Circuit Diagram
 Bill of Materials
　Previously, I created PA-30 (two 15W units combined) using Mitsubishi MOS-FET (RD16HHF1), but the challenge at that time was to make an amplifier that did not drop higher order distortion (5th/7th/9th order distortion) than I expected. This time, the basic design was done by my acquaintance JA4DUX, but after various experiments such as opinions from each station that created it and changing the core/device, I started with MRF-150, but by changing to SD2931-11, IMD decreased and power gain increased. This time, I am using it with bias in the AB class area. The supply voltage depends on the output power setting, but I am driving it at 70V because I can get the final power I need with 20W output, but if I am conscious of up to 50W, I think it would be better to drive it at 60V to be safe. Currently, there is a difference in power gain between bands, so I am only using it in the 3.5M/7M bands. I have considered various ways to use it in common on the HF-Hi bands, but I have not yet been able to achieve it without sacrificing performance on the Low bands. It seems that it is necessary to consider a dedicated unit for the Hi band separately from the Low band. This time, I will introduce the amplifier I created for 3.5M/7M.


 　　MAIN-PWB　　　　　Parts-Side　　　　　　　    　　　　Solder-Side　　　　    　　　　   　　　SUB_1PWB　　SUB_2PWB
The SUB_1 and SUB_2 boards were etched by an acquaintance of mine, JA4DKS, which helped to create a nice finish.


　　　　　　　　　　　　MAINPWB-Assy完　　　　　　　　　　　　　　　　　　SUB_1Assy
　　　　　　　　　　　　　　　　　　　　　　　　　　　　　　　　　　　　　　　　SUB2_Assy
　　　　　　　　　　　　　　　　　　　　　　　　　　　　　　　　　　　　　　　SUB2_Assy（Panel mounting surface）


Internal structure
A 24V fan is attached to the insole plate, drawing air in from the bottom, blowing it onto the radiator and expelling it out the side. The heat sink (all copper) has plenty of room, so strong airflow is not necessary. The power supply consists of two 3.5A power supplies in series with a voltage range of 24V to 36V, allowing the voltage to be set between 48V and 72V. The MAIN unit and SUB_1 unit are attached to the heat sink.
     


The power gain is 20W output for 150mW input at 7MHz. Power gain = 21dB At 3.5MHz, the gain is higher than at 7MHz, so the coupling PAD is set to 7.5dB so that the input power is the same 150mW but the output is 20W. By switching to 7MHz = 0.5dB and 3.5MHz = 7.5dB, 20W output is obtained with 150mW input. The SUB_2 board switches the band between 3.5MHz and 7MHz (PAD switching) and controls the fan. There is room in the heat sink, so a strong wind is not required, but since it is installed in an operating room, the microphone absorbs as much fan noise as possible, so it is controlled to rotate at high speed in receive mode and low speed in transmit mode. However, since there are times when it is used in continuous transmit mode with a signal source other than a microphone, a FAN-SW is provided to force high speed rotation even in transmit mode.
The bias current is set to a total of 1.4 to 1.5 A. I adjusted the bias while checking the 2TONE waveform, and it was possible to further reduce IM3, but it is related to IM5, and if the absolute value of IM3 is lowered, IM5 will float. There seems to be no problem from the 7th order onwards, so the bias current is determined by where you set IM5 and IM3 depending on the power you use. I use it at 20W output, so I measured it under these conditions.  
Output: 20W 90Vpp  
3.6MHz 20W output 2TONE waveform IM3=-59dB IM5=-67dB  
7MHz 20W output 2TONE waveform IM3=-49dB IM5=-69dB
IM3 is worse than 3.6MHz, but it gets better below 20W, so I'm satisfied.		  
I am considering a case, but Takachi will cost about 10,000 yen, so I am still considering it. The device I used this time, the MRF150, does not provide the IMD results shown above, there is a difference of about 7 dB, and the power gain of the MRF-150 is also lower. Next I will try it for the Hi band. Thanks to JA4DUX/JA4DKS. Thank you very much.