MRF101 150W Broadband HF Amplifier

This is a homebrew amplifier for 160 through 10m bands.

The amp works well on 30m and higher frequency bands. While operating FT8 on lower frequency bands, T2 and T3 are heating up and causing efficiency, and hence output power, to drop. I need to investigate this.

Amplifier installed in HP436a Power Meter chassis.

Low Pass Filters must be added in order to comply with FCC requirements. Suitable Low Pass Filters include https://www.w6pql.com/a_1.5kw_lpf_for_160-6m.htm.

The heatsink is a Communications Concepts, Inc., Model 55HS.

Ground the Bias Control, pin 1 on J3, to enable the amp by applying bias to the MRF101s. With gate bias voltage initially set to zero volts, R7 and R8 are adjusted for 100mA quiescent drain current on the MRF101AN and MRF101BN, respectively.

MRF101 150W Broadband Amplifier Schematic Diagram
MRF-101 150W HF Broadband Amplifier
MRF-101 150W HF Broadband Amplifier
View of T2
MRF101AN and MRF101BN Mounting Detail

6 Replies to “MRF101 150W Broadband HF Amplifier”

  1. Very nice project! I just ordered up a few pairs of the mrf101. I’d like to develop a larger ldmos amp at some point, but my knowledge on the matter needs some complimentary experience before going bigger.

    I imagine the input power to this is almost nothing. I’ve been contemplating a way to measure the input match, first thought was an SWR meter but I can’t find one sensitive enough on a hobbyist budget. Second thought was to use my oscope to compare the phase of voltage and current. Were you able to get any measurements on the input? I’m curious to know how well your match is across the bands. Of course an attenuator will buffer most, if not all, of the mismatch.

    For the output transformer, is there any reason you chose not to use transmission line?

    1. Thanks for the feedback!

      The -3dB pad at the input ensures the transmitter sees a nominal 50 ohm load across all bands. The transmitter drive level is very low and the transmitter behaves properly when driving the amplifier, so I didn’t find it necessary to investigate the SWR between the transmitter and amplifier.

      I chose a conventional transformer based on past experience. I would be curious to see the results with a transmission line transformer if you go that route.

      Regards, Rod

  2. Hello, The result looks fabulous. I am also looking into to build such a rf amp. But looking at the spec sheet of the MRF101AN the input impedance is about 25.3 + J10.2 for 13 Mhz. If I look to the input circuit I see 3db attenuator 50 ohm than a transformer which I think creates 2 time 50 ohm balanced loaded by 2 times 56 ohms plus the 25.3 + J10.2 ohm of the mosfet. Would it not be better to change the input transformer to 4:1 to have better matching? The + J10.2 can be compensated by 1 NF in series with the gate instead of 100NF but probably does not matter so much. But any way the results is impressive.

    1. Hi Paul,
      Thanks for the feedback!
      Keep in mind this is a broadband amplifier. The 3dB pad in the input provides a broadband match for the transmitter driving the amplifier. The 1:1 input transformer is providing drive to each MRF101AN that is 180 degree out of phase. The 0.1uF capacitors are for DC blocking. The amplifier has sufficient gain to reach 150W output with ~0.5W drive. As long as the transmitter driving the amplifier is happy, the input matching is less critical. Improvements in this design are needed in the output circuit as noted.

      The matching circuit impedance you are referencing from the MRF101 datasheet is specified for optimized gain, output power, and stability with a reasonable return loss at specified frequency, bias and input power. Since very high return loss was not an objective, the input impedance of the MFR101 is not necessarily the conjugate of the impedance presented by the matching network.

      Regards,

      Rod

      1. Hi Rod, Thanks for the reply. One more question. I see you use T2 to apply power to the mosfets. Is this prefered over using the middle tap on the output transformer? That would save one transformer. Or is it necessary for balancing the fets better.

        Regards,
        Paul

        1. Hi Paul,

          I have typically used a separate choke. Electrically it shouldn’t make a difference. I have seen others do this, and also use a separate choke with the two center taps connected together. From a PCB layout perspective, I prefer to keep the DC circuitry more towards the RF input end of the board which is facilitated by the separate choke.

          Regards,
          Rod

Leave a Reply

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