Logic divider question

[throbscottle]

I made an ajustable interval timer using a 4060 oscillator/divider, which works well, but quite naturally gives a maximum useful interval of t/2^13. Since it's a 14 stage divider, is there a 'neat trick' to get an interval using the full cycle of the output instead of just the high or just the low portion?

 

[Pommie]

Sorry, you need to explain that better as I've no idea what you mean. Mind you, could be the wine, will look again in the morning, may make sense then.

Mike.

 

[dr pepper]

You can make the counter programmable by passing 0's through diodes to the reset input, after inverting it.
However that only reduces the max interval, sounds like you want to increase it.
I'm not stating but I dont think you change the max cycle count, as the clock signal between stages isnt accessible.
One 'trick' you might be able to do if you want a multiple of 2 division, maybe even /4 is to run the 4060's osc at 1/2 or 1/4 your incoming clock freq, and use said clock freq signal to synchronise the 4060's osc.

 

[throbscottle]

Ah, I've led everybody up the garden path, and I wasn't even trying!
It's being used as a stand alone 1-shot timer. So if the oscillator is running at 2kHz, for example, the 4060 is a 14 stage divider, so the 14th stage gives me an output of 0.122Hz, the period of which is 8.192 seconds.
But as a simple timer I can only use the high part or the low part of the output, so the oscillator needs to run at half the speed. So I was wondering if there is some way to use both half cycles and then have the timed interval end. OR'ing all the outputs together gets me most of the way. I was hoping there might be a "well known method" to use for this?

 

[AnalogKid]

Nope. I've used the 4060 in a lot of timing applications, and if you want the convenience of the Q14 output driving the downstream circuit or device, then you get only the half-period.

BUT - if you need only an edge to indicate the end of the period (as opposed to a level throughout the period), then a series capacitor and resistor to Vdd on the Q14 output will produce a negative edge at the end of the full cycle. This is a negative-edge differentiator.

ak

 

[throbscottle]

Thanks AK. I was trying to get my head around doing that. For some reason I can't remember, I'd discounted it, but I'll re-visit the idea.

 

[throbscottle]

I think it was because the output keeps a transistor on for the interval being timed and I didn't want to change the circuit...

 

[rjenkinsgb]

You could use a couple of sections of NAND gate to make a flip-flop.

Set it at power on or with the 4060 reset, then reset when the appropriate counter stage goes high or low, directly or via an C-R network to get an edge?

That allow any stage of the counter to be used, either high or low state.

 

[throbscottle]

That's a good idea - but also defeats the idea of a 1 chip circuit!

 

[Nigel Goodwin]

That's a good idea - but also defeats the idea of a 1 chip circuit!

Then chuck the 4060 away, and replace it with a PIC

Then you make it do whatever you want, and as long as you want.

 

[dr pepper]

There is another oscillator counter chip in the cmos range, it has loads of functions, cant remember the no. but you might be able to find it with google.
Otherwise might as well go with 2 4060's, or how about a 4040 series'd up to it.

 

[AnalogKid]

The CD4521 is a 24 bit oscillator/divider. For the same oscillator components, the longest output period is 1K times longer than the 4060. However, the outputs start at Q18, so if you want something only 2x or 4x of your 4060 circuit, at least one osc component will have to change.

ak

 

[gophert]

The CD4521 is a 24 bit oscillator/divider. For the same oscillator components, the longest output period is 1K times longer than the 4060. However, the outputs start at Q18, so if you want something only 2x or 4x of your 4060 circuit, at least one osc component will have to change.

ak

I just used a cd4521 to make a clock from a stepper motor with a cd4017 as a divider. Clicks once every 18 seconds to move the minute hand around once per hour for a 200-step motor. I stumbled on the cd4521 last week in my junk box as I was struggling with the cd4060. I didn't realize it had the onboard oscillator circuitry like the 4060. Just add crystal and two caps.

 

[schmitt trigger]

The CD4521 has such a long chain of dividers, that you can still employ common 32.762Khz crystals and still have a precise 512 second period.
For longer periods of many hours, you can use the powerline frequency.

 

[gophert]

The CD4521 has such a long chain of dividers, that you can still employ common 32.762Khz crystals and still have a precise 512 second period.
For longer periods of many hours, you can use the powerline frequency.

or you can cascade a second cd4521 and get a delay of more than 6 months. Or anything between by cascading a cd4060.