banner
 Home  │ Audio Home Page 


Copyright © 2016 by Wayne Stegall
Updated September 2, 2016.  See Document History at end for details.

One-Bend Amplifier

Part 4: Improvements include lower hum

Introduction

After the number of output transistors was increased to eight to improve the thermal analysis it was deemed also that an improvement in hum was likely if the inner feedback loop returned to the cascode of the first stage rather than directly to the second.  Therefore I deemed it necessary to implement the changes and re-validate the SPICE analysis.  Circuits and analysis follow.

Circuits

Note:  Component values can be obtained from the SPICE models for now.

Figure 1:  Amplifier Schematic
ampmoscfb4.jpg
SPICE model

Figure 2:  Unregulated power supply
pwrsupply2.jpg
SPICE model


Figure 3:  Regulated power supply
pwrsupply2.jpg
SPICE model

Spice Results


Begin hum measurements with unregulated supply to first two stages

Fourier analysis for vout:
  No. Harmonics: 16, THD: 67.8647 %, Gridsize: 200, Interpolation Degree: 3

Harmonic Frequency Magnitude  
 Decibels

1 120 0.00295888
-50.5775
2 240 0.00144109
-56.8262
3 360 0.000940647
-60.5315
4 480 0.00066636
-63.5258
5 600 0.000506254
-65.9126
6 720 0.000385225
-68.2857
7 840 0.00030367
-70.352
8 960 0.000233349
-72.6399
9 1080 0.000184023
-74.7026
10 1200 0.000138508
-77.1705
11 1320 0.000106726
-79.4346
12 1440 7.62605E-05
-82.354
13 1560 5.59989E-05
-85.0364
14 1680 3.58895E-05
-88.9007
15 1800 2.37921E-05
-92.4713

Calculate total hum from apparent THD.

(1)
 vhum = vhum1 ×  (1 + THDapparent) = 0.00295888 × (1 + 0.678647 ) = 4.96692mV  

(2)
 vhum = 20log(4.96692mV) = -46.0783dBV 

Because hum is higher than suspected, go back to supply regulating first two stages and analyze again

Fourier analysis for vout:
  No. Harmonics: 16, THD: 74.2546 %, Gridsize: 200, Interpolation Degree: 3

Harmonic Frequency Magnitude  
 Decibels

1 120 9.80761E-07
-120.169
2 240 4.71169E-07
-126.536
3 360 3.25877E-07
-129.739
4 480 2.36987E-07
-132.506
5 600 1.93989E-07
-134.244
6 720 1.59653E-07
-135.936
7 840 1.37583E-07
-137.220
8 960 1.20571E-07
-138.375
9 1080 1.06839E-07
-139.425
10 1200 9.66625E-08
-140.295
11 1320 8.76482E-08
-141.145
12 1440 8.05205E-08
-141.882
13 1560 7.44112E-08
-142.567
14 1680 6.9202E-08
-143.198
15 1800 6.45515E-08
-143.802

Calculate total hum from apparent THD.

(3)
 vhum = vhum1 ×  (1 + THDapparent) = 9.80761e-07 × (1 + 0.742546 ) = 1.70902µV  

(4)
 vhum = 20log(1.70902µV) = -115.345dBV 

Now that hum is acceptable continue with distortion analysis at 1kHz.

Fourier analysis for vout:
  No. Harmonics: 16, THD: 0.0152017 %, Gridsize: 200, Interpolation Degree: 3

Harmonic Frequency Magnitude  
 Norm.Mag  
 Percent  
 Decibels



1 1000 21.1954
1
100
0
2 2000 0.00321471
0.00015167
0.015167
-76.382
3 3000 0.000216674
1.02227E-05
0.00102227
-99.8087
4 4000 1.73009E-05
8.16258E-07
8.16258E-05
-121.763
5 5000 1.55152E-06
7.32006E-08
7.32006E-06
-142.71
6 6000 1.34958E-07
6.36731E-09
6.36731E-07
-163.921
7 7000 2.64258E-08
1.24677E-09
1.24677E-07
-178.084
8 8000 3.92902E-08
1.85372E-09
1.85372E-07
-174.639
9 9000 2.10337E-08
9.92371E-10
9.92371E-08
-180.067
10 10000 1.5393E-08
7.26245E-10
7.26245E-08
-182.778
11 11000 1.76767E-08
8.33989E-10
8.33989E-08
-181.577
12 12000 4.92896E-09
2.32549E-10
2.32549E-08
-192.67
13 13000 3.43396E-08
1.62014E-09
1.62014E-07
-175.809
14 14000 2.99262E-08
1.41192E-09
1.41192E-07
-177.004
15 15000 2.39516E-08
1.13004E-09
1.13004E-07
-178.938

Validate the remainder of the design.

Figure 4:  Stability analysis  showing 85º of  phase margin.
stabilityol.jpg


Figure 5:  Small signal error curve.
Shows dominance of second harmonic.
 		 
 Figure 6:  Large signal error curve.
Shows dominance of second harmonic then third.
trerrsss.jpg
trerrls.jpg


Figure 7:  Bode shows frequency response down 3dB at 495kHz
bode.jpg



Conclusions

Hum is considerably improved relative to the last implementation with regulation however not enough to be rid of the regulators altogether.  The remainder of the analysis remains as good as before.

1Note:  Raw Fourier analysis data has been processed in spreadsheet to calculate related results and reformatted.
2See related article:  Thermal Design Part 2.
3See related article:  Floating Source-follower Regulator.

Document History
August 20, 2016  Created.
August 27, 2016  Added note to get component values from SPICE models added missing equation numbers for calculations.
September 2, 2016  Minor correction to caption to figure 7.