Connect with us

Tracking Generator

Discussion in 'Electronic Basics' started by Paul Burridge, Nov 27, 2005.

Scroll to continue with content
  1. AFAIK (which is not a lot on this subject) it's a sig gen in sweep
    mode that's synchronized to its complimentary receiver.
  2. sck0006

    sck0006 Guest

    Could someone please straighten me out. What is a tracking generator?
    I know they go with spectrum analyzers, and I have always just assumed
    they were a controlled swept oscillator or something like that. What
    is their primary function? The reason I ask is I have to run through
    an adjustment procedure on a spec an, and I need to tune a bandpass
    filter, and it calls out for a tracking generator. We unfortunantely
    don't have one, so I'm trying to find out what they actually do and
    can't find much on google. Is something that a sig generator in sweep
    mode could accomplish (i.e. an 8660...) I'm sorry for the vagueness
    of my question, and will give more info if necessary.

    Thanks for your time, and as always, all replies greatly appreciated.
  3. A tracking generator is nothing more than an RF signal generator
    with a swept output frequency. Said sweep is usually slaved to, and
    synchronized with, whatever spectrum analyzer is driving it.
    Tracking generators are indeed used for, among other things, the
    tuning of bandpass filters. The idea is that you feed the signal from
    the generator into one side of the filter, then use the SpecAn to
    monitor the output. You then tune the filter for the smoothest 'hump' on
    the spectrum display.

    I've done this plenty of times with older Motorola radios. Their
    VHF Syntor X series, for example, bottoms out at 150MHz. This won't work
    directly on the amateur 2-meter band, so the filter gets retuned, using
    the exact equipment you're seeing called out, so that its passband is
    changed to cover down to 144MHz on the low end.
    If you don't have one, you're not going to be able to tune that
    filter accurately. You -might- be able to fake it with a generic sweep
    generator and a spectrum analyzer, but it'll be tricky because you need
    to EXACTLY match the sweep rate of the generator with that of the
    spectrum analyzer. If they're not matched, you'll never get an accurate

    If you can tell me which SpecAn you're working with, I may be able
    to suggest a matching generator.

    Happy tweaking.

    Dr. Anton T. Squeegee, Director, Dutch Surrealist Plumbing Institute.
    (Known to some as Bruce Lane, ARS KC7GR,
    kyrrin (a/t) bluefeathertech[d=o=t]calm --
    "If Salvador Dali had owned a computer, would it have been equipped
    with surreal ports?"
  4. sck0006

    sck0006 Guest

    Actually, that's what I've been doing so far. Since we don't have a
    tracking generator, I've just been using our specan, (hp 8568a) with
    our 8660b in fast auto sweep mode, and it just so happens to sync up
    with the spec an w/ a sweep time of 100mS (very tricky indeed), and I
    can only use this for ballpark adjustments. To tune it in I have to
    put it in single slow sweep mode, and put the specan in max hold and
    wait for the sweep to finish (takes forever each cycle). Now,
    however, our spec an failed its performance test, specifically the
    resolution bw switching uncertainty test.

    The table is:
    all measurements in dB, tol at 3Mhz and 10Hz +-1dB, all else +-.5dB
    3Mhz .07 in
    300kHz -.28 in
    100kHz -.92 out
    30kHz .69 out
    10kHz .41 in
    3kHz .32 in
    1kHz 1.84 out
    300Hz -.13 in
    100Hz -.37 in
    30Hz -1.08 out
    10Hz -7.70 way out

    This is where I stopped the performance tests, I know I should go
    through them all before I go adjusting, but it is a loooong procedure,
    and if I just have to redo it all again....
    Hoping that it is just adjustment, not repair needed,...
    P/ the S.M., the related adjustments for this procedure are
    3MHz b.w. filter
    21.4 B.W filter
    Down/Up converter

    The 3MHz filter adjustments call out for the tracking generator. To
    set up the equipment, verbatim from the manual:

    8. Disconnect 97 (white/violet) cable from A4A8J1 and connect
    tradcking generator to A4A8J1 using BNC to SMB snap-on cable.
    9. Set spectrum analyzer SCAN WIDTH PER DIVISION to 20kHz and red
    scan width knob to ZERO. Set TUNING STABILIZER switch to ON. Set
    tracking generator output level to -25dBm and tune spectrum analyzer
    FREQUENCY for a tracking generator output frequency of 21.4000MHZ.

