Single transistor, 32768 crystal oscillator

[Bill Bowden]

Can anyone suggest improvements to the oscillator below
to move the frequency closer to 32768? It works the way it
is, but the frequency is too slow using the capacitors
shown and it loses about a second in 4 hours. I have
removed the 22pF cap (which isn't necessary for startuo),
but the frequency is still slow and the oscillator will
not start with 15pF in place of the 24pF. So it looks like
it's near the minimum capacity and yet the frequency is low.
The crystal is a regular watch, tuning fork variety, but I
only tried one. Maybe a JFET would work better?


+5 Volts
|
\
/ 68K
\
|
+ --------- +---------+ -----> Out
| | |
\ \ |
/ 300K / 3.3M |
\ \ C
| | |/
+--Crystal--+-----B| 2n3904
| | |\
--- --- E
--- 24pF --- 22pF |
| | |
| | |
GND GND GND

-Bill

 

[Robert C Monsen]

Can anyone suggest improvements to the oscillator below
to move the frequency closer to 32768? It works the way it
is, but the frequency is too slow using the capacitors
shown and it loses about a second in 4 hours. I have
removed the 22pF cap (which isn't necessary for startuo),
but the frequency is still slow and the oscillator will
not start with 15pF in place of the 24pF. So it looks like
it's near the minimum capacity and yet the frequency is low.
The crystal is a regular watch, tuning fork variety, but I
only tried one. Maybe a JFET would work better?


+5 Volts
|
\
/ 68K
\
|
+ --------- +---------+ -----> Out
| | |
\ \ |
/ 300K / 3.3M |
\ \ C
| | |/
+--Crystal--+-----B| 2n3904
| | |\
--- --- E
--- 24pF --- 22pF |
| | |
| | |
GND GND GND

-Bill

Have you tried the same crystal using an inverter rather than the
common-emitter amp?

The impedance looking into the transistor base is pretty low (beta * re).
Adding 1k or so might help.

My reading today indicates that the the load capacitance pulls the frequency
around, and that larger total capacitance lowers the frequency. Since the
22pF is in parallel with the input capacitance of the 2N3904, seems like
deleting it would decrease the total capacitance, and thus increase the
frequency a bit.

A JFET will usually have more input capacitance, due to the reverse biased
pn junction, thus causing the frequency to drop even further, I think.

If you have any variable capacitors, you could experiment with the
capacitance on the left. Also, a larger feedback resistor is called for in
the data sheet I saw (10MEG instead of 3.3MEG.)
http://www.ecsxtal.com/pdf/ecs-3x8.pdf

One final note. The crystal above has a max error tolerance of 20ppm, so if
its within spec, it should vary a max of 1 second every 14 hours. Thats like
a minute per month, so you aren't too far off.

Good luck. And, also, thanks for putting all those circuit schematics up on
the web. Lots of good stuff.

Regards,
Bob Monsen

 

[Bill Bowden]

Have you tried the same crystal using an inverter rather than the
common-emitter amp?

The impedance looking into the transistor base is pretty low (beta * re).
Adding 1k or so might help.

My reading today indicates that the the load capacitance pulls the frequency
around, and that larger total capacitance lowers the frequency. Since the
22pF is in parallel with the input capacitance of the 2N3904, seems like
deleting it would decrease the total capacitance, and thus increase the
frequency a bit.

A JFET will usually have more input capacitance, due to the reverse biased
pn junction, thus causing the frequency to drop even further, I think.

If you have any variable capacitors, you could experiment with the
capacitance on the left. Also, a larger feedback resistor is called for in
the data sheet I saw (10MEG instead of 3.3MEG.)
http://www.ecsxtal.com/pdf/ecs-3x8.pdf

One final note. The crystal above has a max error tolerance of 20ppm, so if
its within spec, it should vary a max of 1 second every 14 hours. Thats like
a minute per month, so you aren't too far off.

Good luck. And, also, thanks for putting all those circuit schematics up on
the web. Lots of good stuff.

Regards,
Bob Monsen

The crystal is slow using the either the transistor or the
4069 inverter circuit so it appears to be out of tollerance.
It came out of an old watch with a bad display (bleeding LCD),
but I don't know how old it is.

I tried a new crystal in the inverter circuit as shown in
the PDF file, but it's also slow. I removed the capacitor on
the inverter input side to try and speed it up but it's still
slow by about 1 second in 12 hours. I have 3 or 4 more crystals
to experiment with, maybe I will find one that can be set closer
to 32768. I would have guessed they were all cut for a slightly
higher frequency so that they could be set very close to the mark
with a little extra capacity.

The way I read the spec of +/-20ppm indicates the frequency can
be high or low with the recommended load capacitance of 12pF.
The example circuit in the PDF file uses two 22pF caps which are
in series yielding 11pF total load. But they don't say what the
frequency shift should be with more or less capacitance. I read
somewhere the reactance of the capacitor should equal the series
resistance of 330K which would be 15pF, but the oscillator will
not start with only 15pF. I haven't tried altering the 330K resistor
yet. Guess I'll try that next.

-Bill

 

[Watson A.Name - Watt Sun, Dark Remover]

The way I read the spec of +/-20ppm indicates the frequency can
be high or low with the recommended load capacitance of 12pF.
The example circuit in the PDF file uses two 22pF caps which are
in series yielding 11pF total load. But they don't say what the

Total capacitance is greater than 11 pufs because the chip itself and
wiring add some parasitic capacitance. So don't forget to count that
in the _true_ total capacitance.

frequency shift should be with more or less capacitance. I read
somewhere the reactance of the capacitor should equal the series
resistance of 330K which would be 15pF, but the oscillator will
not start with only 15pF. I haven't tried altering the 330K resistor
yet. Guess I'll try that next.

-Bill

Don't forget that watches were meant to be worn on a warm wrist, which
isn't the usual 25 degrees C.

--
@@F@r@o@m@@O@r@a@n@g@e@@C@o@u@n@t@y@,@@C@a@l@,@@w@h@e@r@e@@
###Got a Question about ELECTRONICS? Check HERE First:###
http://users.pandora.be/educypedia/electronics/databank.htm
My email address is whitelisted. *All* email sent to it
goes directly to the trash unless you add NOSPAM in the
Subject: line with other stuff. alondra101 <at> hotmail.com
Don't be ripped off by the big book dealers. Go to the URL
that will give you a choice and save you money(up to half).
http://www.everybookstore.com You'll be glad you did!
Just when you thought you had all this figured out, the gov't
changed it: http://physics.nist.gov/cuu/Units/binary.html
@@t@h@e@@a@f@f@l@u@e@n@t@@m@e@e@t@@t@h@e@@E@f@f@l@u@e@n@t@@

 

[Erik]

A standard 32.768 kHz crystal is 'on-frequency' with a 12,5 pF load.
They should work fine atleast from -10°C to +60°C.

-Erik-

 

[Bill Bowden]

Total capacitance is greater than 11 pufs because the chip itself and
wiring add some parasitic capacitance. So don't forget to count that
in the _true_ total capacitance.



Don't forget that watches were meant to be worn on a warm wrist, which
isn't the usual 25 degrees C.

--

Yes, the total capacitance is greater due to the stray
capacitance, but I removed one of the caps altogether and
the thing is still slow. Haven't tried the other crystals
yet, but I just thought of a simple fix. The oscillator
is feeding a PIC timer input, so I can probably correct the
error with software.

-Bill