by erichayes » Tue May 03, 2005 12:16 am
Hi All,
Good grounding is good grounding, regardless of the method (star, bus or plane) employed. And yes, it can be a gold plated supercharged pain in the ass to achieve good grounding.
Shannon, your remark about the superiority of a star ground method over plane and bus methods prompted me to drag my 50/50 watt almost-clone prototype of the HF-89 onto the bench for some spec.ing. This is a point-to-point version with minor component changes (primarily in the FB loop) using iron of my design and a choke-input filter network. A 10 ga. silver plated copper ground bus is utilized, with the chassis grounding point at the power entry module. The test equipment used were Tektronix SG-505 signal generator, AA-501 distortion analyzer, SC-503 dual channel 10 meg 'scope, and a DC-504A freq. counter (to keep the '505 honest). The residual noise floor of the test setup was 0.2mV.
Absolute values for noise floor were as follows:
Left chan: .902 mV, Right chan: 1.78 mV
With High Pass Filter engaged:
Left chan: .212 mV, Right chan: .238 mV
Relative S/N+N Ratio referred to 1 watt
Left chan: -70 dB Right chan: -62.8 dB
HPF in LC -83.1 dB RC -81.4 dB
Relative S/N+N ratio at .5% THD:
LC: -87 dB RC: --80.4 dB
HPF in LC -99.6 dB RC -98.3 dB
The distortion at one watt was .0089% on the left channel and .023% on the right. I think the distortion/hum discrepancies are due to a squirrely 6SN7 in the right channel.
My point is that bus bar grounds are every bit as good as star grounds, if they're implemented correctly (and, in my opinion, a damned sight better). Ground planes, on the other hand, create capacitave disasters. You might get rid of hum with a GP, but you're going to screw up all the other parameters of an amp that separate the good ones from the mediocre.
Eric in the Jefferson State