DS3234 Real rime clock (lack of) precision

Hello,

I just received 3 modules from an eBay store and found out that 2 out of thee gain more than a second per hour, the third is pretty on the second. Did it occurred to one of you? Except from returning them, is there anything I can do to improve things

Thanks

The Maxim chip that is on there says it’s +/-2 min per year. There is no calibration. Sounds like you have a bad one. Could it be a counterfeit chip? I recently had 2 shipments of counterfeit Texas Instrument chips from China.

I did again the measurements with the following setup
a master board with a ds3234 from sparkfun
a slave board with one of the eBay parts
every hour the master prints the time of it’s RTC while sending a char via serial2 to the slave which then prints it RTC

All three parts gain from ~1 to ~4 seconds per hour and must be fake as you suggested

Thanks!

unfortunately, could not leave feedback as window closes after 60 days of purchase regardless delivery delay. Beware of RTCs from eBay: they cost much less but some are worthless

Did it occurred to one of you? Except from returning them, is there anything I can do to improve things

Yes of course there is something you can do to get it to run on time.

The DS3234 has an Aging Offset Register (10h/90h) that you can pull the oscillator with. You could try using that to see how far you can pull it. But seconds per hour is just plain not gonna happen with real devices.

tomCrawford:
The DS3234 has an Aging Offset Register (10h/90h) that you can pull the oscillator with. You could try using that to see how far you can pull it. But seconds per hour is just plain not gonna happen with real devices.

Thank you!
I am not sure what you mean by 'is just plain'. Does ir mean its too much for the fine tuning offered by the aging reg? Indeed, the part is speced +-2ppm and is actually ~1000 ppm so even if it can be set right good chances the temp comp won't work

Anyway I brought sparkfun parts and have thrown the eBay ones to the bin.

This discussion should warn forum member to be more careful and less naive than I was

Sorry I was ambiguous. What I meant was, seconds/hour error is a strong indicator of a part that never went through Maxim's quality control.

screenshots i sent to the eBay seller showing the three parts drift with time. The master has a sparkfun part found precise over more than a year

1 ran at night, 2 at morning, 3 during the afternoon

edit: Sorry, for some of the screenshots, got message below.

“There was a problem during the uploading of 2019-05-28 19_50_05-master3.txt - Notepad.jpg.
Your post has been made, however the above attachment was not attached. Please use the Back button to edit your post and submit any required changes.
Your attachment has failed security checks and cannot be uploaded. Please consult the forum administrator.”

Here is a pointer to a thread on the Parallax forum that discusses the aging register.

I dont have a precise frequecy meter: it took me a long time to understand that i have worthless chips

Maxim sells DS3234 chips for two to three bucks each, in very large lots, such as a Chinese manufacturer would buy. If the breakout board with connectors, resistors, capicitors, battery holders, etc. is less than that, they probably didn't buy the chips from Maxim, or they may be used, discards or improperly handled.