Numbers

[Lenp]

Anyone that has been around this 'business' will probably know what these numbers are related to, but, what was the rationale that created them?

50 Hz and 60 Hz
256Khz, 455 Khz and 10.7 Mhz
3.579545 Mhz

Add on if you will!

Len

[Bigglez]

*** Redacted by poster ***

[MrAl]

Hello,

3.14159kHz, 1.2345678kHz (ok just kidding)

For my Freq counter i used 32768Hz. The rational there was that
that frequency is divisible by perfect powers of 2. For example,
32768Hz/2^15 is equal to 1Hz, which makes a nice time base.
32768Hz/2^15 is equal to 0.5Hz, of which a square wave at
this frequency produces a nice gate signal for the frequency counter
(1 second open, 1 second closed). That's why i do, using a LSI
divide chip along with a crystal oscillator running at 32.768kHz.
Many watches use this freq too for same reason.

The other reason i used that freq was because i could build the
whole oscillator/gater circuit with only two ic packages:
one package is the osc, the other the divider, and it's done.
I didnt use PICs at the time, but even so using a PIC i would want
to use a crystal or crystal osc anyway to get reasonable accuracy
for the gate period...i would not depend on the internal osc for that.

[Lenp]

Bigglez

Like most others, you know what the numbers are for, but the real question was WHY were THEY chosen?

MrAl
Logical choices stand the test of time!

Len

[Bigglez]

*** Redacted by poster ***

[Robert Reed]

I believe 60 Hz is the optimum frequency for power companies as a comprimise between iron mass (Transformers) and line losses ( long High tower transmission lines).
455 khz and 10.7 Mhz were originally mathematically computated as the optimum IF frequencys in regards to their associated carriers ( 455KHz AM - narrow band & 10.7 Mhz FM - wideband). As time went on they became popular and cheap (due to mass production) that they were universally used in many other radio services receivers. However, for every radios operating carrier frequency, there is an optimum IF frequency as witnessed by me over the years working with a myriad of radio equipment. You want numbers - I could type for an hour of all the IFs i have seen in use. And there is good reason why all these different frequencies depending on the radios service band.
Most of these oddball frequencies we see for various uses were carefully thought out by people smarter than us so as to satisfy the conditions and environment they were used in.

BTW MrAl- I love those 32KHz cylinder crystals. It saves so much extra division when shooting for a 1.0 second time base. All I need then is just one leftover section of a 'Nand' gate (Oscillator) and Voila, its done!

[Bob Scott]

Anyone that has been around this 'business' will probably know what these numbers are related to, but, what was the rationale that created them?

3.579545 Mhz

The video sync and colour subcarrier can all be counted down from 14.31818..MHz. It's an infinitely repeating decimal.

14,318,182 / 4 = 3.57954545... MHz colour subcarrier
14,318,182 / 455 = 31,468.53 Hz
31,468.53 / 2 = 15,734.26 Hz Horizontal Sync
31,468.53 / 525 = 59.94 Hz Vertical Sync

Notice that the colour subcarrier (3.58M) is an odd multiple (455) of half the horizontal frequency to produce frequency interleaving. Theoretically this places the frequencies of the multiple sidebands of chroma in between the multiple sidebands of horizontal frequency and vertical picture elements.

Using these frequencies, the TV sound subcarrier at 4.5 MHz* is exactly the 286th multiple of the horizontal frequency. I find that odd.(??) But it must have been planned that way by the original NTSC comittee in order to explain why they shifted from the original H and V frequencies of 15,750Hz and 60Hz to such a weird system with repeating decimal places.

* I know the sound is broadcast independently on a carrier with a frequency 4.5 MHz above the video carrier but a normal TV set sees a 4.5MHz audio SUBcarrier come out of the video detector. It is frequency modulated, so its centre frequency not exactly locked to any video signal.

BTW, I have been a TV Broadcast engineer at CKY Winnipeg and CKVU Vancouver.

[Dean Huster]

Numbers aren't the only oddity. In the SI system, every negative integer-multiple-of-3 power of 10 ("engineering notation") has as its metric prefix a lower-case letter: m, µ, n, p, f, a, etc.. But every positive integer-multiple-of-3 power of 10 has as its metric prefix an upper-case letter (M, G, T, etc. EXCEPT for kilo- which is a lower-case "k". Makes absolutely no sense. In my teaching, I always told the students that it's capitals for positive powers, little letters for negative powers, not worrying about the slight error of kilo- simply because it doesn't matter and having it the way I deliberately and erroneously teach it makes all the sense in the world and is consistent.

