No Script, No Fear, All Opinion
RSS icon Home icon
  • EEVblog #265 – Philips PM6672 Timer Counter Teardown

    Posted on April 4th, 2012 EEVblog 14 comments


    http://eevblog.com/files/PM6672-Service-Manual.pdf
    http://eevblog.com/files/PhilipsPM6673-76-Service-Manual.pdf
    PHOTOS: http://www.flickr.com/photos/eevblog/sets/72157629731591013/

    The first of the new regular segment on the EEVblog – Teardown Tuesday.
    Dave cracks open the Philips PM6672 1GHz Timer Counter, and more interestingly, the PM9690 10MHz oven stabilised reference crystal oscillator.

    Be Sociable, Share!
     

    14 responses to “EEVblog #265 – Philips PM6672 Timer Counter Teardown” RSS icon

    • It may be that the ALOX hybrid is actually the heater… The transistor on the copper shield may be the drive transistor. The ‘power’ resistor could be printed on the hybrid, together with a diode as temp sensor. this gives excellent thermal coupling.

      they probably compare external temp to internal temp to make sure there is no gradient.

      That ‘ceramic’ of the hybrid is called ALOx, short for Aluminum oxide. Yep, you got it right, that white slab is actually aluminum oxide powder pressed into a plate and sintered at high temperature. Just like ferrite is ironoxide compressed and sintered.

      Aluminum oxide is a very very good electrical insulator , while at the same time still being a good thermal conductor (since it’s a metal ) It is only surpassed by BeO ( BerylliaOxide ) in terms of thermal conductivity, but BeO is a controlled and restricted substance and very dangerous, as inhaled BeO dust causes a lung disease called Berylliosis ( similar to tuberculosis, TBC is bacteria causing the growth of granulomas in your lungs. Beryllium also causes granulomas in your lungs. No cure, and it will kill you over time )

    • I can confirm the hybrid is very likely the heater itself. Years ago I got a bunch of surplus TCXOs I couldn’t find any data on the net so I had to saw one to find the actual pinout being the “black box” approach not that easy (turned out they had two different voltage supply pins and two regulation inputs). Anyway, they had an ordinary xtal sandwiched into a lot of thermal grease between two of those ceramic hybrids each of them containing a few low power resistors. The power transistor probably acts as a current regulator that keeps temperature constant through some PTC/NTC.

      Here’s some info on how to build a dirt simple ovenized xtal.
      http://www.romanblack.com/xoven.htm

    • As you said Dave, the frequency counter part is a bit unimpressive, but when you got to the TCXO, with the shiny copper box, it was like being in front of a treasure chest. I thought to myself “gogogo, pop that open, gogogo”, and you did it. Thank you Dave for that great EEporn moment.

    • Hey there’s a 555 timer in there! At 8:40, right in the middle of the image. Awesome.

    • Crystal holder looks like an HC-6/U type.

      (That WW II era designation stands for holder, crystal, model 6, universal application I believe.)

      There are some subtle options to HC-6/U holders for long term stability.

      The cheapest versions solder the base and can together with soft solder. The base insulator is phenolic type material.

      An improved version cold welds the base and can with the base pin insulators being glass beads.

      The highest quality versions use a fused quartz “can,” anda vacuum inside the can.

      Most crystals age down in frequency as impurities inside the enclosure outgas and add small amounts of mass to the quartz plate. Another source of contamination is ingress through the can to base seal and around the pin seals. Hence the effort to improve the hermetic seal.

      Operating at an elevated temperature also increases outgassing.

      • having a vacuum in the can would not be a good idea in an ovenized crystl. vacuum does not transfer heat … so your ‘keep the temperature constant’ woud do zilch …

    • Thanks for the tear down which I found interesting. I have used various including that particular model (without the ovened crystal). The last time I used a counter was for checking the number of NAND flash accesses at boot up of a handset. I almost ran out of digits on the counter. Talk about software bloat.

    • RyanFarrelly1985

      IRELAND!! OMG we made real stuff before?lol

    • Nice one again, Dave! And I’d like to add, Teardown Tuesday is a lovely idea! Hope this will never cost any device’s life though.

    • Teardown tuesday is great. Keep ‘em coming!

      Greetings from WESTERN Germany ;-)

    • yes, please keep those teardowns coming, you know as well we love it at least as much as you do Dave!

    • I love the teardowns! The ones with older gear are always interesting to see the ingenuity applied to parts and circuits required to achieve the precision required. The boxed oscillator is really precious!

      That gives me lots of ideas to use in my own circuits…

      Not only Germany and Ireland, but also Portugal is there in a 74LS74 at 7:04… I still have some of these.

    • Dave,

      I used the MC10116 and 95H90′s in 1973/74
      so that should give you an idea of the vintage.

      When I was a EE intern in 73/74 the MECL was a hot ticket!

      Sam
      W3OHM

    Leave a reply