    There is only a single cable from the tracking generator going into
    a4a8j1, and there aren't any sync lines between the specan/tracking
    generator and the 8568a.
    **So is this just a center frequency of 21.4MHz w/ a sweep width of
    200kHz, or is it CW at 21.4MHz?

    The procedure starts with filter peak adjust, and to connect scope
    chA to tp7, chB to tp5, adjust tracking generator output frequency to
    peak ch a display, then adjust a cap for maximum peak to peak level on
    chB. This is then repeated at different test points, and adjusting
    different caps,.. and finally using the spec an crt to achieve final
    Then for center and symmetry. Still nothing about syncing the
    tracking generator to the 8568a. It says to bypass, one by one, the
    xtal filters and adjust for minimum peaks and symmetry.
    This all makes sense, but there isn't any adjusment of sweep time, or
    syncing the two spec an's, or anything like that.

    The other tests only require DVM, step attenuator, power meter, etc..

    So, what would you make of this? Is the tracking generator inputting
    a swept frequency to the 8568, or is it CW? Think it's possible to
    use the 8660b for this?

    Also, the manual calls out for an HP 141T/8552B/8553B/8443A.

    If I need to clairify further, please let me know. Sorry if this is
    confusing, and thanks for the quick response.
    Thanks, Steve
  5. Jamie

    Jamie Guest

    that is a programmable signal sweep source needed to inject into your
    filter so that the monitor can view the status of the filter.
    service monitors (most any ways) have this option in places where you
    may need to align duplexers and such etc...
    my service monitor is getting old now even though it's in mint shape
    (IFR type). it only covers up to 1 Ghz in sweep.
    maybe i will sell it for a newer unit! who knows.
  6. Gotcha. You're in luck. I believe the HP 8444A-OPT 059 tracking
    generator is compatible with that analyzer. What happens is that the
    SpecAn outputs a swept IF freq that is in sync with its sweep rate. Said
    swept output is plugged into the tracking generator, and the output of
    said generator then goes to the filter. Vyola! Perfect match.

    You shouldn't have much trouble finding the 8444A (they turn up on
    Greed-bay periodically), but be absolutely POSITIVE it's the OPT 059
    model. This option adds a self-contained third local oscillator to the
    tracking gen, and makes it truly usable with many anlayzers it might not
    otherwise be able to work with.

    Happy hunting.

    Dr. Anton T. Squeegee, Director, Dutch Surrealist Plumbing Institute.
    (Known to some as Bruce Lane, ARS KC7GR,
    kyrrin (a/t) bluefeathertech[d=o=t]calm --
    "If Salvador Dali had owned a computer, would it have been equipped
    with surreal ports?"
  7. Ben Jackson

    Ben Jackson Guest

    I'm confused. If you have a spectrum analyzer, couldn't you just use
    broadband noise as the filter input and look at the resulting spectrum
    to see the characteristics of the filter?
  8. sck0006

    sck0006 Guest

    Thanks very much for your response, I'll see if my boss wants to buy
    one (hopefully).

    Thanks again,
  9. sck0006

    sck0006 Guest

    I'm confused too. For our purpose, we need signals in the range of
    0-10dBm. We use the swept signal generator to tune in transponder
    taps. You have to use an antenna to transmit into the tap, and the
    tap attenuates the signal by 25dB, so after antenna loss and the
    attenuator loss, the signal is significantly less. Having a tracking
    generator for this would be much more agreeable than using the
    ballpark method.

    Not sure about using the broadband noise for tuning the filter...

  10. Ben Jackson

    Ben Jackson Guest

    Oh, you should ignore me, I wasn't responding to your original query,
    just just hoping someone could explain why the tracking generator was
    necessary when the output was being interpreted by a spectrum analyzer
    (vs a scope).
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