central ground in leach amp

[Wagener]

there is two central grounds in the leach amp. One at the output and on next to signal and signal ground.

I suppose I should use the one next to signal

The pad used here is very small. What size wire should I use to run the ground from the transformer to here?

As far as I know the same power will run in this ground wire than the power that will run in one of the rails? Why did Dr. Leach make such a small ground pad? I am not sure where to run the 0 of my transformer to the board.

[Tony]

hi,

both of them go to the central ground star point, a ga22 wire is enough....the speaker ground should also go to the central ground point....

tony

[djk]

There are NO central ground points on the Leach amplifier boards.

There is a input signal ground.

There is a non-signal ground (power supply bypass caps).

R51 goes between these two grounds, its value is usually less than 10R.

The input jack must not be grounded to the chassis.

R50 and C25 mount directly to the output binding posts.

Construct a power supply central ground.

I use a copper strap between the two main filter caps. I drill three holes in the middle. The middle hole gets the transformer center tap. Each of the other holes get the speaker ground and the input and non-signal ground from either the left or right channel.

The input and non-signal grounds are returned to the central ground each with 18ga wires.

After Nelson Pass, I use a 4R inrush current limiter from the center tap to the power cord ground. Mark Levinson uses 10R with reverse parallel diodes and a 0.1µF capacitor.

If you have a bad ground loop usually R51 blows.

The heatsinks and chassis should all be grounded. Paint and anodizing does not conduct electricity, scrape it back as needed.

Rubber insulator washers filled with boron nitride are better than mica and grease. HOWEVER, how do you know what they are selling you? Non-filled ruber washers are much worse than mica. Good washers cost more than the transistors they mount. How well do you know your supplier? I won't buy the high priced washers unless they are in a sealed bag identified by the manufacturer.

If you substitute plastic for metal outputs just remember to use twice as many. Metal cased Motorola (ON Semiconductor) outputs can take higher temperatures than plastic devices, thus more devices are need when using plastic. A MTBF study by RCA showed that metal had much longer life expectancy than plastic.

[pooge]

Originally posted by djk

I use a copper strap between the two main filter caps. I drill three holes in the middle. The middle hole gets the transformer center tap. Each of the other holes get the speaker ground and the input and non-signal ground from either the left or right channel.

It is best if you don't drill the holes directly in line between the filter cap studs. Offset them from that line an inch or so. Doug Self showed a measurable distortion phenomenon from mounting central ground in line with studs. It is due to the high pulse current between the studs. The current between the studs will create a voltage divider with your central ground point. Since you will not likely mount the central ground points at the exact electrical midpoint between the studs, you will not have a true zero voltage point there. Offsetting the CG point will get it out of that heavy current path. Again, the distortion was measurable and surprising, as well as the simple cure.

[djk]

" Doug Self showed a measurable distortion phenomenon from mounting central ground in line with studs"

If using PC board or such.

There is about 50A peak current in this buss so you must be careful.

I have a couple amps running stereo 2R loads from a common ±94V supply. This works out to about 80A peak through this buss. The loudspeakers are 107dB/W and the listening distance less than 10ft. No hum here.

My work with Hafler DH200/220/500s show that the ground must be dead center on this buss, or it humms. Hafler used a 16ga wire, totally inadequate for the peak currents involved. Moving the ground point a fraction of an inch changed the hum.

I bought a roll of copper strap about 7/16" X 3/32", no hum here.

My holes are drilled in a small triangle, with about 1/16" between each hole. The middle hole is at the apex of the triangle and is dead center lengthwise. I've tried one larger hole, but it doesn't work as well as the wires can twist and move during soldering.

Feel free to use the Self method, it doesn't work for me.