Best way to make a 555 retriggerable?

As a practical man I'd say that one gets more and more accurate estimates for the frequency when one measures longer and longer. Just like the (dc) power gain estimate of a latching relay increases without bound (which could be practically called 'approaching infinity') when measured longer and longer. The dc-SQUID coincidentally has the same property: power gain approaching (practically) infinity when signal frequency approaches dc. MOS transistors would behave similarly if their gates did not leak, if I'm not mistaken?

In the pedantic viewpoint the 18.5 kHz measurement accuracy could be brought down to femtohertz level, maybe, before the sun expands into a red giant. Even then one can imagine fleeing in a spaceship with the measuring apparatus onboard. Reaching atto-Hz accuracy then depends on whether the universe is cosmologically open or closed.

Actually, I'd expect the convergence towards the 'exact' frequency to be slower than linear in time, once the reciprocal of the 1/f noise corner of the apparatus is exceeded. I'm too tired to think now what consequences the Heisenberg uncertainty would bring in - it works to the other direction I figure?

Regards, Mikko

No "use a pic" posts are allowed without a schematic, software (firmware) listing, and programming instructions.

Thanks, Rich

To the ordinary roob in the street, "infinity" is "how much electricity is in the light socket." ;-)

Cheers! Rich

I guess that depends on the parameters of "exactly." To the nth degree, obviously forever, and that _still_ wouldn't get it, because before the signal started, its frequency was zero.

;-) Rich

Does the OP really need that accuracy, or is he attempting to know H-sync so he can switch to the appropriate synchronization? ...Jim Thompson

Probably the only thing you need to change in the source code is: processor 12c509 __config 0xfea

If using a 12c508 you should use processor 12c508 instead of processor 12c509

The _config 0xfea line gives the PIC the configuration word. That word configures the PIC for how you need it to operate in general, before it attempts to execute any program instructions. The configuration word in the source code tells the 12c508 to:

I mention this, because you might need to change the configuration word if you use a different PIC, and you might need to change the hardware, too.

For example, AIRC the PIC16F84 doesn't have an internal RC timer, so you would need to wire the R and C externally to that chip. That information is seen in the configuration word definition. The configuration word is defined in the datasheet. It all sounds very complicated, and it is the first few times. But you can get help here.

As to why you can't get the 12c508 to work, assuming you changed the processor 12c509 in the source above to processor 12c508: I'm guessing your MPASM suite has nothing for the 12c509 (or 12c508 if you changed it), so you could look for that at microchip. Here's what I have for the 12c508 in my MPASM Suite\LKR directory:

And in my MPASM Suite directory I have: P12C508.INC P12C508A.INC

I haven't tried the code - don't have a 12c508 - so what I posted is just for reference and not a tested solution. :-( But if you're missing the link & include files for the '508 you need to add them to get it to work.

Ed

Jim Thompson wrote: : On Thu, 10 Nov 2011 19:33:16 +0200 (EET), Okkim Atnarivik : wrote: : > In the pedantic viewpoint the 18.5 kHz measurement accuracy could : >be brought down to femtohertz level, maybe, before the sun expands into : >a red giant. Even then one can imagine fleeing in a spaceship with the : >measuring apparatus onboard. Reaching atto-Hz accuracy then depends on : >whether the universe is cosmologically open or closed.

: Does the OP really need that accuracy, or is he attempting to know : H-sync so he can switch to the appropriate synchronization?

True, the OP doesn't need the femto-Hz accuracy.

Regards, Mikko

Me, I'd run a XTAL (or tight R/C) controlled count-down to 18.5kHz, use a digitally phase-jerkable oscillator and compare at the end of each cycle. (Basically a frequency-only comparison... accuracy =

I've done this at much higher frequencies... as in satellite down-links. ...Jim Thompson

Life's easier if you remove the requirement to measure exact equality... of anything physical.

It's fun to trigger a scope from one atomic clock and look at a second one on a vertical trace. Zoom up to, say, 1 ns/cm. It looks triggered. Come back in half an hour or so, and it's drifted a little.

I'm not sure how people compare atomic clock frequencies in the big leagues. They are talking stabilities of 1 part in 1e18 these days.

John

ISTR some of the ultra stable devices are optical, like aluminum or mercury ions. They have a technique to count down to rf frequencies, so they can do the comparison at high GHz frequencies. This makes it much easier to get down to 1e-18 in a reasonable time frame.

Mike

Larkin follows Popular Science Magazine closely ;-) ...Jim Thompson

Have I mentioned lately what an idiot Thompson is?

John

I've been busy on a project, so I'm late to this discussion. Retriggering a 555 one-shot _accurately_ is non-trivial.

But here's a way to do an accurate frequency compare...

which is a variation of a floppy data extractor I did in 1983...

at GenRad, to handle extreme floppy-drive speed variations in portable (briefcase) testers.

It boils down to being a "jerkable" VCO... in the frequency-compare case, clocked by 1MHz... use a 555 if you must ;-)

I've also used this scheme on a satellite down-link for clock restoration. ...Jim Thompson