F2 Gain

[FYC]

Hi, I took some time during the Chinese New Year Break to measure the frequency response of the F2 clone I made a few months ago. Using rather simple equipment (a 20 years old signal generator and a hand held 5M Hz Oscilloscope), I got the attached graph.

I am amazed by the high end extension, -3dB at 300kHz, no wonder Nelson said this amp is very "fast". But what annoy me a bit is that there is a 4dB difference in terms of the gain in the L and R channels. My measurement was made with an input signal of pure sine wave at about 250mV p-p and output impedance of 600 Ohms. The F2 is loaded with a 50W 8 Ohm resistor and the Gain is measured as the voltage across the 8 Ohm load.

I did notice a slight shift in the image to the L side during audition, but was surprised to see such a big difference. Can anyone help to suggest some cure? My thinking is that this must solely be due to the difference in the transconductance of the 2 IRFP044 MOSFETs in the 2 channels, as I cannot imagine anything else. Would like to hear some suggestions.

And by the way, if anyone had made any measurement of the F2, please confirm if you also get a -3dB point around 300kHz?

[Nelson Pass]

The bandwidth is typical.

I have not seen variations in gain, so something is off on your build.

Check the Source resistors and the Load resistors, since they are
the primary gain setters. The transconductance of the Mosfet figures
into it, but it would be a broken Mosfet which would make a difference.

[FYC]

Thanks Nelson for the hints.

I used the same load resistor 8 Ohm in all the test, so this eliminate 1 factor. You are right, the source resistors are rather suspicious. I remembered you mentioned once the gain is about 16dB, so seems that the R channel is closer to expectation (gain = 16dB at 8 Ohm load) while the L channel is the one to hunt down (gain = 20dB at 8 Ohm load).

[Nelson Pass]

Originally posted by FYC
Thanks Nelson for the hints.

I used the same load resistor 8 Ohm in all the test, so this eliminate 1 factor. You are right, the source resistors are rather suspicious. I remembered you mentioned once the gain is about 16dB, so seems that the R channel is closer to expectation (gain = 16dB at 8 Ohm load) while the L channel is the one to hunt down (gain = 20dB at 8 Ohm load).

Perhaps the output load resistors (3 # 47 ohms) are not connected.

[FYC]

Confirmed it is the source resistor. I use 3 2R 5W resistor to form the source resistor. One of the 2R resistor is open for some reason. Thanks.