. They will plug into a
DIP socket, which is what I'm using in my measurement jig. Buffered
oscillator running off 9 V, no pulling when I hook up the counter.
http://tinyurl.com/3dpdn

73,

"PM"


Yes I built a similar oscillator, a bit cruder though
I clipped my scope probe to the output and connected the
probe to the counter. I noticed that the frequency drifts somewhat
when the crystal is first plugged into the test oscillator. I
guess I need to let the frequency counter warm up for about a
half hour before taking any measurements, the oscillator may
also need a warmup period, and perhaps even the crystal.
If you handle the crystal while inserting it into the socket
the heat from your hands probably affects the crystal, though
I noticed that if I continue to hold the crystal with
two fingers things stablize more quickly (is my hand acting
as a crystal oven?) Total drift is under 250hz.

Yes I built a similar oscillator, a bit cruder though
I clipped my scope probe to the output and connected the
probe to the counter. I noticed that the frequency drifts somewhat
when the crystal is first plugged into the test oscillator.

Highly recommended reading;

"High-Performance crystal filter design" by Bill Carver, Communications
Quarterly, Winter 1993.

Addresses the same problem and lots more, and unless you plan on permanently
attaching your fingers to your filter, you'd better pay attention.

W4ZCB