[sdiy] A440 -- almost

[fmg]

I'm currently building a 440 hz reference (with some harmonics) to
test VCO linearities (tracking) based in a PAL 3.582056 Mhz xtal.
I'm using a CD4040/CD4082 combination to divide Xtalfq by 8141.
It currently produces a 440.0012 hz. I think that a sligh tuning
to the xtal would be capable to make it run more close to 3.582040,
but I'm not sure if I could reach that freq. I haven't a freq counter
and using a lend one just for a while (and it's hard to figure out a
frac of Mhz counting milimeters in an oscope grid :)

Once I have it ready I'll post it in a week or two. If I can use
a freq counter for more than an hour...

I think that a not-beyond-manufacturer-spec tuning will produce
a nice 440.0007 Hz. (+0.0003/-0.0001 Hz under thermal circumstances)

Will be it enough for a sdiyer ?
(I hope it will not...)

I'm also working in a combination of CD4040/CD4068/CD4013/CD4017
to produce 4400hz - 880hz - 440hz - 55hz from the same PAL xtal. Due
to the grossly division implemented the 440 will not be so nice, but
it doesn't care for me cause in this circuit I have in mind "tracking"
and not "tuning".

happy waves.

Fabio Gonzalez
Posadas, Argentina



Michael Baxter wrote:
> 
> Hi all synthDIYer friends,
> 
> I'm designing and building a modular system, and I'm very excited to see
> all of the activity and innovative designs in this community! As I get
> circuits designed and built, I'll post them somewhere. I did of lot of
> this kind of design in the 1970's, so I'm hoping to contribute to "module
> diversity," if you will. More later.
> 
> In this first post, thought I'd share some quick notes on the A-440
> reference frequency or calibrator thread that came up recently.
> 
> I discovered that the NTSC color burst reference frequency is very close
> to a necessary superharmonic. 3.579545 MHz / 8134 yields 440.07192 Hz.
> 
> Compared to a true 440.0 Hz, this is still pretty close. The "correct"
> reference frequency would need to be 585 Hz lower, and this is probably
> out of the the crystal adjustment range, esp. as the NTSC frequency is
> supposed to be within +/- 10 Hz. However, quantitatively, the actual
> difference seems small compared to a +/- 1-cent interval above A4.
> 
> The hack to "adjust" crystal mechanically is very interesting, I had not
> seen that before. It definitely has the feel of amateur radio hints
> and kinks lore, and probably works. Starting with a cheap crystal, one
> could try several times, to eventually get it right. Not so different than
> Vbe matching transistors, :-).
> 
> I did see the later posts on the 4.40 MHz crystals, which would be of
> course "spot-on." But I was very surprised at how close one could get with
> the garden variety NTSC crystal, the kind laying around in your junkbox.
> 
> Kindly yours,
> Michael Baxter
> California, USA