Simple frequency devider - prescaler using a counter?

[bbiandov]

Hi

I'm wondering if one can build simple frequency devider with a simple BCD 4 bit counter such as 74LS169? The RCO will be the output.

Yes, I realize that the duty cycle of the output will be a departure from the perfect 50% but so what? Is it a viable option to use a counter as devider?

~B

 

[ericgibbs]

Hi

I'm wondering if one can build simple frequency devider with a simple BCD 4 bit counter such as 74LS169? The RCO will be the output.

Yes, I realize that the duty cycle of the output will be a departure from the perfect 50% but so what? Is it a viable option to use a counter as devider?

~B

hi,
Yes thats OK.
If you are concerned about the 50% form factor, look at the datasheet and see that the ripple carry output is equal to a clock input period.

 

[Nigel Goodwin]

Hi

I'm wondering if one can build simple frequency devider with a simple BCD 4 bit counter such as 74LS169? The RCO will be the output.

Yes, I realize that the duty cycle of the output will be a departure from the perfect 50% but so what? Is it a viable option to use a counter as devider?

That's all prescalers are - to get a 5% duty cycle, simply add a divide by two flip flop on the end (giving an extra divide by two).

 

[Dean Huster]

That was always the nice thing about the old 7490 decade counter (old in that it was within the first offerings Signetics had for their 7400-series TTL). The 7490 (and all its offspring and several similar counters) consisted of a divide-by-5 and a divide-by-2 counter. If you ran the output of the 2 into the 5, you had standard BCD output but a non-square decade output. Or you could run the output of the 5 into the 2 and get the same decade result with a 50% duty cycle at the sacrifice of not being able to use the ABCD outputs to drive a display decoder because things would be all scrambled up.
If you want a decade divider AND a square output (adding a binary divider to the output of a decade divider will give you a very inconvenient divide-by-20 and you'd have to remember to convert your counter display), consider modifying a decade or 4-bit binary counter to do a divide-by-five and then add a binary counter to the end. Two or three chips (depending up the chip type) for the output you might desire, although there's no need to preserve a square input for a frequency counter.

Remember that if you use a prescaler like that, you destroy your counter's resolution by a factor of 10. With a 1-second gate time, you'd normally read the frequency to the nearest Hertz. With a prescaler added and using the same gate time, you'll be reading the frequency to the nearest 10 Hertz.

If your use for this circuit is as a prescaler to extend the upper frequency of a counter, don't forget that the chips you use for this "front end" need to have a frequency capability that is 10 times that of the counter's capability for maximum range or your prescaler will limit your top end.

Dean

 

[bbiandov]

Thanks Dan

Actually I am using this as a speed controller to a stepper motor. I plan on loading the devide factor at the counter (4 bits) and then that will effectively give different speeds since the output frequency will change based on the count-from value (I am doing reverse count to zero when the RCO is triggered and that is my output). The loading comes out of an external HP jet direct print server so I can control the devide factor via IP over Ethernet and hence my stepper speed is controllable via IP.

The problem is that I can't get good reading on my output. The freq counter goes heywire with 1s gate time. Its not my cable nor the instrument cuz I get perfect reading on the clock source (produced by 555 timer) which is higher than the output no matter what. My freqs are in the 50Khz range so I am not pushing the envelope with this.

My only guess is that due to the extremely skewed wave on the output I am not able to get good reading with the freq counter (which BTW goes up to 1GHz)

Any suggestions guys? I am cool with sacraficing another /2 in exchange of better duty cycle ONLY if that will allow me better reading on the output signal. I can crank up the 555 much higher if I need to

Thanks
B