Connect with us

Is there a device that...?

Discussion in 'Electronic Components' started by Mark Folsom, Dec 9, 2003.

  1. Mark Folsom

    Mark Folsom Guest

    Is there a (preferably tiny surface mount) device that starts out
    non-conducting and then conducts a short time after a voltage is applied,
    like maybe 100 microseconds? If so, what's it called?

    Mark Folsom
     
  2. Steven Swift

    Steven Swift Guest

    capacitor.

    but I think you meant something more detailed.
     
  3. Boris Mohar

    Boris Mohar Guest

    What is it that you do at Novatech?
     
  4. Mark Folsom

    Mark Folsom Guest

    A capacitor conducts when voltage is applied, then stops. That's the
    opposite of what I asked for.

    Mark Folsom
     
  5. Dave Platt

    Dave Platt Guest

    A capacitor conducts when voltage is applied, then stops. That's the
    opposite of what I asked for.[/QUOTE]

    Correct.

    "Inductor" is one possible answer, which might work if you don't need
    a particularly "sharp" turn-on, and if you're simply concerned about
    passing DC.

    Another approach would be to use a FET, with a suitable gate voltage
    being applied through an R/C circuit of some sort... or perhaps using
    a bidirectional FET-type optocoupler.

    Without knowing more about your application and needs, it's going to
    be difficult for anyone to give you an answer precise enough to suit
    your needs.
     
  6. Mark Folsom

    Mark Folsom Guest

    Correct.

    "Inductor" is one possible answer, which might work if you don't need
    a particularly "sharp" turn-on, and if you're simply concerned about
    passing DC.[/QUOTE]

    For the first 100 to 1000ms of operation, the powered circuit draws about 5
    microamps. The initial voltage, from which we would like to protect the
    powered circuit, could be from 50 to 60 volts. After about 50 microseconds,
    drawing off a tiny bit of charge will bring the voltage down to half of its
    initial value, which is acceptable to the powered circuit. Since we need
    this to be tiny, an inductor is unsuitable.
    What I envisioned, if only we had room, is an RC circuit connected through a
    zener to the gate of an SCR. The capacitor would charge until it reached
    the threshold of the zener and then switch the SCR on--and it would draw off
    the bit of charge needed to pull the voltage down. A small surface mount
    chip with this set of components or an operational equivalent would help a
    lot.

    Mark Folsom
     
  7. Steven Swift

    Steven Swift Guest

    Whoops, in my attempt at "wit" I missed the point. Sorry.

    But, all wit aside, you can buy "reset generators" that take one cap to
    set the delay time. A SOT-23+cap=100uS delay.
     
  8. mike

    mike Guest

    7 posts in the thread and it's still not at all clear what you want.
    Way too many "about", 10:1 time interval range, tiny bit of charge...

    Just clamp the voltage with a zener.
    mike

    --
    Bunch of stuff For Sale and Wanted at the link below.
    Toshiba & Compaq LiIon Batteries, Test Equipment
    Honda CB-125S
    TEK Sampling Sweep Plugin and RM564
    Tek 2465 $800, ham radio, 30pS pulser
    Tektronix Concept Books, spot welding head...
    http://www.geocities.com/SiliconValley/Monitor/4710/
     
  9. Well, inductor would be closer, but the OP is probably looking for
    some kind of avalanche device. OTOH, 100 uS is a pretty long time, in
    the grand scheme of things, so maybe some kind of time delay circuit.

    It _would_ be helpful to know what the OP had in mind for this part to
    _do_...
     
  10. Mark Folsom

    Mark Folsom Guest

    I think it's pretty clear what I want. If there's something specific you
    don't understand, maybe you could ask. Unlike most electronic applications,
    this one needs to avoid throwing away the energy that would be lost in
    clamping with a zener. I may seem stupid to you, but I already thought of
    clamping with a zener.

    If you think 7 posts is too much, you can stop reading.

    Thanks,

    Mark Folsom
     
  11. Steven Swift

    Steven Swift Guest

    Okay, I see what you need.

    I'd use a tiny mosfet with an RC on the gate, along with a diode to make
    sure the gate cap is reset upon power down. The bad news is that this won't
    work if your power comes up slowly. You'll have to design the RC time constant
    for your anticipated slew rate.

