Here is the Alpha Nirvana v2 schematic with Hugh's latest tweaks:
Here is the behavior at clipping - notice how symmetric it is:
Here is the FFT for 8Vpp into 8ohms:
Here are the distortion components with 0.0039% THD:
Code:Harmonic Frequency Fourier Normalized Phase Normalized Number [Hz] Component Component [degree] Phase [deg] 1 1.000e+03 3.937e+00 1.000e+00 -0.68° 0.00° 2 2.000e+03 1.538e-04 3.906e-05 -113.08° -112.40° 3 3.000e+03 8.049e-06 2.045e-06 56.41° 57.09° 4 4.000e+03 1.027e-06 2.608e-07 164.55° 165.23° 5 5.000e+03 1.262e-06 3.207e-07 -177.53° -176.85° 6 6.000e+03 1.038e-06 2.637e-07 179.93° 180.61° 7 7.000e+03 8.854e-07 2.249e-07 179.88° 180.56° 8 8.000e+03 7.748e-07 1.968e-07 179.92° 180.60° 9 9.000e+03 6.887e-07 1.749e-07 179.93° 180.61° 10 1.000e+04 6.198e-07 1.574e-07 179.94° 180.62° 11 1.100e+04 5.635e-07 1.431e-07 179.94° 180.63° 12 1.200e+04 5.165e-07 1.312e-07 179.95° 180.63° 13 1.300e+04 4.768e-07 1.211e-07 179.95° 180.63° 14 1.400e+04 4.427e-07 1.125e-07 179.95° 180.64° 15 1.500e+04 4.132e-07 1.050e-07 179.96° 180.64° Total Harmonic Distortion: 0.003912%(0.004251%)
And here is FFT for 28Vpp into 8ohms (12.5w):
And here are the distortion components for 0.016% THD:
Code:Harmonic Frequency Fourier Normalized Phase Normalized Number [Hz] Component Component [degree] Phase [deg] 1 1.000e+03 1.417e+01 1.000e+00 -0.68° 0.00° 2 2.000e+03 2.278e-03 1.608e-04 -115.76° -115.08° 3 3.000e+03 5.799e-04 4.092e-05 51.25° 51.93° 4 4.000e+03 1.615e-04 1.139e-05 20.66° 21.34° 5 5.000e+03 6.764e-05 4.773e-06 -110.87° -110.19° 6 6.000e+03 2.623e-05 1.851e-06 -177.51° -176.82° 7 7.000e+03 8.719e-06 6.152e-07 98.94° 99.62° 8 8.000e+03 3.366e-07 2.375e-08 -37.39° -36.70° 9 9.000e+03 2.908e-06 2.052e-07 -157.53° -156.85° 10 1.000e+04 2.602e-06 1.836e-07 178.74° 179.43° 11 1.100e+04 1.990e-06 1.404e-07 176.05° 176.73° 12 1.200e+04 1.803e-06 1.272e-07 -179.77° -179.08° 13 1.300e+04 1.709e-06 1.206e-07 -179.55° -178.87° 14 1.400e+04 1.588e-06 1.121e-07 179.91° 180.60° 15 1.500e+04 1.477e-06 1.042e-07 179.90° 180.58° Total Harmonic Distortion: 0.016635%(0.016717%)
Looks like fun!
Question - what about using JFETs for LTP?
Also - is there a way to get more power? Say 100W?
Maybe you could run a pair in balanced configuration for 80 watts, but would be kind of a waste. I ran a pair of F4 stereo amps that way, great way to heat a room but you just don’t listen to 80 watts continuously. The extra watts are good for headroom and a class d or ab is far more practical.
Bill
Bill
Besides the points that wirewiggler brought up, also bear in mind that doing a balanced bridged configuration will cancel 2nd harmonic distortion, and this amplifier design was specifically tuned for a certain harmonic structure profile (see 1st page of this thread); at which point, it’s not an Alpha Nirvana anymore. Hugh Dean has written about his philosophy with regards to the harmonic profile.
Best,
Anand.
Best,
Anand.
Gee, those Modushop 5RU, 350mm deep cases (single heatsink each side) that I bought for my AN 4Rs, certainly work well as total heatsinks! 
Winter has arrived so I have put my ANs back in the system ... after most of the day turned on, panel temps (outside) are as follows:
* heatsinks (not the fins thereof) in the middle of the bolts holding the output devices - 72 deg C
* front panel, centre - 48 deg
* back panel, centre - 47 deg
* top panel, centre - 47 deg.
Winter has arrived so I have put my ANs back in the system ... after most of the day turned on, panel temps (outside) are as follows:
* heatsinks (not the fins thereof) in the middle of the bolts holding the output devices - 72 deg C
* front panel, centre - 48 deg
* back panel, centre - 47 deg
* top panel, centre - 47 deg.
Haha - nooo ... that would be going backwards, Scott.
My first build of these amps used a 3RU Modushop case, with a CPU cooler for each pair of output devices - each set of cooler fins having an 80mm supposedly-quiet, Noctua fan. This "worked" very well - in the sense that the heatsink temperature (from memory) was about 60 deg C - but even though I had slowed the fans by reducing their voltage, I didn't like hearing the noise of 4 fans. So I rebuilt the amps using 5RU Modushop 'Dissipante' cases, having a 300mm long heatsink, each side.
@Anand - yes, mine are the 4 ohm version. +/-22v DC rails, 3a bias.
Andy
Hi Vunce,
Not quite - I have a pair of stereo AN4Rs ... so 4 channels of amplification.
So each 300mm long x 200mm high heatsink is taking care of 3a x 44v = 130w, right?
And Anand ... if you think that my 72 deg heatsink temp is a problem - it would be possible to re-house the ANs a second time, by putting them into the 400mm deep Dissipante cases. This would give each output device its own 200x200 heatsink ... instead of them sharing the one 300x200 heatsink.
Or even 500mm deep! Here, each output device would have a 250x200 heatsink.
Do you think this would make a worthwhile difference to the heatsink temperature? (These are my "forever amps", after all! )
Not quite - I have a pair of stereo AN4Rs ... so 4 channels of amplification.
So each 300mm long x 200mm high heatsink is taking care of 3a x 44v = 130w, right?
And Anand ... if you think that my 72 deg heatsink temp is a problem - it would be possible to re-house the ANs a second time, by putting them into the 400mm deep Dissipante cases. This would give each output device its own 200x200 heatsink ... instead of them sharing the one 300x200 heatsink.
Or even 500mm deep! Here, each output device would have a 250x200 heatsink.
Do you think this would make a worthwhile difference to the heatsink temperature? (These are my "forever amps", after all!
)