ad: elecraft

600 watt HF Linear Amplifier Project

Discussion in 'Homebrew and Kit Projects' started by KD8UYQ, Feb 9, 2017.

ad: L-HROutlet
ad: l-rl
ad: Left-2
ad: Left-3
ad: abrind-2
ad: l-BCInc
ad: L-MFJ
ad: L-Geochron
  1. AF6LJ

    AF6LJ Ham Member QRZ Page

    The main mode of failure when the amplifiers in question are operated while be driven on a lower band other than what is selected by the bandswitch is the gate to drain voltage is exceeded.
  2. N1OOQ

    N1OOQ Ham Member QRZ Page

    Fascinating reading... please do keep us updated.
    AE5SB and W4IJ like this.
  3. WA4SIX

    WA4SIX Ham Member QRZ Page

    I think he has the band selector board controlling the Diplexer/LPF & that should clear this issue up, right?

  4. KD8UYQ

    KD8UYQ Ham Member QRZ Page

    Hello All,

    I'm back from my long weekend working with my buddy in Ohio on his "Boat Anchor" AM station and attending the hamfest in Mansfield, Ohio.

    Sue, I have done several bone headed things during this build including running the transceiver on one frequency while the amplifier was on another and luckily have not blown up the power FETs yet. Here is an example below. Lesson learned; don't push power into the PA board with the bias line disconnected. RF blew up the capacitor but the transistors were fine.


    While that little decoupling cap was giving up the ghost the fire and smoke coming out of it was pretty impressive.

    To Ed's point I do have manual band selection but 90+ percent of the time my Icom IC-7300 will be connected to it and will switch the band automatically when I select a different frequency on it. Plus if I do manage to kill a MRF150 or two , or four they are much less expensive to replace then a BLF188XR.

    Where was I? Setting the bias. As I mentioned before you have to work kind of fast setting the bias current or you could lift the thermistor so that it is not touching the copper heat spreader but that messes up the bias voltage. So what I did was to set the bias voltage while the heat spreader was at room temperature then quickly set the bias current of each power FET before the heat spreader warmed up to much. Look at the PA board again.


    Say that the FET on the left is #1, the one to it's right is #2 and so on. You want to set the bias current for #1 first than #4 than bounce over to number #2 and finally #3 (you could do #4 first than #1 it doesn't matter about the order as long as you avoid bringing up a pair at a time). Once again keep in mind that you need to terminate both the input and output of this board or the amplifier will break into oscillation when you try to bias power FET pair #1 and #2 or FET pair #3 and #4. I speak from experience here. Another thing to note is a counterclockwise turn of each bias current trimmer pot increases the current indicated on your DMM. You would think it would be the other way around but it's not.

    Adjusting the trimmer pot for each power FET adds to the total current draw that you are monitoring on you multi-meter. So you start cranking counterclockwise turns into the 20 turn trimmer pot for FET #1. The first ten turns or so nothing happens but then you can see that the current reading starts to climb pretty quickly. When you reach 15o mA stop and go to power FET #4. Start turning the adjustment screw on trimmer pot #4 counterclockwise and watch the current reading on your meter. The first ten or so turns nothing happens than the current reading on the meter starts to decrease, don't panic. Keep going and you will see the current level rise again back to 150 mA then start increasing rapidly from there. When you have reached 300 mAs stop. Go to the trimmer for FET #2 and keep turning the adjustment screw counterclockwise until you have reached 450 mA. The current will do that roller coaster thing again where it goes down from 300 mA before it finally starts climbing up to the 450 mA stopping point. Lastly, set the idle current for power FET #3 and stop when you see 600 mAs showing on your current meter.

    You are done. The PA board is ready to make little signals bigger.

    KK4CUL likes this.
  5. AF6LJ

    AF6LJ Ham Member QRZ Page

    This has been a cool project to follow, Thanks again for posting it.
  6. ZS1ADC

    ZS1ADC Ham Member QRZ Page

    Bob I'm watching this thread with great interest as I too have bought the parts from Russia and I'm almost ready to start my build. My kit however has the sd2933 fets. I know there are many mods so I gather as much info as I can not wanting to blow up expensive parts. So this is a great help to me and I thank you for sharing your build with us.
    Keep up your great work, Regards, Reg.
  7. WA7PRC

    WA7PRC Ham Member QRZ Page

    Soft metals such as copper and aluminum are "gummy" and don't like to be cut. Generally, especially when tapping small holes in soft metal, I use a thread forming (aka roll) tap. The drilled hole is larger than what you'd use for a cutting tap. As always, LOTS of appropriate lube is used.
  8. N1OOQ

    N1OOQ Ham Member QRZ Page

    What's different about a thread forming vs another kind of tap?
    Agreed, this is a cool project. This PCB is (I believe) briefly mentioned in SSD... I don't think they realized all of the pitfalls yet back then.
  9. WA7PRC

    WA7PRC Ham Member QRZ Page

    It seems flippant but, a thread-cutting tap removes material, while a thread-forming tap forms the material. The latter is also called a "roll/rolling tap". Roll taps have three or four lobes that push the material to form the threads, and are used with softer materials (aluminum, copper, brass, etc).
    AF6LJ likes this.
  10. N1OOQ

    N1OOQ Ham Member QRZ Page

    Hmpf... the taps we have at work have, I think, 4 lobes. They work ok for aluminum and steel as long as you go slowly.

Share This Page