How to fine tune frequency from 4060 IC and crystall

My goal is to generate 40.00 Hz signal (NB, not MHz, not KHz).

The 4060 IC with a 4MHz crystal works nicely and gives 4E6 / 2^14 = 244.14 Hz from 4060 pin 3. Then, dividing by 6 using 7490 IC should give 40.69. The breadboard setup now works very well and delivers a consistent approximately 40.6 Hz signal, evidenced by the sound heard through a speaker close to the ~low Eb (40 Hz) heard at:
https://www.szynalski.com/tone-generator/. (The exact musical Eb is supposed to be closer to 38.891).

Using a variable ("trim" ?) capacitor "at the crystal stage" supposedly can make the adjustment from 244.14 to 240, to be closer to exactly 240 which would get (dividing by 6), 40.00 Hz. Using a trim capacitor on pin 9 of the 4060 (in addition to the crystal, resistor and caps on pins 10 and 11 of the 4060) seems to be recommended to makd the adjustment. But this does not seem to work for me, so far.

Am I missing something, (e.g. knowledge)?

Isn't there some straightforward way to make the adjustment, that is, to shave 4.14 Hz off of the signal out of 4060 pin 3, or another alternative short of going to a programmable IC?

Thank you in advance for the very valuable advice. This a wonderful forum for me and I appreciate your expertise and the especially your fee (good will!)

JB

Hi,

I don't know how much you can trim the Xtal frequency with a capacitor.
I assume a couple of 100ppm. But yiu need about 50000ppm.
I don't think this is possible.
(That's the great benefit of XTALs, that the freqency has very good accuracy and can't be influenced in a wide range from external circuits.)

I'd use a simple microcontroller with 16 bit timer/counter. Easy to get any frequency in the 40Hz range with 0.1% accuracy.

Or you may use a PLL.

Or a VCO or an RC oscillator.

Klaus

I agree with Klaus.
To get exactly 40Hz with a division factor of 98304 you have start with 3.932160 MHz which is nearly 68KHz different from the 4MHz you have at the moment. A typical 4MHz crystal will only 'pull' two or 3KHz at best so you have no chance of reaching your target.

Consider something like a 6-pin or 8-pin (you can get both packages) PIC10F200 processor which has an accurate internal clock which you can divide to produce 40Hz very easily and they cost almost nothing. If you need other frequencies you may be able to generate them simultaneously too.

Brian.

Hi,

one of the best fit is to use a 4.5MHz XTAL.
divide by 1024
divide by 113
get 38,8897 Hz

This means a -32ppm deviation. I assume this can be adjusted with a capacitor... if needed ;-)

Klaus

Or better still, utilize a programmable-modulo divider like the CD4059.
I've generated 50 or 60Hz signals utilizing a 32.768Khz crystal with it.

the 32.768 can get me 40.96, dividing by 2^4 and then by 50. I'm searching for a 3.932160 M, which evidently was made for me, giving 40.000, a Baud legacy frequency. No such crystal found here (Philippines) as yet. Parts are very affoardable here, around 35 pesos (<$1.) Meanwhile I'm wiring up a 2.000 M crystal which promises 39.8597 ((/2^10)*49) and 9.9649, very close to another target frequency (10 Hz). This week I will move to circuit board which might coral some of the stray capacitance which has been plaguing me on the breadboard since day 1, a couple of months ago. Replacing the 4017 with 7490 (plus a 4060 / 4040 pair) already helped tremendously towards precision, the 4017 being some kind of Christmas tree lighting debacle. Without a real meter, I'm gotten pretty close (within 1-2 Hz) by seat-of-pants counting and a speaker that plays the Eb in a variety of octaves, albiet a little sharp. the 2.000 MHz soldered will likely get me close enough for my purpose, which is gamma wave synchrony.

Reference: Dr. Tsai at MIT's Picower lab and the December 2016 paper about 40 Hz therapy for Altzeimer's (in mice). [I am independent retired programmer now hobbist and amateur 'optigeneticist' - maybe that's why I am dragging my feet about the microcontroller....]

Thanks again.