Connect with us

pin drivers

Discussion in 'Electronic Design' started by Jonathan Kirwan, Aug 8, 2003.

Scroll to continue with content
  1. I'm suddenly interested in pin drivers. Not for production work
    or for a commercial product. But for hobbyist exploration.

    Here are the specs as I believe I'd like them:

    High impedance current source/sink:
    +/-35V compliance range
    10mA full scale, 100uA resolution, 10uA worst case offset
    1mA full scale, 10uA resolution, 1uA worst case offset
    100uA full scale, 1uA resolution, 100nA worst case offset
    10uA full scale, 100nA resolution, 10nA worst case offset
    1uA full scale, 10nA resolution, 2nA worst case offset
    (The last scale is a "wish for")

    Low impedance voltage (+ or -):
    35V full scale, 20mV resolution, 5mV worst case offset
    dissipation limited to 15mA maximum current would be fine

    Ambient Temperature range: 5-45 Celsius

    I think I could get away with using the low impedance voltage
    settings when I want to "ground a pin."

    I'd plan to use the lowest feasible current source sink scale
    and set it to as close to 0nA as I could, in order to implement
    the "high impedance off/disconnect" state of the pin driver.

    Control wiring should allow selection of the current range or
    the voltage mode for the pin, as well as a single voltage input
    used to specify either the voltage output or the current.

    I imagine implementing this as an "upper half" and a "lower
    half," if done discretely, each with their own separate
    controls. I'm open to other ways, though.

    Doesn't need to be fast. I can live with 100ms settling time
    when I change the settings.

    I've looked at the LT1970, for example. But I have at least one
    serious problem with it: it can stand off +/-18V, but not
    +/-35V and more. Also, cost is about $6.50 in 25's, about twice
    that in 1's, and less expensive is better because of the number
    of pins I may want to implement (32 or more.) And it's designed
    for rather higher current limits and it's offset and bias
    currents may be some trouble (I want 10mA and down from there.)

    I've done some googling for various ATE/pin drivers. There's
    the Linear LT1970, the Apex PA08, and Elantec EL devices, to
    name some. But nothing yet quite like what I'm looking for,
    without more complexity than what I've already come up with for
    less cost -- a discrete arrangement which I *think* might do
    what I want at reasonable part cost, using 16 BJTs (under $1 for
    all) and both sections of a cheap LM358 (about 20 cents.) But
    it seems someone must have needed something in the +30V
    compliance area and only rather small currents that may already
    be "out there."

    I'm only looking for around 1/2 watt or so dissipation for the
    pin driver and I thought there might be an "all in one" swiss
    knife in this power range that doesn't cost me more than $4 per
    pin at 32 pins. (I might be pressed into the $6.50 for the
    LT1970, but it's voltage range is too limited and I'd still need
    to surround it with more parts.)

    Specs I should think about more closely? Discrete circuit ideas
    I might consider? Suggested IC parts to look at, for ideas?
    Questions about sanity?

    Jon
     

  2. When I was recently browsing in the news of Analog devices, i remember
    having seen some pin driver circuits. Don't remember the specs though.

    Rene
     
  3. Yes, Analog was one of the early places I looked through. I'm
    looking for $4 per pin, hopefully substantially less. The max
    of 1/2 watt per pin probably suggests something about packaging
    limitations for combining several into an IC, but I know very
    little about how many and what kind of packaging would make
    sense.

    I know that the nice metal can (TO-5, I think) 2N2222 can handle
    about 1/2 watt to air, and I could imagine a bunch of
    those-sized things -- but the number of pins per TO-5 would be a
    problem.

    I expected that there might be an IC out there which combined
    lots of pin drivers. If combined into an IC with support for
    serial-shifted commands, then the pin count would be much more
    manageable. The dissipation issue could be managed (in my case,
    anyway) because no more than a few pin drivers would actually be
    dissipating much. I'm thinking about testing systems where the
    total power consumption of the circuit in-test is under 1/4 watt
    (and usually *way* under) -- and even if most of the +/-35V
    (let's say all but 1V of the 70V of it) must dissipate away 69V
    for a 1V, 1/4 watt circuit-under-test, then we are talking about
    some 69V/4 or less than 16 watts combined. That's doable, I'd
    imagine.

    Of course, I don't expect anyone to have produced an IC just for
    my cares. But I was wondering if anyone had done something in
    the neighborhood. The LT1970, for example, is a 500mA kind of
    animal. Way too much for my needs and rather expensive, even in
    quantity. Plus, I'd need a lot of externals and I didn't even
    look enough at it to see if it could really manage the very low
    current end of things -- I'd already excluded it because of the
    limited voltage rail tolerance, not to mention the cost.

    I suppose the discrete approach is what I'll have to do. I'll
    just get four pins up and working on a proto and worry about the
    issues of calibrating and operating it and see if there are any
    gotchas I hadn't anticipated. I'm just a hobbyist, so I've
    probably missed some important details, too. Fix those, if I
    can, and then proceed. If that works out and I like the
    results, I can get an SMT board done up with the 32 pins. I can
    get arrays of PNPs and NPNs and resistors, so it will work out
    okay and won't cost all that much. But I was just hoping.

    Jon
     
  4. Jim Thompson

    Jim Thompson Guest

    It only had a few selections for PU/PD sources/sinks.

    ...Jim Thompson
    --
    | James E.Thompson, P.E. | mens |
    | Analog Innovations, Inc. | et |
    | Analog/Mixed-Signal ASIC's and Discrete Systems | manus |
    | Phoenix, Arizona Voice:(480)460-2350 | |
    | [email protected]_innovations.com Fax:(480)460-2142 | Brass Rat |
    | http://www.analog-innovations.com | 1962 |

    For proper E-mail replies SWAP "-" and "_"

    Why is it that Democrats can't debate politely?
    And are only rude and interruptive.
    Lack of mental capacity?
     
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

-