Two unbal. power amp vers based of Pass "X" (SuSy) with sim. results (Fig. 3 from US patent 5376899)
I have read the article about US-Patent 5376899 All topology versions from there you can find by
http://www.pat2pdf.org (please fill in 5376899 before download)
Short form describtion about the basic schematics from Fig. 1-4
Fig 1: single-ended folded-cascode voltage gain stage/current source, inverting mode
Fig.2: complementary-symmetry version of folded-cascode voltage gain stages, inverting mode
Fig.3 and Fig.4: same as Fig. 1 and 2, but include push pull power follower, that is in NFB loop
All versions from Fig. 1-4 are balanced versions, i. e. to use only for bridged mode.
I looked by "circuitmaker" at some relationships between two NFB variations and THD at low and high frequencies. At bottom the results - but please note, that is not really an orig. X-model, because I choise unlal. (unbridged) version. Please note secondly follow: by this simulation all capacitors and resistors so as the voltage source (power supply) and input source (signal generator) are idealized (i. e. no internal resistance by the power supply). In additional there are no lead respective wire inductance and spread capacitance through PCB layout.
Only the effects of pure circuit - are to see.
In real life (few years ago) only version with lower current and voltage I have tested (I have used as evaluation unity gain preamplifier with additional compensation capacitor) - so beginners should not be evaluate and not trying for diy. Proffesionals should be use excellent high performance heatsinks, even by the MOSFETs for the voltage amplifier (my rule is follow: heatsink temperature above ambient: < 20 degrees - aprox the same value as my hands)
Also of interest it could be, whether there are great differences by simulation results in other cad software (e.g. Orcad, Microsim or LTspice). Perhaps one of you this one can check out because the circuit is very simple.
Following parameters are valid for all versions:
Voltage gain = 10-times, input 1Vss, output 10Vss, wave form: sine wave.
Both PDF images are directly comparable, after performing download of all pdf files.
My simulations based on the unbridged version of Fig. 3 of us patent 5376899 for inverting mode matched for 60 VDC single ended power supply. I choise a unsymmetrical power supply, so I don't have trouble with dc offset so as dc speaker protection. Both versions I have simulated also with single ended power follower version, but the difference compare to push-pull version is not to observe by 2 times of idle (quiescent) current though the SE output. A good feature is to work with variable source resistor for choise the open loop gain (possible through two independend current source)
Follow sequence for the Images I have choise:
A: Pass X unbal. Var. I NFB incl. PP-output
1) schematic for Damping factor test
2) AC-Analysis (damping factor over frequency)
3) AC Analysis (normally frequency response)
4) Fourier Analysis lin 10 KHz (THD) K2: 9mV K3: 10,5mV K5: 0,54mV
5) Fourier Analysis log 10 KHz (THD)
6) Fourier Analysis lin 100 KHz (THD) (K2: 75mV K3: 74mV)
7) Fourier Analysis log 100 KHz (THD)
8) Fourier Analysis log 19/20 KHz (IM)
9) Transient Analysis 1MHz
10) schematic
No 9) only to observe, if there are
- sawtooth-deforming (risk of TIM, in generall still present by NFB about two voltage gain stages)
- or not critical and audible low pass-deforming (no risk of TIM, only delay in setting time)
I have read the article about US-Patent 5376899 All topology versions from there you can find by
http://www.pat2pdf.org (please fill in 5376899 before download)
Short form describtion about the basic schematics from Fig. 1-4
Fig 1: single-ended folded-cascode voltage gain stage/current source, inverting mode
Fig.2: complementary-symmetry version of folded-cascode voltage gain stages, inverting mode
Fig.3 and Fig.4: same as Fig. 1 and 2, but include push pull power follower, that is in NFB loop
All versions from Fig. 1-4 are balanced versions, i. e. to use only for bridged mode.
I looked by "circuitmaker" at some relationships between two NFB variations and THD at low and high frequencies. At bottom the results - but please note, that is not really an orig. X-model, because I choise unlal. (unbridged) version. Please note secondly follow: by this simulation all capacitors and resistors so as the voltage source (power supply) and input source (signal generator) are idealized (i. e. no internal resistance by the power supply). In additional there are no lead respective wire inductance and spread capacitance through PCB layout.
Only the effects of pure circuit - are to see.
In real life (few years ago) only version with lower current and voltage I have tested (I have used as evaluation unity gain preamplifier with additional compensation capacitor) - so beginners should not be evaluate and not trying for diy. Proffesionals should be use excellent high performance heatsinks, even by the MOSFETs for the voltage amplifier (my rule is follow: heatsink temperature above ambient: < 20 degrees - aprox the same value as my hands)
Also of interest it could be, whether there are great differences by simulation results in other cad software (e.g. Orcad, Microsim or LTspice). Perhaps one of you this one can check out because the circuit is very simple.
