Tweet
There is some good information here I believe, especially Wilder's comment "...no current should flow through the chassis whatsoever. If you lay out your ground scheme properly you will only have 2 connections to the chassis...amp circuit ground and the AC mains safety ground. With this arrangement the chassis only serves two purposes...(1) acts as an RFI/EMI shield and (2) acts as a "mains fuse blower" should the hot side of the AC mains come in contact with the chassis."
And O'Connor "Each node of the ground is a collection of _related_ elements and is a star, but these stars are not tied independently to the chassis. rather, they tie to a buss in an organized fashion that follows the loops of circulating currents. The entire ground system is _wired_ and related portions are twisted with supply feeds that are themselves stars on the supply feed buss. The buss has only one connection to the chassis and can be lifted if required."
I've just finished a prototype build modeled around a Tremolux 5EG and for the first time ever I had absolutely no hum and noise generated from the initial layout and wiring scheme. Zip, zero, nothing to fiddle with, on instabilities and no hum to buck as I always have had to do in the past when building an amp. They amp layout adheres to the principles in the above quotes and that is no doubt why it is so quiet.
Mine is a turret board layout. The ground bus (6 turrets connected in series in a line) begins on the far left of the turret board starting with the high current connections - Rectifier plate, first filter, and OPT center tap, then pie filter caps, output tubes and tremolo all connecting to the ground bus progressively downstream (to the right) as component layout moves in that direction. The phase splitter ground is via a wire to the right end of the ground bus, volume and tone grounds are wired directly pot-to-pot and star connected to 1st stage cathodes and the input jacks, then via wire to the right end of the ground bus. The input jacks serve as the only ground connection point to the aluminum chassis. The AC plug ground wire is bolted to the chassis near its entry point. All filaments are twisted wire AC. The chassis of the tweed era style and size. I'm going to follow this method religiously in the future.
And O'Connor "Each node of the ground is a collection of _related_ elements and is a star, but these stars are not tied independently to the chassis. rather, they tie to a buss in an organized fashion that follows the loops of circulating currents. The entire ground system is _wired_ and related portions are twisted with supply feeds that are themselves stars on the supply feed buss. The buss has only one connection to the chassis and can be lifted if required."
I've just finished a prototype build modeled around a Tremolux 5EG and for the first time ever I had absolutely no hum and noise generated from the initial layout and wiring scheme. Zip, zero, nothing to fiddle with, on instabilities and no hum to buck as I always have had to do in the past when building an amp. They amp layout adheres to the principles in the above quotes and that is no doubt why it is so quiet.
Mine is a turret board layout. The ground bus (6 turrets connected in series in a line) begins on the far left of the turret board starting with the high current connections - Rectifier plate, first filter, and OPT center tap, then pie filter caps, output tubes and tremolo all connecting to the ground bus progressively downstream (to the right) as component layout moves in that direction. The phase splitter ground is via a wire to the right end of the ground bus, volume and tone grounds are wired directly pot-to-pot and star connected to 1st stage cathodes and the input jacks, then via wire to the right end of the ground bus. The input jacks serve as the only ground connection point to the aluminum chassis. The AC plug ground wire is bolted to the chassis near its entry point. All filaments are twisted wire AC. The chassis of the tweed era style and size. I'm going to follow this method religiously in the future.
Comment