i tried 2 different 555 chips so far

Where did you purchase the 555's?

The faded marking on the top one screams counterfeit to me. The bottom one also looks suspicious. Compare the texture of the top and bottom of the chips to see if they're the same and examine the shoulder area of the leads for pitting.

 

Where did you purchase the 555's?

The faded marking on the top one screams counterfeit to me. The bottom one also looks suspicious. Compare the texture of the top and bottom of the chips to see if they're the same and examine the shoulder area of the leads for pitting.
View attachment 152445

Bottom one is purchased on DigiKey and top one is from small local shop.

 

Bottom one is purchased on DigiKey and top one is from small local shop.

Take the top one back to the shop and tell them it's counterfeit and see what they'll do. A reputable seller would test it to confirm, or show you that it isn't. Manufacturers want their package markings to survive typical cleaning.

Does DigiKey give guarantees that parts are genuine?

At any rate, check both parts to make sure the top and bottom of the packages have the same texture. There should be no pitting on the shoulders of the leads; pitting in an indication of sandblasting to remove the original markings.

I don't recall if surface mount parts have mold markings. If present, they should be well defined. In this case, the bottom hasn't been altered, the top looks like it's been remarked.

 

Any update?

The circuit I breadboarded has been running for almost 4 days. The period has been pretty consistent, varying about 5 us (from 1.55x ms); but that might be from not letting the frequency counter heat up.

 

First I have to thank everyone for great response and help.
You helped me a lot to troubleshoot weird problem .

Circuit is now running for past 24+ hours without issue. After few hours frequency stabilized and is now fluctuating between +-0.1 what is more than enough accurate.

So for all of you who helped with suggestions i have to do some explaining. I tested all of your suggestions and some improved stability but none of them solved issue...

I did not think i will have to say it but i did big rookie mistake. Same time as i bought 555 in local shop i bough also new flux for soldering (old one was gone empty)... And of course i did not check what seller sold me. i trusted him it will work. Indeed it worked ok when soldering, but i did not check that flux is conducting and requires cleaning. With old flux i did not need to clean and i usually used just cloth with some alcohol to wipe excess flux. Funny thing is i checked with multi-meter only resistance between vcc and GND and i visually expected all pins...

Lesson learnt... i cleaned all old soldering, used new flux and its funny that my first no modification circuit works perfectly...

Once again thanks for all suggestions and help sometimes basics are bigger issue than we think.

 

This is the second time in the past week or so someone has had a problem with conductive flux!

Fluxes suitable for electronics use should most definitely not be conductive, with the exception of water-soluble organic acid fluxes. Water soluble flux must be removed within hours, otherwise corrosion products that are very difficult to remove can be formed. For very high impedance circuitry, it may be necessary to remove flux, regardless of the type. No-clean flux is not no-residue flux and is incredibly hard to remove.

There are many many vendors on ebay selling counterfeit fluxes, claiming they are from a major US manufacturer.

 

Ah the 555: simple, cheap and good, the Corn Flake of analog ICs.
Been a Crescent wrench in my toolbox since '78.
Never had the issue described here, I agree it's something thermal.
Hmmm, I've been bit by bad flux residue too.
But, if freq drift is a problem, I'd go crystal osc & a counter/divider.