Following parameters are valid for all versions:
Voltage gain = 10-times, input 1Vss, output 10Vss, wave form: sine wave.
Both PDF images are directly comparable, after performing download of all pdf files.
My simulations based on the unbridged version of Fig. 3 of us patent 5376899 for inverting mode matched for 60 VDC single ended power supply. I choise a unsymmetrical power supply, so I don't have trouble with dc offset so as dc speaker protection. Both versions I have simulated also with single ended power follower version, but the difference compare to push-pull version is not to observe by 2 times of idle (quiescent) current though the SE output. A good feature is to work with variable source resistor for choise the open loop gain (possible through two independend current source)
Follow sequence for the Images I have choise:
A: Pass X unbal. Var. I NFB incl. PP-output
1) schematic for Damping factor test
2) AC-Analysis (damping factor over frequency)
3) AC Analysis (normally frequency response)
4) Fourier Analysis lin 10 KHz (THD) K2: 9mV K3: 10,5mV K5: 0,54mV
5) Fourier Analysis log 10 KHz (THD)
6) Fourier Analysis lin 100 KHz (THD) (K2: 75mV K3: 74mV)
7) Fourier Analysis log 100 KHz (THD)
8) Fourier Analysis log 19/20 KHz (IM)
9) Transient Analysis 1MHz
10) schematic
No 9) only to observe, if there are
- sawtooth-deforming (risk of TIM, in generall still present by NFB about two voltage gain stages)
- or not critical and audible low pass-deforming (no risk of TIM, only delay in setting time)
Two unbal. power amp vers based of Pass "X" (SuSy) with sim. results (Fig. 3 from US patent 5376899) continued
B: Pass X unbal. Var. II PP-output not in the NFB loop
1) schematic for Damping factor test
2) AC-Analysis (damping factor over frequency)
3) AC Analysis (normally frequency response)
4) Fourier Analysis lin 10 KHz (THD) K2 25mV K3: 11,5mV K4: 1,94mV
5) Fourier Analysis log 10 KHz (THD)
6) Fourier Analysis lin 100 KHz (THD) K2: 190mV K3: 70mV K4: 7,8mV
7) Fourier Analysis log 100 KHz (THD)
8) Fourier Analysis log 19/20 KHz (IM)
9) Transient Analysis 1MHz
10) schematic
No 9) only to observe, if there are
- sawtooth-deforming (risk of TIM, in generall still present by NFB about two voltage gain stages)
- or not critical and audible low pass-deforming (no risk of TIM, only delay in setting time)
B: Pass X unbal. Var. II PP-output not in the NFB loop
1) schematic for Damping factor test
2) AC-Analysis (damping factor over frequency)
3) AC Analysis (normally frequency response)
4) Fourier Analysis lin 10 KHz (THD) K2 25mV K3: 11,5mV K4: 1,94mV
5) Fourier Analysis log 10 KHz (THD)
6) Fourier Analysis lin 100 KHz (THD) K2: 190mV K3: 70mV K4: 7,8mV
7) Fourier Analysis log 100 KHz (THD)
8) Fourier Analysis log 19/20 KHz (IM)
9) Transient Analysis 1MHz
10) schematic
No 9) only to observe, if there are
- sawtooth-deforming (risk of TIM, in generall still present by NFB about two voltage gain stages)
- or not critical and audible low pass-deforming (no risk of TIM, only delay in setting time)
hallo Mr. Pass
yes, therefore I must introduce the second bridge half so as two additional resistors connected to the two ends of the existing resistor (for the presently setting of wanted open loop gain) to earth (GND). But the question is, which value is the best choise; there are no dc and signal about the resistor by identical halves but in real life flows the correction current for the distortion cancellation through residual tolerances between both bridge halves. It is interesting to know, what happens, if I introduce tolerances in one bridged half - but it can take some time. I think 2-3 weeks. I am happy, if you can give me advices about the values regarding these resistors. Perhaps in this time I have some comparative results of simulation from some other members (e. g. j.carr or j.curl), in order to assess how accurate is the modeling of my.
Unfortunately I don't know enough about the handling with ORCAD, LTspice or MICROSIM to compare the results.
All the best
Andreas Kirschner
yes, therefore I must introduce the second bridge half so as two additional resistors connected to the two ends of the existing resistor (for the presently setting of wanted open loop gain) to earth (GND). But the question is, which value is the best choise; there are no dc and signal about the resistor by identical halves but in real life flows the correction current for the distortion cancellation through residual tolerances between both bridge halves. It is interesting to know, what happens, if I introduce tolerances in one bridged half - but it can take some time. I think 2-3 weeks. I am happy, if you can give me advices about the values regarding these resistors. Perhaps in this time I have some comparative results of simulation from some other members (e. g. j.carr or j.curl), in order to assess how accurate is the modeling of my.
