Get Inside A TCXO Clock Chip

[Pete] wondered how real-time clock modules could be selling on eBay for $1.50 when the main component, the Maxim DS3231 RTC/TCXO chip, cost him more like $4 apiece. Could the cheap modules contain counterfeit chips?

Well, sure they could. But in this case, they didn’t, and [Pete] has the die shots to prove it. He started off by clipping the SOIC leads rather than desoldering — he’s not going to be reusing this chip after he’s cut it in half. Next was a stage of embrittling the case by heating it up with a lighter and dunking it in water. Then he went at it with sandpaper.

It’s cool. You can see the watch crystal inside, and all of the circuitry. The DS3231 includes a TCXO — temperature-corrected crystal oscillator — and it seems to have a bank of capacitors that it connects and disconnects depending on the chip’s temperature to keep the oscillator running at the right speed. [Pete] used one in an offline situation, and it only lost sixteen seconds over a year, so we’d say that they work fine.

If you’d like to know more about how crystals are used to keep time, check out [Jenny]’s excellent article. And if sixteen second per year is way too much for you, tune up your rubidium standard and welcome to the world of the time nuts.

28 thoughts on “Get Inside A TCXO Clock Chip

          1. Comparing the date code with the date of the blog post i wonder if these chips were pulled out of old equpiment and are now beeing “recycled”.

          1. You have a real gift. It’s very uncommon these days to find someone who can write more than one sentence and not use the word pseodo-word “proactive”, or the indignity “signage”, or the new howler, “gifted”; which, of course, is millennial-speak for “given”, or “gave”.
            Keep up the good work.

    1. IMHO it could be done with single pair of additional capacitors in a sort of PWM fast-slow regulation, but that would require more work, meaning more battery juice expenditure, and that’s probably why they opted for what seems to be a binary series of capacitances, which allows for more static setting of temperature compensation over the full temperature range.

      If one doesn’t need that much battery autonomy, has a bit more space on PCB, and doesn’t mind doing calibration on one’s own, temperature compensated timekeeping probably could be done using a dedicated general purpose low power micro with 32kHz crystal and some additional circuitry (three additional capacitors, a thermistor, and perhaps two small NMOSFETs) instead of RTC.

      1. The Devil is in the details. Phase Noise and Jitter problems increase dramatically with non-linear techniques without draconian analog efforts to tame them – which brings you back to square-one.

    1. It could well be, but I have no way of knowing. The Maxim rep I contacted says the package and die markings correspond to a legit lot made back in 2011, but would not provide any additional details (to prevent counterfeiting, they said). Of course, a cloner could have faked such markings as well, but that leads to a deep rabbit hole of being unable to determine such things.

      Also, is the use of a moderately specialized, not-widely-used clock chip worth the effort of cloning compared to, say, FTDI chips? Maybe. I’m a physicist, not a detective or one familiar with the economics of cloning chips.

  1. “[Pete] used one in an offline situation, and it only lost sixteen seconds over a year, so we’d say that they work fine.”

    Nice! But the data sheet says “The TCXO provides a stable and accurate reference clock, and maintains the RTC to within ±2 minutes per year accuracy from -40°C to +85°C.” so those 16 seconds could have been a lucky shot. Still much better than most 32.768 KHz crystals though.

    1. From -40 to +85! He’s well in the center of the operating temps, and gets much better than the worst-case results. Sounds about right. I have one running in our basement, and it hasn’t lost a minute in 1.5 years.

      YMMV, but as you say _much_ better than an uncompensated crystal setup, which can lose a minute in a month easy (20 ppm).

      1. They don’t ship internationally. You will have to ship the components to some local warehouse or shipping service, and have them ship internationally. It makes sense if you want to buy a lot of components together, to amortize the shipping cost. Otherwise, eBay is a better choice.

  2. These chips are made in China. I guess, as with lots of other products, they produce more than they were asked to do and after a while they dump the overproduction on Ebay. A wild guess but I have bought products way too cheap but as real as could be.

    1. A commenter on my site who used to work at Dallas Semiconductor said the wafers are made in Dallas (Texas), tested, then sent off to a third-party company for cutting into individual dies, packaging, etc., then returned to Dallas for more testing. That was his knowledge from a few years ago.

      While I don’t know what they’re doing these days, it wouldn’t surprise me if they kept the production of the wafers in-house.

      Another commenter says that DS3231s are often used in Chinese-made utility meters made by the millions and that much of the chips sold on eBay and the like are otherwise-good leftover chips from the production of such meters. I have no way of verifying this, but it seemed interesting.

    2. A collegue once asked his chineese provider how they could have such low prices. The answer was : when a big company asks for a batch of product, say 1 million pieces, we actually make 1.5 million, because as at this point, it costs the same. We keep the over production and ‘sell’ it for free, the price you pay is actually the delivery cost.

  3. “And if sixteen second per year is way too much for you, tune up your rubidium standard and welcome to the world of the time nuts.”

    Or a GPSDO, they are getting pretty cheap.

Leave a Reply

Please be kind and respectful to help make the comments section excellent. (Comment Policy)

This site uses Akismet to reduce spam. Learn how your comment data is processed.