• About EEVblog

    Check Also

    EEVblog #1047 – Solar Roadways FINALLY BUSTED! Colas Wattway

    Dave finally puts an end to the idea of Solar Roadways. We have the test ...

    • Hey Dave, if you ever want to part with that Hybrid IC, I need one for a 3468!

      Maybe you have a franken-meter, repaired by using two 3457’s to make one that worked.

      Nice tear-down mate.


    • Yep, those are pesky HP part numbers. The 1820 prefix denotes digital ICs; the 1826 prefix denotes analog ICs; the 1810 prefix is for resistor networks; etc. The HP part-numbering system made it easier for HP components engineers to spec multiple vendors under one HP part number, so that you didn’t know whether TI, National, or Fairchild would be supplying your TTL part in production. If it had to be a particular vendor, you had to create a special HP part number for that. Makes troubleshooting at the component level harder, unless the parts are double marked (some are) or unless you have the part cross reference from the back of the manual.

    • Hey Dave, found this on All Programmable Planet.

      ‘All of this to say, I think it would be wonderful for us to have a PCB layout expert as a blogger here on All Programmable Planet — someone who could keep us all abreast of the latest developments in the circuit board world, and also answer any questions we may have when creating our next masterpiece.

      What do you think? Do you know anyone who might fit the bill? If so, it would be great if you could ask them to contact me at [email protected].’

      Sounds like an opportunity for Dave.


      • An opportunity to produce content to benefit someone else’s website. No thanks, I’d rather produce content for my own site.
        Just a heads up for fellow bloggers, or wannabe bloggers – I don’t know how much All Programmable Planet would pay for such a gig, and these comments are not directed at them, but to the industry in general – it is very common in the industry for people to expect you to do it for free in exchange for “the exposure”, or to work really cheap like for free gear or some such. Don’t do it unless you are being paid professional rates for it, otherwise it drags the whole industry down.
        Also, it is very common for companies to offer you highly inflated viewership figures to entice you to write for them, beware.

    • GregU

      At 36:20 in the video, we see the bottom of the switching hybrid. What is the hook-shaped trace on the left?

      • John Kolb

        I would guess that J shaped etch to be a guard ring. Note that it is near the corner pin which would be a good location for the input to be measured.

    • li

      Hello, The agilent 34410A have a fan is that bad for the accuracy of the meter?

      • Not necessarily. They would have done their homework to ensure that it’s not a problem. Ideally you don’t want a fan moving air with temperature gradients within a precision instrument. But if you have to have a fan for some reason, then you carefully design to avoid any issue due to it.

    • Pete

      Hi Dave,
      I found a very interesting PDF on HParchive.com which seems to contain the 1826-# (starts on pg.11) references as well as many more. I wonder if these match up with what you were seeing Dave?!



    • @Pete, here’s a list put together by another Aussie –


      Have found it useful as well when repairing HP equipment.


    • Excellent teardown on an excellent instrument! Please keep up the great work.

    • Hi Dave,

      I have had one of those for years. And recently I had to have it certified. Mine was last calibrated over 10 years ago and it was still in specification so you are dead on about stability.

      The 3455A’s and the 3456A’s LED 6 1/2 digit meters are also quite nice with higher input Z DC ranges up to 10 or 100V as compared to 10M for ranges abive the 3V range on the 3457A’s. The 55’s and 56’s go for cheap on ebay (unless you review them!) and they have outstanding stability. Both Jim Williams and Bob Pease confirmed their experience with me and said that those meters was top shelf. They take up a bit more space on the bench but in my humble opinion they are well worth it. The only disadvantage of the 55’s and 56’s is that they do not have current but I can always use my uCurrent for that!

      The 3Gohm range is handy on the 3457A and I use mine alot just for the 3GOhm and current ranges.

      One can never have too many DMM’s. My latest project is a Solartron (Schlumberger 7071/7081 7 1/2 (or is it 8 1/2))Digit DMM that needs some TLC. It uses a PWM method converter and supposed to be able to hold stay in cal for years. I guess I will have to check that once I fix the 0V fail message on self test!

      Now one of those would make an interesting tear down for you as it was designed to directly compete with the HP3458A.

      Maybe 3458A’s will be cheap someday. By the way, they use a more advanced version of the LT version of the LM399. Sometimes you can find the reference cards from the 3458A on ebay but they do not go cheap.

      Take cards and all the best and happy teardowns!

      Moderator – Independent LeCroy_Owners_Group on
      Yahoo! Groups

      P.S. All of the major, and some minor, test equipment manufacturers (pick one) use in-house part numbering schemes. Tek had a huge catalog with house numbered parts in it that in many cases were standard parts tested and selected for given parameters. Even RCA and Zenith did that for TV’s and radios.

      We did the exact same thing when I was at VDO (an automotive instrument cluster company)in the 90’s and also a couple of decades before when I worked at Pulsecom a telecomm company. We were famous for having our custom capacitors that were labeled 125V parts when they were actually 250 or 450V parts.

      Useful Info:

      For HP instruments add 60 to the first two digits of the serial number and that will give you the year of manufacture.

    • Analogue Man

      Hi Dave, the voltage reference was probably the LTZ1000.


    • Serge Lamothe

      Date codes for older instruments is basically the first two numbers plus 61. That will give you the year of fabrication.
      The last two digits are the week.

      so 2703Axxxx means made in 61+27= 1988 week 3.


    • zlymex

      Very nice teardown.
      However, one thing very important to be worth mentioning is the internal resistance standard if that is not equally important as the voltage standard. I recon the hermetically sealed 40k Vishay resistor costs much more than LM399. Even metrology grade 3458A shares the same resistor.

    The EEVblog Store generally ships twice a week, on Tuesdays & Fridays, Sydney time. Dismiss