normal X-Amp

[uli]

I did some drawings, maybe the wizzards may comment them...

Uli

[uli]

data size

[uli]

data size 2

[uli]

the idea behind that circuit was to combine some already
working designs. The output section is very similar to the
proven Pass/Thagard A-75. The idea of leaving the amount
of overall feedback to the builders choice fascinated me and
therefore I implemented this feature with R30 and R47 (XX).
How it works can be read in the Pass/Thagard article in
AudioAmateur or at passdiy.com. I simply adapted their
frontend to modern Mosfets. The amount of bias (ClassA)
can be adjusted easily by trimming R39 and R51 to about
0.14 volts across one of the 0R33 source resistors. This
amount of bias is good for 25W RMS ClassA and leads to
about 180W to be handled by the heatsinks, 15W per fet.
For thermal stability T11, T14 must be mounted on the
heatsink near the output devices.
The trimming of the CCS´s is needed for offset adjustment
after a reasonable warm up time. To achieve best results
both inputs should be grounded during adjusting R1, R3.
The amp is capable of about 200W/8R and 400W/4R RMS.
The only tricky thing is the PSU. I use a hammond toroid with
1000VA and 48V CT. and a small one 2x 15V (independent!)
for the auxiliary PSU (+-55V) Powercaps are 4x Siemens
47mF/40V

Uli

[Peter Daniel]

fixed

[uli]

phantastic service you moderator guys!!!
TX Uli

[Joe Berry]

Regarding the power supply, you could of course use a voltage
doubler to generate a raw +/-70VDC, which you can filter and/or
regulate down to +/-55V. This eliminates the need for a separate
transformer and its associated wiring, mounting, and fusing.

If the finished amplifier has audible background hum or can't
make full output at high frequencies, you could reduce the 100R
value between the sources of the input diff pair to increase the
open loop gain of the voltage section and make the feedback
loop more effective. Also, as you've referenced both input gates
to ground, I don't think you'll need the 630R to ground at the
sources of the input diff pair.

I'm curious as to why you referenced the cascode gates to the
filtered output stage supply rails rather than to a zener string in
the front end section. Does this ensure that the voltage section
will clip before the output stage does?

[uli]

youre right
Uli

[Charles BL]

Sound?

[uli]

as I never heard an original X-series amp, I can´t tell.
From listening to my Aleph 2´s (tweaked: 35V supply rails but
about 5A bias) the X has better ("clearer") bottom with more
punch!
I use them now for the bass below 95Hz for 2x12" per side
biamping.(the Alephs for the rest)
Uli