3-500Z amp drive me crazy! Guru needed

Discussion in 'Amateur Radio Amplifiers' started by VA2NA, Apr 26, 2019.

ad: L-HROutlet
ad: l-rl
ad: L-MFJ
ad: MessiPaoloni-1
ad: Left-2
ad: Subscribe
ad: Left-3
  1. AA5CT

    AA5CT Ham Member QRZ Page

    Another question, do you know the resonance points of your PA plate choke?

    PA plate chokes are not all the same, and can exhibit various resonances, including series resonance, at some number of frequencies often interspersed between operating ham bands.

    W8JI has done a good job of exploring and explaining this:

    https://www.w8ji.com/rf_plate_choke.htm
     
  2. AA5CT

    AA5CT Ham Member QRZ Page

    I did some characterization of an SB220 amplifier some years back, looking for odd resonances in the PA output stage ... here is some material you might peruse just for ideas' sake:

    https://sb220amplifier.wordpress.com/test-page-ading-a-new-page/

    https://sb220amplifier.wordpress.com/
     
  3. KD2ACO

    KD2ACO Platinum Subscriber Platinum Subscriber Life Member QRZ Page

    Up at the frequencies of interest, layout is going to be key.
    I'd consider 3 separate grid grounding tabs per tube, one for each pin.
    I'd consider keeping the filament feed wires twisted from the choke, feed the filaments in parallel (sorry about that filament transformer already there... what voltage is it? Are you feeding the filaments in series now? Can't quite see under there), and include a cap across the choke output as it feeds the first tube.
     
  4. W1QJ

    W1QJ Ham Member QRZ Page

    Tom W8JI in his original design of the Al-82 amplifier DID NOT use a single point ground per grid pin to ground. He used a continuous moderately wide strap and circumvented the tube socket tying all the grids together as he hit various ground spots around all 4 corners. The intention for this was to obviously to get away from the single point ground per tube pin. As time went on Ameritron developed their own tube socket which is now used in all their 3-500 based amps. The newest design is what I call a "ground plane" type socket where the entire bottom of the PC type socket is is grounded. Obviously this type of grid grounding is about as low in inductance as you will probably ever get. Short of using this socket I always use the the old method Tom used of running all the grids together and grounding at multiple points along the way. If I was building new from scratch I would use the new Ameritron "ground plane" type socket. The sockets are available in their MFJ catalog.
     
    WA7PRC likes this.
  5. WB2WIK

    WB2WIK Platinum Subscriber Platinum Subscriber QRZ Page

    Just glancing at the photos:

    • The plate RFC looks to have too many turns for 50 MHz; how many are required varies but that choke has more turns than any plate RFC I've used for 6m amps
    • It appears the bottom end of the plate RFC isn't bypassed at all where it should be, which is right at the bottom of the choke
    • It appears B+ is fed from a feedthrough on the "other side" of the tubes by a long red wire that runs between the tubes to feed the bottom end of the plate RFC -- lousy layout, asking for trouble
    • It appears the only B+ bypass for the plate RFC to work with are far way, at the other end of that long red wire where the feedthrough is -- lousy layout, asking for trouble
     
    WA7PRC likes this.
  6. K2XT

    K2XT Premium Subscriber QRZ Page

    Are the parasitic choke resistors 100 ohms? Aren't they usually more like 2-3 100 ohm resistors in parallel?
    And have you played with the coil dimensions in the parasitic choke?
     
  7. VA2NA

    VA2NA Ham Member QRZ Page

    I have made some progress yesterday,

    Using the VNA i can see a real big dip in the VHF band, around 140 Mhz

    For the plate choke i was thinking the same ( a part of this amp have been done by another one) and i have disconnectec the plate choke just to see if the plate choke is part of the VHF oscillation....not.. it change anything, the DIP is still present and very strong.

    There is a bypass capacitor very close to the coil but not seen in the photo, i am agree to relocate the feed trought close to the plate choke and not riding across the tube.
     
  8. VA2NA

    VA2NA Ham Member QRZ Page

    I have try 2 different parasitic and the last one using 1 100 ohm was just a ''test'' as soon as i have apply bias, the 2 tubes going in oscillation and blow the carbon resistor because the winding on the parasitic appear as a resistance and the 100 ohms was the best path.
     
  9. VA2NA

    VA2NA Ham Member QRZ Page

    The SK-410 socket used here was originaly in this HF amplifier and was working without oscillation, but it is true that the grid have long interconnection.

    I will add more picture this morning to show you how it is done.

    I have add 2 more grounding strap on each grid,

    Using the VNA, i am under impression that the RG-400 coax ground point is affecting the oscillation a lot ( RG-400 that is feed the cathode across a .01 on the filament and shield is fixed on the chassis...not very far away from the grid connection... mutual couplage???
     
  10. VA2NA

    VA2NA Ham Member QRZ Page

Share This Page