Accurate one second time base

[D'COUNCILOR]

Hi everyone!

I've just finished my digital timer but as of now I'm using the LM555 timer as the one second time base. But I need a time base for my project which is accurate.

I would really appreciate any help. Thank you.


D'councilor

 

[Nigel Goodwin]

Hi everyone!

I've just finished my digital timer but as of now I'm using the LM555 timer as the one second time base. But I need a time base for my project which is accurate.

I would really appreciate any help. Thank you.


D'councilor

The usual way is to use a 32KHz clock crystal and divide it down to give a one second timebase. Is your project battery or mains powered?, if mains powered you can divide the incoming mains frequency down (either 50 or 60Hz) - it's usually more stable than a crystal - in the long term.

 

[stevez]

How accurate a time base do you need? What might be important is being able to check whatever you do against a known standard. You might build a crystal controlled oscillator at a frequency that can be compared to a known standard - then divide from there. I might use WWV at 5 mHz in my area though I can't always receive it. I'd set things up to be able to hear both WWV and my oscillator in a reciever then adjust the oscillator to zero-beat against WWV. That might allow me to come within 30 or 40 Hz for a short time. You might use a local broadcast station as a point of reference as their accuracy requirements are quite high. These are crude "poor man's" methods.

You might also purchase a complete crystal oscillator. If manufacturer's instructions are followed then accuracy indicated by manufacturer should be expected, I would think. A crystal, all by itself is also not bad as long as the circuit accurately replicates the manufacturer's suggestions.

 

[laroche73]

check previous message threads

There is some information on this topic available in previous threads, including the one you started in October . A simple search is the best way to see which topics have already generated some discussion in these forums:

https://www.electro-tech-online.com/threads/what-is-good-program-for-pcb.135/

https://www.electro-tech-online.com/threads/any-one-of-these.4056/ As Phasor, csaba911 and Nigel pointed out, a simple, inexpensive and fairly accurate portable timebase consists of a 32 Khz crystal and a CD or 74HC 4060 divider/oscillator. The 4060 has a 14-stage divider, so you'll get a 0.5 sec output from a 32.768Khz crystal (2^15/2^14 = 2Hz = 0.5 sec). Add another flip-flop (half a 74HC74, for ex) to get the 1 second output.

 

[D'COUNCILOR]

One second time base

Hello!

Thank you very much for your response. I'll try all your suggestion. By the way I'm from the Philippines and were using 60HZ here.


D'councilor

 

[john1]

The easiest way i have found to get a very very acurate one
second reference is to use the guts from an unwanted digital
clock. They cost less than nothing, and they are very very
accurate.

 

[D'COUNCILOR]

One second time base

Hello all

Thank you very much to all your suggestion about my question regarding the one second time base. I've got a one second time base for my circuit by using the 60 HZ frequency from the mains power. It works great.
Now I'm enjoying my timer very much.

Can I ask another question? I am now doing a counter circuit using the CD4510 and CD 4511 as the counter and decoder. My question is how can I increment the count of this counter circuit by 3,6, and 9? I would really appreciate any help that you may extend to me.

Thank you very to all of you and more power to this forum.

d'councilor

 

[Dean Huster]

Accurate timebases

Using the line voltage as your frequency reference has it's limitations. For a frequency counter with the line voltage as a base, it cannot possibly have more than a five-digit accuracy even at best. Line frequency is accurate for digital clock use for the long-term, not the short term. Over weeks, problems with line frequency are averaged out. You can have some short term hiccups that limit you to only two or three digit accuracy at best. Five-digit accuracy translates into about ±0.01% while 3-digit accuracy is more like ±1%. A digital line-referenced clock is no more accurate than the analog version.

You can get oscillator-on-a-chip ICs from lots of sources, in nearly any frequency you want, including the 32,768 Hz (32.768 KHz) mentioned which you can divide down with a 15-stage binary counter string. Another option is a 1 MHz oscillator divided down with six decade dividers. The problem with an oscillator-on-a-chip is that you can't fine-tune the frequency like you can with a "discrete" design where you can pad the crystal with a trimmer cap to set the frequency dead-on. The advantages of the oscillator-on-a-chip are the low cost and tremendous space savings over a discrete design.

Here's idea #4853 for deriving an accurate 1 pps: Use the guts from a quartz clock movement that provides 1 pps to the solenoid that drives the second hand around the dial. Just get rid of all the mechanics and use some simple circuitry to get the logic to interface to your application.

Dean