    I still like the idea of a micro-power reset generator, and then use it to
    drive a Mosfet. Maxim, et al make these things. You can also look for "hot
    switch controllers" which are important to make hot plug and play work. USB
    has these devices.

    Am I on the right track now?

    Sorry about the original flippant remark.

    Steve.
     
  12. Boris Mohar

    Boris Mohar Guest

    Mark,

    It might be clear to you what you want but it is not clear to us. Perhaps it
    would be more useful to also describe what you have in more detail.
    That bit about "throwing away energy" catches attention. You might have a good
    reason not to waste any, but sharing that with us might help in coming up with
    a practical solution. How much current do you want the device to switch and how
    much power is it allowed to dissipate doe to it s finite on resistance?
    sci.physics could not help yo because they invent this stuff. We use it. Unlike
    sci.physics, playing ten questions is out of the question. Spill your guts and
    you will get help. sci.electronics.design is another place to ask.
     
  13. mike

    mike Guest

    That's often the problem. Customer is SURE he knows what he wants.
    But that's usually not what he asks for. It's the old communication
    problem. Getting the requirements out is the hardest part of any
    project. I called the psychic hotline, but Miss Cleo didn't know either.

    If there's something specific you
    Nobody said you were stupid, just unspecified.
    Here's what you stated:

    If it's really just a tiny bit of charge, why is clamping such a
    problem? If you had your magic device, how would that tiny bit of
    charge be of benefit??? Stated another way, how are you gonna get the
    voltage down without drawing off that tiny bit of charge?
    I'm trying to be helpful. We're up to 10 posts and there's still
    something yet unstated that's driving your requirement.
    If you're upset with this, you would have been REALLY upset by
    my response to your original vague request. I chose not to send that one.

    You're welcome,
    mike


    --
    Bunch of stuff For Sale and Wanted at the link below.
    Toshiba & Compaq LiIon Batteries, Test Equipment
    Honda CB-125S
    TEK Sampling Sweep Plugin and RM564
    Tek 2465 $800, ham radio, 30pS pulser
    Tektronix Concept Books, spot welding head...
    http://www.geocities.com/SiliconValley/Monitor/4710/
     
  14. Mark Folsom

    Mark Folsom Guest

    I have two capacitors that I am charging in series, with a lot of noise, in
    about 50 microseconds--then discharging in parallel. I use some diodes to
    allow that to happen--I modeled it and then built it and it works and allows
    me to get a lot more energy than I do by charging them from the same source
    in parallel. The capacitors then power a custom timer ASIC that draws about
    5 microamps. At the end of a programable interval that ranges between 100
    and 1000 milliseconds, the circuit switches the remaining charge into a 2
    ohm resistance. Any device for this function would have to conduct about 8
    amps for a few hundred microseconds and have less than a quarter ohm
    resistance when in the on condition.

    The diodes that we use to switch from series to parallel seem to have enough
    capacitance to keep the output up to the charging voltage until a little
    charge is drawn off. It appears to take less than two microseconds to go
    from 52 volts to 26 volts when the capacitors are loaded with 12kohm on the
    output.

    The whole purpose of the circuit, up to the ASIC, is to gather some very
    scarce energy--clamping with a zener is our default approach, but it reduces
    our margins below where we would like them.

    Is that clear yet? I'm not going to disclose the whole system design or the
    end use.

    Mark Folsom
     
  15. Something like this:




    VCC
    .--------.
    | |
    | |
    | >|
    | |--------.
    .-. /| |
    | | | |
    | | | |
    '-' | |
    | | |
    o--------o |
    | + | |/
    ### '--------|
    --- |>
    | |
    | |
    | |
    | |
    | |
    | |
    | |
    === ===
    GND GND
    created by Andy´s ASCII-Circuit v1.24.140803 Beta www.tech-chat.de

    That's a circuit that short-cuts VCC to GND when the capasitors voltage
    reach about 0.7V.
     
  16. Dbowey

    Dbowey Guest

    RE:
    Hmmmmm! In your order:
    Yes.
    I think the end use is somewhat clear.

    You seem to have a chip on your shoulder. That's not nice when asking for
    help.