Unfortunately I don't know enough about the handling with ORCAD, LTspice or MICROSIM to compare the results.
All the best
Andreas Kirschner
Four unbal. power amp bal. versions based of Pass "X" (SuSy) with sim. results (Fig. 3 from US patent 5376899)
hallo Mr. Pass
here the results of simulation, now for bal. versions. Please note, that I must find solution for GND by simulation. Now there are "actually GND" and "GND only for simulation" by all 4 versions. Please not in additional, that the basic circuit is that of Fig. 3 of US patent 5376899, but device numbers are from Fig 1 respective from first fig. on page 1 of this one. From Var. II and III I don't get the spectrum of IM. Instead of this I get the message
"simulation errors occured" - "warning: line too long"
Evaluation, analysis and compare to unbal versions I will make the next days.
A: Pass X bal Fig3 Var I
Gain=10x: 40=100R 36/37=15K Open-Loop Gain: 28x In/out 1V/10Vss
1) schematic for Damping factor test
2) AC-Analysis (damping factor over frequency)
3) AC Analysis (normally frequency response)
4) Fourier Analysis lin 10 KHz (THD) K3: 4,2mV
5) Fourier Analysis log 10 KHz (THD)
6) Fourier Analysis lin 100 KHz (THD) (K3: 15,4mV)
7) Fourier Analysis log 100 KHz (THD)
8) Fourier Analysis log 19/20 KHz (IM)
9) Transient Analysis 1MHz
10) schematic
hallo Mr. Pass
here the results of simulation, now for bal. versions. Please note, that I must find solution for GND by simulation. Now there are "actually GND" and "GND only for simulation" by all 4 versions. Please not in additional, that the basic circuit is that of Fig. 3 of US patent 5376899, but device numbers are from Fig 1 respective from first fig. on page 1 of this one. From Var. II and III I don't get the spectrum of IM. Instead of this I get the message
"simulation errors occured" - "warning: line too long"
Evaluation, analysis and compare to unbal versions I will make the next days.
A: Pass X bal Fig3 Var I
Gain=10x: 40=100R 36/37=15K Open-Loop Gain: 28x In/out 1V/10Vss
1) schematic for Damping factor test
2) AC-Analysis (damping factor over frequency)
3) AC Analysis (normally frequency response)
4) Fourier Analysis lin 10 KHz (THD) K3: 4,2mV
5) Fourier Analysis log 10 KHz (THD)
6) Fourier Analysis lin 100 KHz (THD) (K3: 15,4mV)
7) Fourier Analysis log 100 KHz (THD)
8) Fourier Analysis log 19/20 KHz (IM)
9) Transient Analysis 1MHz
10) schematic
Four unbal. power amp bal versions based of Pass "X" (SuSy) with sim. results (Fig. 3 from US patent 5376899) continued
D: Pass X bal Fig3 Var IV
Gain=83x: 40=20R 36/37 = NC (15M) Open-Loop Gain: 83x In/out 120m/10Vss
1) schematic for Damping factor test
2) AC-Analysis (damping factor over frequency)
3) AC Analysis (normally frequency response)
4) Fourier Analysis lin 10 KHz (THD) K3: 11mV
5) Fourier Analysis log 10 KHz (THD)
6) Fourier Analysis lin 100 KHz (THD) (K3: 27mV) 200mVss input
7) Fourier Analysis log 100 KHz (THD) 200mVss input
8) Fourier Analysis log 19/20 KHz (IM)
9) Transient Analysis 1MHz
10) schematic
Unfortunately I must reduce the capacity of
pdf files from version II and III, therefore this comes later
D: Pass X bal Fig3 Var IV
Gain=83x: 40=20R 36/37 = NC (15M) Open-Loop Gain: 83x In/out 120m/10Vss
1) schematic for Damping factor test
2) AC-Analysis (damping factor over frequency)
3) AC Analysis (normally frequency response)
4) Fourier Analysis lin 10 KHz (THD) K3: 11mV
5) Fourier Analysis log 10 KHz (THD)
6) Fourier Analysis lin 100 KHz (THD) (K3: 27mV) 200mVss input
7) Fourier Analysis log 100 KHz (THD) 200mVss input
8) Fourier Analysis log 19/20 KHz (IM)
9) Transient Analysis 1MHz
10) schematic
Unfortunately I must reduce the capacity of
pdf files from version II and III, therefore this comes later
Four unbal. power amp bal versions based of Pass "X" (SuSy) with sim. results (Fig. 3 from US patent 5376899) continued