I wonder what the reasoning was for k vs. K? I've not found an answer in all the NIST pubs on the subject.

Oh, I believe that 455KHz was selected to be below the BCB and above the regular LF band transmissions. Maybe. But the actual exact selection of 455? Who knows? It's 10.7 for both FM and TV audio. Both FM and TV broadcasting developed about the same time, so it would make sense that they'd be the same since TV was just FM radio with a picture added.

Dean

[Robert Reed]

Bob
In your reply of
"Using these frequencies, the TV sound subcarrier at 4.5 MHz* is exactly the 286th multiple of the horizontal frequency. I find that odd.(??) But it must have been planned that way by the original NTSC comittee in order to explain why they shifted from the original H and V frequencies of 15,750Hz and 60Hz to such a weird system with repeating decimal places."
I would like to bat that around a bit. The bandwidth of a TV channel is 6 MHz and that consists of AM Vestigal side band modulation (halway method between double side band and single sideband) . More specifically, the lower sideband being
1.25 MHz and the upper sideband being 4.5MHz. That totals 5.75 MHz of its allotted space and the remaining 0.25 MHz is known as the 'Guard band' to complete the full 6 MHz bandwidth of one channel. The Guard Band is a Dead Zone in which steep low pass filters take their effect. This absolutely eliminates any carry over frequency to the next adjacent channel. Now, it was always my understanding that the 4.5 MHz subcarrier for sound was placed at that very edge of the channels bandwidth for two reasons. First and naturally to keep that info out of the video stream and secondly because of the beginning stages of sharp low pass filtering that takes place at that point, any unintentional reduction of the audio subcarrier would have little effect since the audio is not only less effected by a weak signal, but also has its own IF stages to further amplify that signal. Please correct me if I am wrong.

[Bigglez]

*** Redacted by poster ***

[philba]

44.1KHz. sampling rate for "redbook" (CD) audio. It was chosen as a compromise between Philips and Sony. One of the requirements was that the digital audio have 20Khz cut off thus dictating a minimum sampling frequency 40Khz, per Nyquist. Sony originally had used 47.25 KHz so that lower order filters could be used. 44.1khz was apparently good enough.

[Bigglez]

*** Redacted by poster ***

[Bigglez]

*** Redacted by poster ***

[dyarker]

1.2288MHz divides by powers of 2 for standard serial data rates.
For example: 1228800Hz / 128 = 9600Hz.
Why the data rates 300, 600, 1200, etc? I can't say that I know for certain, but 75bps is very close to what was a common electro-mechanical current loop (teletypes) bit rate. 75 * 2 = 150, 150 * 2 = 300. As technology allowed faster rates, they just kept doubling. So, 1.2288MHz crystals are easy to find.

[Bob Scott]

I would like to bat that around a bit. The bandwidth of a TV channel is 6 MHz and that consists of AM Vestigal side band modulation (halway method between double side band and single sideband) . More specifically, the lower sideband being
1.25 MHz and the upper sideband being 4.5MHz. That totals 5.75 MHz of its allotted space and the remaining 0.25 MHz is known as the 'Guard band' to complete the full 6 MHz bandwidth of one channel.

I checked the illustration from my old Sam's book that I bought way back in 1969. The video sidebands extend from 1.25MHz below the carrier (the vestigal one) to 4.25MHz above the video carrier. This leaves room for the sound carrier at 4.5 MHz above and its multiple sidebands, and a guard band.

TV sets have quite sloppy video IFs and add a 4.5MHz "sound trap" to keep the audio subcarrier out of the video. Chroma trap too for the Y signal.

If we are going into this deeply, let's discuss the chroma signal. It's centered at ~3.58MHz above the video carrier. It is modulated by two signals, "I" and "Q" which are in phase quadrature with each other. The "I" signal sideband extends 1.5 MHz below and 0.5 MHz above the 3.58. The "Q" signal is limited to +/- 0.5 MHz. The upper sidebands can't extend any higher because there is not enough channel bandwidth.

Here is where it got interesting for me and I'm making an assumption that because the upper sideband of the "I" signal is truncated at 0.5 MHz, the lower sideband that extends to 1.5 MHz below uses up quadrature space above .5 MHz because the upper sideband is truncated. I think that is why the "Q" signal can't also extend down there and has a different bandwith.

I guess I mean that I don't think you can independently quadrature modulate an AM carrier (with two signals) and make it SSB at the same time, sort of like Gerald Ford trying to walk and chew gum.