Continue to Site

# How I can do? (Oscillator Crystal with 74HC4060)

Status
Not open for further replies.

#### Cao Tung Lam

##### New Member
I want to build an oscillator using IC 74HC4060 with output frequency 50Hz or 60Hz. I read the datasheet of the IC and I found the circuit that I don't know. The question are :
If I want to make 50Hz or 60Hz output frequency, how do I
calculate the value of the crystal that I must use?

generating 50/60Hz from a 74HC4060

Lamct, you won't be able to generate 50 or 60Hz using only a 74HC4060 and a crystal, because the necessary input frequencies are not standard crystal frequencies. In addition to the 74HC4060 (14 stage binary counter), you will need an odd modulo divide-by-N counter to get 50/60Hz.

In other words, the 4060 can only divide by powers of 2. Trying (60Hz x 2^14) gives a required input frequency of 983.04Khz. You won't find a standard crystal at that frequency or at any integer multiple.

However, by following the 4060 with a two stage divide-by-N counter, a 50 or 60Hz output can be generated.

The easiest way to do this is to pick a standard crystal frequency in the 1 - 20Mhz range. Just don't use simple integer multiples like 6.000 or 10.000Mhz, we need something that a binary counter can divide down to a useful intermediate frequency.

Examples:

1) Try 11.0592Mhz (a common value for baud-rate generators)

(11.0592Mhz/60Hz) = 184320; (184320/2^12) = 45

A two stage divide-by-N counter with N1 = 5 and N2 = 9 will produce a 60Hz output.

Likewise,

(11.0592Mhz/50Hz) = 221184; (221184/2^11) = 108

N1 = 9, N2 = 12 will produce 50Hz.

2) Try 6.144Mhz (another standard crystal frequency)

(6.144Mhz/60Hz) = 102400; (102400/2^12) = 25; N1 = 5, N2 = 5.

(6.144Mhz/50Hz) = 122880; (122880/2^12) = 30; N1 = 5, N2 = 6.

A simple two stage divide-by-N circuit can be found here:

The CD4018 also works well for building divide-by-N counters.

There are plenty of other ways to generate 50/60Hz from a crystal, such as dividing down to 1Hz and using a phase locked loop like the 4046 to multiply back up to the desired frequency. However, a two stage divide-by-N counter is still required in the PLL feedback loop to get 50 or 60Hz (where N1, N2 = 5, 10 or 6, 10).

National Semiconductor used to make the MM5369, a nice part that used a cheap TV colorburst crystal (3.58Mhz) to accomplish what you're trying to do, although only at 60Hz.

The part is now obsolete, but you may still be able to find some. Note that it shouldn't be too hard to program a small PIC micro (like the 12C508) to duplicate the function of the MM5369.

- CAL

Divide by N counters

I should mention that the divide-by-N counter doesn't have to be done in two stages. Divide-by-N counters are commonly built by cascading 4-bit binary or decade counters. A single chip alternative for this application is the 74LS592 or HC592. It's an 8-bit loadable down-counter with a terminal count output. You use the RCK and /CLOAD pins to load in the appropriate divisor. This IC is harder to find than 4-bit counters like the 74HC161 and CD4018.

- CAL

60Hz oscillator chip

Lamct, I hadn't seen your previous post on the subject or I wouldn't have mentioned the MM5369. I see you are trying to replace it. The previous posts describe a valid way to generate 50/60Hz, but if your clock is line powered, why not use a simple zero-crossing detector to replace the MM5369 circuit?

60Hz oscillator

Hi Raloche 73,

I don't know use a simple zero-crossing detector. How can I do? Can you help me expand it!

Thanks and best regards!

Lamct

60 Hz zero-crossing detector

Lamct, attached is a basic circuit for getting a 60Hz square wave off the secondary of your power supply transformer. Technically, it's not a true zero-crossing detector, since the nominal switching point of the 74HC14 will be 2.5V with a 5V supply. However, the schmitt trigger will switch at the same point in every 60Hz cycle, and that's what matters if you're just using the output as a timing source. (True zero-crossing detectors usually operate off the AC mains and are used to switch triacs or scrs when no current is flowing).
Notice that you feed the schmitt trigger with a half-wave rectified, NON-filtered waveform that has been scaled down by the resistor divider. The total resistance of R1 + R2 is not critical and can be between 10K - 100K. Pick the divider ratio to keep the peak input voltage to the 74HC14 under the 5V supply rail. The schmitt trigger's output frequency will be 50 or 60Hz, depending on your AC system, but the duty cycle will not be 50%. Again, this shouldn't matter if you're just using the output as a timing source. The 74HC14 could be replaced by a comparator like the 311 as long as you add in some hysteresis circuitry.

- CAL

#### Attachments

• schmitt_850.gif
7.6 KB · Views: 2,097
Although the line frequency is a simple way of getting the desired 60 Hz, and one ive used myself at times, it doesnt allow for much portablity.

There is away around the problem of the 74HC4060's division by multiples of 2 by using a decade counter ,such as a 7490, just before the crystal oscillator. You can get division by 2,4,5,6,7,8, and 9 by manipulating the binary outputs.

Here are a few ways of doing it:

This will give you more freedom for choosing a crystal.

crystal frequencies

The only drawback to using a single 4-bit odd-modulo divider with the 4060 is that you're limited to using ~ 1 Mhz or less Xtal's (60Hz x 5 x 2^12 = 1.2288Mhz). This isn't so bad, Epson offers a line of low-cost crystals in the range of 32 - 149Khz, C-2 type. You won't be able to use the cheapest though - 32.768 watch crystals. That was the beauty of the MM5369, it used a very inexpensive crystal (3.58Mhz TV colorburst).

- CAL

Laroche, those are some fine ideas you presented. However, your diode bridge is smoke waiting to happen. You swapped the GND and the low side of the secondary connections. I'm sure it's a typo (drawo?).

Good eye Ron, i thought there was something odd about it.

!

Thanks for the good eye, Ron. I felt queasy about posting that so quickly, and I played with the diodes more than once (I'm going to modify TinyCad to include the standard diagonal 4-diode bridge). I'm new to the board and this is the second good mistake I've made in two days. Mechie caught the last one and had the good grace to fix it. They don't call us "novices" for nothing!

It just proves every reader should go through the logic of posted solutions, and decide whether or not they work. Thinking for yourself is a good thing...

- CAL

error error

Oh, great. Now no one's going to trust my schematics. I spent much more time figuring those out than drawing the diode bridge. It just goes to show you that the simplest things are often overlooked. I welcome any feedback on my other posts, humble pie tastes just fine.

Need'nt sweat it Laroche. Just think of it as proof reading. Ive made similar if not worst bloopers than that. Actually ive seen some bad blunders made in published electronics technical books. How they got past editing is beyond me.

Status
Not open for further replies.

Replies
22
Views
4K
Replies
12
Views
2K
Replies
18
Views
2K
Replies
3
Views
2K
Replies
5
Views
2K