My WA2EBY MOSFET Amplifier
A Homebrew Debugging Story
After carefully building the WA2EBY dual-IRF510 MOSFET amplifier exactly as described in the original project article, I was disappointed to see only about 5 watts RF output on the bench. Each IRF510 was correctly biased at 10 mA, the amplifier was powered from a clean 13.8 VDC bench supply, and I even experimented with raising the supply to 19 volts. Yet the output stubbornly remained low. Even more suspicious was the current draw over 2.5 amps on key-down , clear evidence that something in the amplifier was consuming power without converting it efficiently into RF.
The troubleshooting process began methodically. I verified the input attenuator pad values and wiring, checked all toroid windings against the published turns count and core types, and confirmed correct phasing on the transformers. Using an oscilloscope, I examined the RF signal splitting and combining networks; both appeared symmetrical and healthy. To eliminate any possibility of relay losses or T/R switching problems, I bypassed the relay entirely and connected the amplifier directly to a 50-ohm dummy load. Still, the meter stubbornly indicated only 5 watts output.
At that stage, suspicion naturally shifted toward the output transformer T3. In RF power amplifiers, excessive current draw combined with low RF output often points toward a partially shorted winding or a single shorted turn in the output transformer , effectively turning RF energy into heat instead of useful output power. I was already mentally preparing to rewind the toroid.
Before tearing the amplifier apart, however, I decided to perform one more measurement. I quickly built a simple RF probe and checked the RF voltage directly at both MOSFET drains during transmit. The result was surprising: both drains showed a healthy RF voltage swing of approximately 13 volts peak during key-down. That measurement changed everything. The RF voltage was clearly there, meaning the amplifier stage itself was functioning properly. The problem was not the amplifier,it was the wattmeter.
I replaced the suspect power meter with another unit, and instantly the amplifier came alive: approximately +/- 20 watts RF output with only 1 watt of drive at 13.8 volts. Current consumption now made sense, efficiency looked reasonable for an IRF510 design, and the MOSFETs warmed normally without any signs of thermal stress or instability.
The final conclusion was both satisfying and educational. The WA2EBY amplifier design itself was solid all along; the real fault was a misleading RF power meter. Many inexpensive through-line wattmeters, especially older CB-style meters or low-cost HF meters can be wildly inaccurate at low power levels, frequency-dependent, or even introduce mismatch problems that affect measurements. A simple RF probe turned out to be the most trustworthy diagnostic tool on the bench.
One important lesson from this project is never to rely entirely on a single instrument when troubleshooting RF equipment. Cross-checking measurements with an RF probe, oscilloscope, dummy load, or second wattmeter can save hours of frustration and unnecessary rebuilding.
Now that the amplifier is delivering a solid +/- 20 watts output, the project moves into its next stage. The immediate goals are building an RF sampler for safe monitoring with a PC-based oscilloscope and constructing proper low-pass filters for clean 40m and 20m operation. After that comes the most rewarding phase of any homebrew project real over-the-air testing. There is something deeply satisfying about watching a pile of components evolve into a fully functional on-air RF amplifier built with your own hands.
-------------------------------------------------------------------------------
Brainstorming ideas and layout concepts for a Manhattan-style RF circuit board before beginning final construction. Careful component placement, short RF paths, solid grounding, and logical stage separation are critical in Manhattan construction, especially for broadband HF power amplifiers. Much of the early design effort focused on minimizing stray inductance, maintaining symmetry in the push-pull layout, and ensuring adequate space for transformers, heatsinking, and future modifications.
The RF power stage of the amplifier is built around a pair of IRF510 MOSFETs operating in push-pull configuration. Although originally intended as switching devices, the IRF510s have proven to be inexpensive, rugged, and very capable performers in low-voltage HF linear amplifiers when properly biased and matched. Each device was mounted to a large finned aluminum heatsink using insulating pads, shoulder washers, and thermal compound to ensure effective heat transfer while maintaining electrical isolation from the chassis.
Under normal SSB operation, the heatsink remains only moderately warm, even at approximately +/- 20 watts RF output. However, to improve long-term thermal reliability during extended key-down or digital-mode operation, I added a small 12 VDC cooling fan to provide continuous airflow across the heatsink. Particular attention was also given to the physical layout: the MOSFETs were mounted as close as possible to the output transformer and combiner network in order to minimize drain lead length, reduce stray inductance, and maintain good RF symmetry within the push-pull stage.
-------------------------------------------
A short clip with bad focusing and shaking picture frame 😁😂
