This probably isn't the best title either........

I have constructed and oscillator using a 3.2768MHz crystal and I want to utilise the 200Hz output from Q14 as the basis for subsequent 20Hz and 10Hz dividers. These dividers consist of a CD4017 and a CD4027 to achieve these frequencies. The oscillator is held in its reset state initially and when the reset pin is taken low I would expect the internal delay between this action and the pulse output train at Q14 to be no more than 5mS. The Fairchild CD4060BCN I am using exhibits a delay of 7.4 - 7.6mS from reset removal to Q14 output. I have 4 devices all from the same batch and they all exhibit this order off delay.

I purchased a batch of new TI devices and ran the same test and these devices exhibited a delay of only 4.8 - 5.0mS - just as should be expected.

For some unknown reason the Fairchild devices seem to have a fixed internal delay of approx 2.5mS added to the normal propagation through the 14 ripple carry counters. This is causing my 100mS delay circuit to produce a delay of 102.5mS and it should not be the case.

Has anyone encountered this phenomenon before and if not then can you offer an explanation of what might be causing this excessive delay? Shouldn't CD4060 from different manufacturers be interchangeable without any operational changes?

Scope traces attached show CD4060 reset pin 12 top and Q14 output bottom.

 

hi rher,
The CD4000 series are not 'fast' switching devices.
Have you checked the HEF versions of the devices you are using.?
E
Update:
Check this PDF

 

If you look at the internal circuit, you'll see that RESET disables the oscillator.
Pierce oscillators don't start instantly. Have a look on pin 9 to see how long it takes your oscillator to get started. It will depend on the gain of the inverters between pins 11 and 10, which will may vary.

What is your power supply voltage? The 4060 will only manage 3.5MHz at 5V, increasing to 12MHz at 15V. The power supply voltage affects the gain of the inverters, and so will also affect the startup time.
If you are on a 3V to 5V supply, you could try a 74HC4060, but I suspect there will still be a delay.
NXP uses the same oscillator on its microprocessors and states that one should allow 500us for it to start.

 

I’ve gone through several variations of the 4060 and have found that they some have unique qualities. I have a circuit that will only work with one brand and model. If you find that it works with TI I suggest sticking with it. If it’s not on the datasheet, it is not specified or guaranteed.

 

All crystal oscillators require time to startup as already pointed out.
Your solution is to have the crystal oscillator free-running and gate it into your counter.

 

Make a crystal oscillator using a 74HCU04, then feed it into the clock input of a 4020 or even into pin 11 of the 4060. Then the 4060 reset circuit can't stop the clock. You will then only experience a startup delay the first time it is switched on.