Ever notice radios like the Galaxy Saturn or others with built in echo never have the same fidelity in transmit that the same radio board without the internal echo does? For example, the 2950 board is nearly identical to the Saturn but has much fuller sounding audio with wider bandwidth.
The built in echo has very limited bandwidth and will chop off both treble and bass. Unfortunately just because you may opt to never use the built in echo, it still produces all of its negative effects in the audio even when it's off. That's because the TX audio still passes through 90% of the echo circuitry when it's off.
Remove the center wire on the mic gain control and tape it off. Then connect a new wire from the center of the mic gain control to the center pin of Q32. That's the collector of the audio limiter transistor. You will likely find another brown wire connected to the same trace coming from the echo board. It's a good idea to remove this too.
The scope and signal generator confirm a notable improvement in the TX bandwidth after this echo bypass. The mod is the same even if you have removed Q32 although, I recommend you don't. These radios produce objectionable distortion when severely overmodulated.
If you leave Q32 in place and adjust the potentiometer, you don't have to be nearly as critical with the mic gain setting. You can now drive the audio hard enough to produce some compression that gives it punch without all the flat topping and RF cutoff.
I got to see the NPC diode inline with the resistor too. First thing I did after seeing what it does to the carrier was rip the diode out and put it back to stock. All it does is drive the carrier up higher with your audio and the proof of this is that after every period of hard modulation you can see the carrier at a much higher level and quickly fall back to its setting faster than a meter could track it.
It makes the watt meter look like there is lots of forward swing and should have louder audio but since the carrier is moving up, this counters any increase in modulation. I can also see why this is confused with NPC since driving the carrier higher prevents negative peak clipping. I suspect the diode may simply be rectifying the some of the modulated DC and just adding it to the DC carrier setting.
The built in echo has very limited bandwidth and will chop off both treble and bass. Unfortunately just because you may opt to never use the built in echo, it still produces all of its negative effects in the audio even when it's off. That's because the TX audio still passes through 90% of the echo circuitry when it's off.
Remove the center wire on the mic gain control and tape it off. Then connect a new wire from the center of the mic gain control to the center pin of Q32. That's the collector of the audio limiter transistor. You will likely find another brown wire connected to the same trace coming from the echo board. It's a good idea to remove this too.
The scope and signal generator confirm a notable improvement in the TX bandwidth after this echo bypass. The mod is the same even if you have removed Q32 although, I recommend you don't. These radios produce objectionable distortion when severely overmodulated.
If you leave Q32 in place and adjust the potentiometer, you don't have to be nearly as critical with the mic gain setting. You can now drive the audio hard enough to produce some compression that gives it punch without all the flat topping and RF cutoff.
I got to see the NPC diode inline with the resistor too. First thing I did after seeing what it does to the carrier was rip the diode out and put it back to stock. All it does is drive the carrier up higher with your audio and the proof of this is that after every period of hard modulation you can see the carrier at a much higher level and quickly fall back to its setting faster than a meter could track it.
It makes the watt meter look like there is lots of forward swing and should have louder audio but since the carrier is moving up, this counters any increase in modulation. I can also see why this is confused with NPC since driving the carrier higher prevents negative peak clipping. I suspect the diode may simply be rectifying the some of the modulated DC and just adding it to the DC carrier setting.
