Kenwood TL-922A Modifications

Discussion in 'Amateur Radio Amplifiers' started by W4LAC, Jun 5, 2019.

ad: L-HROutlet
ad: l-rl
ad: Left-2
ad: MessiPaoloni-1
ad: L-MFJ
ad: K5AB-Elect-1
ad: Subscribe
  1. W4LAC

    W4LAC XML Subscriber QRZ Page

    Looking for opinions...

    1. Parasitic suppressors: Do you prefer the original Kenwood suppressors, or the newer U-shaped “hairpin” suppressors?

    2. Direct grounding of the grids to the chassis, or leave as Kenwood designed in the amplifier with the RFC’s and caps?

    3. Additional electrical bonding between the tuner chassis and the tube chassis, or sufficient as-is? If additional recommended, what would be your preferred way to accomplish it?

    73's!
     
  2. W1QJ

    W1QJ Ham Member QRZ Page

    I have reworked numerous 922 amps. I find the original parasitic suppressors to be just fine. I would leave them. I would ground the grids and perform the "self bias" mod to protect the filament transformer. Add a glitch resistor. A 1n5408 diode on the minus rail is also good.
     
    VK6APZ likes this.
  3. W4LAC

    W4LAC XML Subscriber QRZ Page

    What's this "self-bias" mod? Is there a drawing/document showing this glitch resistor and diode and where they go? Also, I'm installing the Kessler Engineering refresh kit. Does the Kessler kit negate either of your suggestions?
     
  4. WB2WIK

    WB2WIK Platinum Subscriber Platinum Subscriber QRZ Page

  5. W1QJ

    W1QJ Ham Member QRZ Page

    Last I knew the Kessler kit contained a QSK mod as well. Some of his stuff is questionable and I would stay away from. Last I knew his kit was $400.00 If that is the case I would suggest the W7RY QSK mod instead and just do the simple mods I suggest. Many salespeople create problems that do not exist so they can sell solutions to non problems. Be wary.
     
    WQ4G and WA7PRC like this.
  6. N8FVJ

    N8FVJ Ham Member QRZ Page

    The hairpin suppressors were determined not to be as affective as a resistor with turns of wire around the resistor body.
     
  7. WA7PRC

    WA7PRC Ham Member QRZ Page

    The "self bias" modification reconfigures the bias switching so that, in the event of a tube grid-cathode short, the cutoff bias source is not shorted. This prevents the 80V bias winding on T2 from being blown open. Jim @W7RY had me add it to the TL-922 schematic that I redrew.

    I'm with Lou @W1QJ on the Kessler stuff. $400 seems kinda pricey. The W7RY scheme is a much better price, and works well.
     
    WG7X likes this.
  8. G0HZU

    G0HZU QRZ Member

    A crude model of a GG amplifier based on the datasheet parameters does suggest that the use of 200-220pF caps at the grid pins will protect against negative resistance over a more useful bandwidth than a typical attempt to ground the grids directly with copper tape. So there is method in the madness here even though many people seem to prefer to ground the grids directly.

    The suppressors are there to offset any net negative resistance (at the anode) generated by the tube(s). They achieve this by adding more series resistance than there is 'negative' resistance in the tube. However, adding just a resistor would be no good because the resistor would overheat badly from the high power signals on 10m and 15m. So the compromise is to increase the resistance to maybe 47R and put an inductor across the resistor. The inductor looks like a low reactance clamp across the resistor at 28MHz and it shows a high reactance up at VHF. So the inductor ensures that the net series resistance at 28MHz will be low (to keep the resistor cool) but it allows the net series resistance to be high enough at VHF to offset the negative resistance in the tube.

    Floating the grids with the right value caps (even with the added inductance of the cap connection wires) can in theory at least do a better job of minimising the negative resistance in the tube and it can do it over a wide bandwidth. But the risk here is component ageing (and failure?) and also there will be variation in the component lead lengths during initial assembly. So this method might not perform as consistently as it could/should.
     
    Last edited: Jun 7, 2019

Share This Page