    Don
     
  17. Mark Folsom

    Mark Folsom Guest

    Well, I was told I should use a capacitor, that I should just clamp the
    voltage with a zener, that I should use an inductor, etc., etc. Meanwhile,
    I've given more and more detail on how the item would be used, and most of
    the comments I get back offer nothing whatever of use--either you haven't
    read what I've posted, or you don't understand what I'm trying to do even
    after you've read it. Maybe the vast majority of you don't have a clue and
    you're just jerking me around for your own amusement. Being nice to a bunch
    of ignorant jerks isn't likely to be terribly productive in this context.

    Mark Folsom
     

  18. I read all the postings closely, and it wasn't until the last one or two
    (the one that said you won't disclose more) that I began to have a clue what
    you were looking for. That might be because I'm an ignorant jerk. Or it
    might be because you're being less successful in your communications than
    you had thought. Your initial posting, for instance, gave no hint of
    whether you were wanting to switch currents of 100pA or 100kA.

    If you ask a question, and some people ask for clarification and others give
    answers that aren't what you expected, it is probably because your question
    was unclear. Sadly, your perception of how clear it was is not the gating
    factor.

    If I did indeed understand that recent posting, the bottom line is that you
    gather free energy from noise, charge some capacitors, and then eventually
    dump most of the juice into a load of about 2 ohms, after waiting a little
    while (where the timer is powered from a small fraction of the charge in the
    caps). You want to switch about 8A, with an on resistance of < 0.25 ohms,
    after a controllable delay. (Gee, sounds like Maxwell's demon.)

    I *think* that I understand that the power you want to control is the juice
    to the timer; that is, you want to keep the timer out of the picture until
    some of the charge has been bled off of the caps. Later posts seem to
    suggest that what you're interested in is the switch to the 2 ohm load,
    though, which seems to have thrown some folks off the scent. Perhaps I'm
    confused. Or, just an ignorant jerk.

    If I'm not confused, then I wonder whether perhaps powering the timer off
    just one of the caps, rather than off the series combination, might be a
    better approach. If not, then another consideration is that since you're
    talking very small currents, you might be able to just use a resistor and a
    MOSFET; that is, avoid the C, just use the gate charge of the MOSFET as a
    capacitor. Only two parts, there.
     
  19. mike

    mike Guest

    Ok, so I'm playing 20-questions, reverse engineering a secret design to
    help someone who's calling me names.
    I think that's called masochism.
    But I'm bored, and I really am a nice guy (most of the time), so here goes.

    Interesting little details the 8 amps and 2 ohms.
    Would have been nice to have that information in the beginning.

    Let's do a back of the envelope approximation.

    If your tiny bit of charge is the 5 uA load that takes your caps from
    60V to half that in 50 uS, you have about 8.3 pF of capacitance.
    When you discharge that into 2 ohms, you have a time constant of
    16.6 picoseconds.
    So, you need a deivce that can switch 8 amps and get that 16pS time
    constant to last a few hundred microseconds.

    I expect that's not what you meant. Can you see why some of us might
    be more than a little confused by a descrepancy that's 4 orders of
    magnitude?

    So, I guess I am an ignorant jerk that just can't understand what you're
    trying to do even after I've read it.

    If you're gonna hide the details, at least try to have some consistency
    in the numbers you do post. My subscription to the psychic hotline
    has expired.

    Your move,
    mike

    --
    Bunch of stuff For Sale and Wanted at the link below.
    Toshiba & Compaq LiIon Batteries, Test Equipment
    Honda CB-125S
    TEK Sampling Sweep Plugin and RM564
    Tek 2465 $800, ham radio, 30pS pulser
    Tektronix Concept Books, spot welding head...
    http://www.geocities.com/SiliconValley/Monitor/4710/
     
  20. Well, if you had given sufficient detail in the first place, you
    wouldn't be having this problem. It looks from here like you've been
    given a number of ideas which may work, but without a detailed
    knowledge of whatit is you are trying to do (which you've said you
    won't provide), many of us are just shooting in the dark. The more
    detailed your requirements, the more useful will be the [free]
    responses you get. In either case, you get what you pay for.

    [For instance, if you are looking at the output of a photodiode, maybe
    replacing it with a small solar panel would give you enough power to
    run your switching circuit...]
     
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

-