I previously purchased a 100W shortwave amplifier with the model number BJ300 for less than 500 yuan. I found some information online that suggested it was primarily designed for export. There's also a link on Zhuanzhuan (Soukupiao). After receiving it, I tested it using a G90S and found that the best results were at 1.8 MHz. With an input of 2W, I could achieve a power output of 150W. In other frequency bands, I was able to get over 100W with inputs of 3-5W (although the specifications state a peak output of 300W for short periods, and I'm hesitant to try it). It also has automatic PTT switching and SSB delay. The only major drawback is that the standing wave at low frequencies is quite high, much higher than my previous DIY amplifier. At 14 MHz, the standing wave was 4-5, and at 7 MHz, it was 2-3. However, above 21 MHz, there was no noticeable effect. I decided to investigate further.

The solution I used involved two M1104 tubes. When I disassembled them, they appeared to be in good condition and still worked. The M1104 seems to be a customized version of the MRF247 made by MRF, so there shouldn't be much difference. (Later, I saw versions of the MRF247 in other disassembly videos, and the circuit components were also consistent, which suggests that it was actually one tube with an additional casing.)
I previously posted on the HalloCQ forum asking for advice from experienced users about how to improve my SWR (standing wave ratio). I was considering using the built-in antenna tuner of my G90S, or matching the impedance at the amplifier input.
This device essentially has no impedance matching; it only includes a pair of 50-ohm resistors to ground. After researching and gathering information, I discovered that foreign researchers, specifically HAM, have already attempted to address this issue. The following method is a combination of their approaches:

The results after modification are as follows (using insulated wire and a protective tube):

After modification, a dummy load test was performed, and the HF full-band standing wave remained below 1.1, with minimal overshoot. It could drive up to over 100W within 5W. The modifications were very successful!
Of course, you can also change the capacitor C18 in the upper-left corner of the tube to adjust the delay time when using SSB mode.
Notes: The amplifier does not have a low-pass filter; be sure to include a low-pass filter board as well.