B: Pass X bal Fig3 Var II
Gain=28x: 40=100R 36/37= NC (15M) Open-Loop Gain: 28x In/out 355m/10Vss
1) schematic for Damping factor test
2) AC-Analysis (damping factor over frequency)
3) AC Analysis (normally frequency response)
4) Fourier Analysis lin 10 KHz (THD) K3: 12,3mV
5) Fourier Analysis log 10 KHz (THD)
6) Fourier Analysis lin 100 KHz (THD) (K3: 25mV) 680mVss for 10Vss
7) Fourier Analysis log 100 KHz (THD) 680mVss for 10Vss
8) Fourier Analysis log 19/20 KHz (IM)
9) Transient Analysis 1MHz
10) schematic
B: Pass X bal Fig3 Var II
Gain=28x: 40=100R 36/37= NC (15M) Open-Loop Gain: 28x In/out 355m/10Vss
1) schematic for Damping factor test
2) AC-Analysis (damping factor over frequency)
3) AC Analysis (normally frequency response)
4) Fourier Analysis lin 10 KHz (THD) K3: 12,3mV
5) Fourier Analysis log 10 KHz (THD)
6) Fourier Analysis lin 100 KHz (THD) (K3: 25mV) 680mVss for 10Vss
7) Fourier Analysis log 100 KHz (THD) 680mVss for 10Vss
8) Fourier Analysis log 19/20 KHz (IM)
9) Transient Analysis 1MHz
10) schematic
Four unbal. power amp bal versions based of Pass "X" (SuSy) with sim. results (Fig. 3 from US patent 5376899) continued
C: Pass X bal Fig3 Var III
Gain=10x: 40=20R 36/37=11K3 Open-Loop Gain: 83x In/out 1V/10Vss
1) schematic for Damping factor test
2) AC-Analysis (damping factor over frequency)
3) AC Analysis (normally frequency response)
4) Fourier Analysis lin 10 KHz (THD) K3: 1,65mV
5) Fourier Analysis log 10 KHz (THD)
6) Fourier Analysis lin 100 KHz (THD) (K3: 16mV)
7) Fourier Analysis log 100 KHz (THD)
8) Fourier Analysis log 19/20 KHz (IM)
9) Transient Analysis 1MHz
10) schematic
C: Pass X bal Fig3 Var III
Gain=10x: 40=20R 36/37=11K3 Open-Loop Gain: 83x In/out 1V/10Vss
1) schematic for Damping factor test
2) AC-Analysis (damping factor over frequency)
3) AC Analysis (normally frequency response)
4) Fourier Analysis lin 10 KHz (THD) K3: 1,65mV
5) Fourier Analysis log 10 KHz (THD)
6) Fourier Analysis lin 100 KHz (THD) (K3: 16mV)
7) Fourier Analysis log 100 KHz (THD)
8) Fourier Analysis log 19/20 KHz (IM)
9) Transient Analysis 1MHz
10) schematic
Pass X Var I (unbal) versus X Var V bal (only THD, 10KHz + 100KHz) US patent 5376899
Gain=10x In/out 1V/10Vss - continued
B. Pass X Var V bal Fig.3 THD 10 100KHz-.pdf (40 = 10R, Open Loop Gain = 125)
1) THD 10KHz
2) Schematic
3) THD 100KHz
good to observe is the THD compensation effect by balanced version, even by 100 KHz. Only third harmonic there are, that means at same time, there are no unwanted 1KHz signal by IM measuring by 19+20 KHz
At next I will investigate the complementary-symmetry (push-pull) version of Fig. 4 from US patent 5376899 both unbal and bal. versions
Let's see, whether it is possible to get a design based to this and to the CSPP topology (circlotron) with even more reduction of distortion.
Gain=10x In/out 1V/10Vss - continued
B. Pass X Var V bal Fig.3 THD 10 100KHz-.pdf (40 = 10R, Open Loop Gain = 125)
1) THD 10KHz
2) Schematic
3) THD 100KHz
good to observe is the THD compensation effect by balanced version, even by 100 KHz. Only third harmonic there are, that means at same time, there are no unwanted 1KHz signal by IM measuring by 19+20 KHz
At next I will investigate the complementary-symmetry (push-pull) version of Fig. 4 from US patent 5376899 both unbal and bal. versions
Let's see, whether it is possible to get a design based to this and to the CSPP topology (circlotron) with even more reduction of distortion.
Nelson Pass said:Play nice, you boys.
Don't make me come over there.
Nelson
You are welcome to come over here anytime you like, do some salmon and trout fishing and walking in the mountains, maybe some deer hunting.
Bring JC as well, it wouldn’t harm him if he rested his eyes on something else than the scope.
Cheers
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.