New 2x LDMOS Bias Supply Front End PCB

 

Designed this new little front end PCB this weekend.

My primary goals were as follows:

  1. Provide a regulated bias supply using a precision regulator to provide more accurate regulation and some ripple rejection, unlike a straight zener.
  2. Allow for independent adjustment of each LDMOS bias, with absolutely no interaction, i.e., adjustment to one side has zero affect the other side. The reason I want this is to be able to experiment with harmonic nulling.
  3. Incorporate a bias relay and 12V input conditioning onto the PCB, for less external components.
  4. Design a specific temperature compensated bias slope of -25mV/deg C using a common, highly accurate 3-pin adjustable voltage regulator.
  5. Incorporate a single flange type attenuator into the design for simplicity and lower cost.
  6. Start working on normalization of gain, by first trying to neutralize the RF path post-attenuator through the 4:1 input transformer. First pass attempt here involves providing a termination resistor and trying to offset the capacitance in the transformer.

By far the largest point of interest here is #4. For this I set up SPICE simulations and ran Monte Carlo on resistance values in conjunction with a 2% 5k NTC thermistor, using the value curves from the manufacturer.

Simulating, I was able to come up with component values that produced exactly -25mv/C from 25C to 45C, then becomes steeper 45 to 65C, then steeper still 65C to 85C. Which I would think is ideal.

We’ll see how it all plays after I complete the redesign of the output board. This is much less involved, only dealing with mechanicals for now.

 

 

 

 

Comments

4 responses to “New 2x LDMOS Bias Supply Front End PCB”

  1. Tim Hatch Avatar
    Tim Hatch

    Rob,
    I wondered when someone would start using some sort of voltage regulator on the input. I don’t see anywhere to connect the thermistor. Circuit looks good. Let us know how it works.

    Tim
    WB0WUQ

    1. N4GA - Rob Avatar
      N4GA – Rob

      Tim,

      All the commercial amps use a regulator I think. The therm connects between the pad with the 10uF electrolytic cap and ground. It will get wired in later during final construction, as I need to drill and tap another hole.

      Thanks for the comment!

      Rob

    2. John Cunliffe Avatar
      John Cunliffe

      A way to temperature stabilize the bias is a must for stable performance. That can be as simple as inserting a diode in series with the ground connection of a 3 terminal voltage regulator with the diode typically inside a lug that is attached to the heat sink, to as elaborate as have a separate NTC stabilized circuit for each half of a transistor. I typically set the bias for each element so for this dual transistor circuit I would have 4 bias adjustments. But them I am partial since I work with commercial equipment were every single element in a high power amplifier circuit is individually adjusted for optimal bias and temperature stabilized using either manual means or a micro processor that looks at the output signal and automatically adjusts for best performance..

      1. N4GA - Rob Avatar
        N4GA – Rob

        Thank you John. Good information. I’ve done both the diode/regulator and NTC resistor divider circuits. I’ve done dual bias but not quad. For what I design to, I’ve found anything more than one bias is overkill and very difficult to initially set equal and to have them track dead equal. Now if you introduce a MCU bias control, that’s a whole new situation.

        Thanks again, Rob

Leave a Reply

Your email address will not be published. Required fields are marked *