Connect with us

Increment/Decrement Control Loops

Discussion in 'CAD' started by john jardine, Oct 31, 2004.

Scroll to continue with content
  1. john jardine

    john jardine Guest

    Sounds pretty much like the old 'huff-and-puff' VFO stabiliser the radio
    hams used.
  2. Jim Thompson

    Jim Thompson Guest

    Anyone know the history of Increment/Decrement Control loops?

    There's a paragraph on page 56 of the October "Microwaves and RF"
    magazine about AGCing a VCO using an up/down counter and D-A in an
    increment/decrement control... touting it as if it's the next best
    thing to sliced bread.

    But I hardly think it's novel... I was setting the timing of
    automobile ignition systems in 1968, but used a squirt of charge
    into/out-of a capacitor to increment/decrement.

    The paragraph references an IEEE article in October's "IEEE
    Transactions on Instrumentation and Measurement", which I can't access
    since I object to $35/month for the privilege (as I noted today as I
    renewed my memberships).

    Maybe someone who subscribes to that service can look it up?


    ...Jim Thompson
  3. Jim Thompson

    Jim Thompson Guest

    That's what I thought, too.

    ...Jim Thompson
  4. Try looking for "single-speed floating control". I think it's a really
    old and not-so-great control strategy.

    Best regards,
    Spehro Pefhany
  5. Max Hauser

    Max Hauser Guest

    "Jim Thompson" in news:...
    I haven't seen the original, but from the description here, it sounds like a
    good old digital integrator. In case anyone who is reading this has not run
    into or thought about it, a digital up-down counter is precisely a DSP
    integrator for a one-bit input ("+- 1"). (At times in the past we made very
    good use of that property, and other properties of one-bit arithmetic, in
    digital filters for oversampling data converters.) And an integrator is the
    heart of many control loops.

    In the early days of digital forays into the system-simulation world
    pioneered by analog computers (more on those in classic books by Jackson,
    and by Korn and Korn), people built special-purpose digital integrators, and
    then made loops out of them that would solve differential equations over
    time, just like electronic analog computers did. They were called Digital
    Differential Analyzers (DDAs). A single digital integrator in an AGC
    control loop would be a grandchild of those.

    -- Max
  6. Jim Thompson

    Jim Thompson Guest

    I thought it worked rather well. My scheme looked at dwell, decided
    if it was early or late and moved a threshold accordingly... this
    happened _after_ firing so was not subject to noise from the spark.

    Just like having an elf sitting there turning a knob just a small
    amount after each firing ;-)

    ...Jim Thompson
  7. I think it generally works better if the elf moves the knob more for
    large differences and less for small differences.

    Best regards,
    Spehro Pefhany
  8. Jim Thompson

    Jim Thompson Guest

    If you choose a fixed increment that is small enough it closes at
    "left-right-left-right-left......" ;-)

    Driving a big-assed T-bird (400+CID) in 1968, up and down 101 (near
    Sunnyvale) at over 100MPH, I was never able to cause the ignition to
    stumble... scared the hell out of my boss, though ;-)

    ...Jim Thompson
  9. Jim Thompson

    Jim Thompson Guest

    Perhaps I should elaborate. Waveform form (simplified, it's more
    rounded) from magnetic pickup in distributor:

    _ _ _
    / | / | / |
    | / | /
    | / | /
    ----------------------- THRESHOLD
    | / | /
    |_/ |_/

    Downstroke is firing. When ramp crosses threshold power device
    (driving coil) turns on. So this crossing determines "dwell".

    Device control is simple-minded (my preference :)...

    Device has three states:

    Off... Fired
    Saturated, coil charging
    Current mode, regulated at 5.5Amp

    So I only look at two conditions... when firing occurred was I in
    saturation or in current regulation.

    If device was in saturation, dwell didn't start soon enough so I
    decrement the threshold. If device was in current regulation dwell
    started too soon so I increment the threshold.

    Worked like a champ.

    ...Jim Thompson
  10. Tom Seim

    Tom Seim Guest

    I will send you a copy if that is your real email address.
  11. Jim Thompson

    Jim Thompson Guest

    I have already been sent a copy, Thanks! (The header E-mail address
    is false... read the SIG to obtain a valid E-mail address.)

    ...Jim Thompson
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