Connect with us

0.01% resistors

Discussion in 'Electronic Design' started by Raveninghorde, Jul 30, 2013.

Scroll to continue with content
  1. What are people paying for 0.01% resistors?

    I wanted to reduce production tweaking so thought a few 0.01% 0603 or
    0805 resistors would do the trick. Most suppliers have parts in the
    many dollars range. Digikey do have a few Stackpole resistor values
    for £0.57, about 85 cents.
     
  2. Bill Sloman

    Bill Sloman Guest

    Farnell has got a few parts at a slightly higher prices - 10k, 1k, 100R.
     
  3. RobertMacy

    RobertMacy Guest

    Once, long ago, someone decided to use 0.01% resistors. We found the
    rpcing in US dollars at $1 to $1.50
    Delivery data, now this was the 'interesting' part, 6 to 9 months, but
    NEVER arrived !!!!

    It was cheaper to simply put in excessive parts and let a laser
    automatically cut unwanted ones out. That was assembly was constant, and
    the tester program did the 'adjustment'

    In retrospect, I'd push for the 'design-it-out' principle. Go back to
    Engineering and beat them about the head and shoulders until they had
    designed out any problem like that in Manufacturing. ;)
     
  4. The big dollars are not for close tolerance resistors but for
    resistors that earn that close tolerance number by having commensurate
    low tempco. Without that you're going to get noise and nasty
    transients as temperatures change, at best, even if you calibrate out
    the temperature drift.


    Best regards,
    Spehro Pefhany
     
  5. Bill Sloman

    Bill Sloman Guest

    Farnell stocks a few precision resistor arrays, so your first thought is correct. I bought a couple for my low distortion oscillator last year.
     
  6. These are pretty impressive, unique parts as far as I know.

    <http://www.linear.com/product/LT5400>

    0.2ppm/K ratio matching.
     
  7. I have the same requirement (let me know if you find anything!)

    There seem to be Vishay "bulk metal foil" types north of $10.

    Farnell have a few cheap "0.01%" resistors of unknown origin and
    specification, very peculiar.

    0603 size seem thin on the ground, they seem to be 0805 or above.

    Rhopoint seem to specialise in precision resisistors. There are some
    manufacturers of lower cost (than Vishay) bulk metal foil precision
    resistors, as well as other types.

    <http://www.rhopointcomponents.com/components/resistors.html>

    It is very tempting here to use Larkins susumu types. Problem is if you
    take the worst case datasheet limits on anything bar the metal foils,
    they always seem incredibly pessimistic. But it seems ridiculous to put
    say two $10 chip resistors on a board otherwise full of sub-$0.001 parts
    and a $4 120MHz ARM.
     
  8. Unfortunately, the ones that are 50x better cost about 50 times more.



    Best regards,
    Spehro Pefhany
     
  9. They are out there.. matching "typical" +/-0.1ppm/°C, but not cheap
    (~$20 pair).

    It's strange when the resistor part of a BOM cost dwarfs that of the
    semis.


    Best regards,
    Spehro Pefhany
     
  10. It's not hard to have $1,000 worth of resistors on a board with semis
    worth 5-10% of that.


    Best regards,
    Spehro Pefhany
     
  11. Thanks, that's interesting, and they're relatively inexpensive.

    One of the dangers of using parametric search is that I'd missed those
    due to their rather poor absolute tempco (25ppm/K), but that tempco
    (typical 8ppm) is matched typically to within +/-2.5%.

    Not quite as good as the Z-foil types (0.1ppm/K typical matching with
    1:1 values), but typically better for ratios other than 1:1.


    Best regards,
    Spehro Pefhany
     
  12. Simplifies the buyers job. Forget everything else, just shop around and
    negotiate for a couple of resistor types.

    I have recently (belatedly) realised that is already the case for ICs.
     
  13. I wanted to make precision op-amp gain stages (single ended). The
    LT5400s come in a handful of value combinations, so I got to modify my
    "parallel-series resistor combination" program to work with the possible
    networks of 4 resistors.

    What combination of 4 resistors come closest to a specified ratio?

    I sketched them and got 19 combinations (plus their reciprocals)


    "a+b",
    "a + (b+c)",
    "(a+b) + c",
    "a+ (b+c+d)",
    "(a+b) + (c+d)",
    "(a+b+c) + (d)",
    "(a|b) + c",
    "a + (b|c)",
    "a + (b + (c|d))",
    "(a+b) + (c|d)",
    "a + ((b|c)+d)",
    "(a+(b|c)) + d",
    "(a|b) + (c+d)",
    "((a|b)+c) + d",
    "((a|b) + (c|d))",
    "(a) + (b|(c+d))",
    "(b|(a+c))+d",
    "a + (b | (c+d)) FB from c-d",
    "(b | (a+c)) + d FB from a-c"


    "|" means "in parallel with"
    "+" means "in series with"

    Then a dumb C program to try them all.

    Later extended to 2 and 3 gain stages.
     
  14. Phil Hobbs

    Phil Hobbs Guest

    If you use a common-centroid layout, you can get pretty good ratio
    stability with garden variety ones.

    Cheers

    Phil Hobbs

    --
    Dr Philip C D Hobbs
    Principal Consultant
    ElectroOptical Innovations LLC
    Optics, Electro-optics, Photonics, Analog Electronics

    160 North State Road #203
    Briarcliff Manor NY 10510

    hobbs at electrooptical dot net
    http://electrooptical.net
     
  15. Pease on this stuff:-
    http://www.datasheetarchive.com/files/national/htm/nsc03933.htm

    Irritating that it's so difficult to even come close to the tempco of
    a well-designed and well-constructed rack mount ratio transformer.
    (< 1ppb/°C typically)
     
  16. Bill Sloman

    Bill Sloman Guest

    It would be more irritating if there were more applications where you couldexploit this. I've known about ratio transformers for about forty years now, but have yet to find an application where I could use one.

    And they don't have to be rack-mounted to be that good. A ferrite core wound with Litz wire could be hooked up as a ratio transformer, though making the connections would be both fiddly and labour-intensive. I've long cherished the idea of winding one with round-to-flat cable. You couldn't put on all that many turns, but you wouldn't have to sort out the wires.
     
  17. I think that misses at least

    r+(r|(r+r))

    to use my "notation".
    Yes that is what I needed, otherwise you don't get to take advantage of
    the ratio matching of an integrated network. (Or of resistors from the
    same reel possibly?).

    Also some of the LT5400 parts have two values in the package, so you
    need to actually get the formulae so you can plug in the values.
     
  18. Phil Hobbs

    Phil Hobbs Guest

    Well, they're used all the time in capacitive gauges.

    Cheers

    Phil Hobbs

    --
    Dr Philip C D Hobbs
    Principal Consultant
    ElectroOptical Innovations LLC
    Optics, Electro-optics, Photonics, Analog Electronics

    160 North State Road #203
    Briarcliff Manor NY 10510

    hobbs at electrooptical dot net
    http://electrooptical.net
     
  19. No no, in my sketch the bottom resistor is moved over underneath the
    right hand one!

    Oh all right, sorry. I plead a cold coming on.
     
  20. My question is why? Even brewing beer does not take that kind of precision..
     
Ask a Question
Want to reply to this thread or ask your own question?
You'll need to choose a username for the site, which only take a couple of moments (here). After that, you can post your question and our members will help you out.
Electronics Point Logo
Continue to site
Quote of the day

-