Might I suggest before you copy unknown amplifiers, you get a copy of Bob Cordell and Doug Self books on amplifier design and read up on it.
The very question " how does it sound" is misplaced unless you are tuning a circuit for specific defects to add "character" . With no global feedback, it is likely to have a lot of character. Besides, you did not even include what the devices are, so no way to evaluate the performance even for simulated transfer function and distortion.
The very question " how does it sound" is misplaced unless you are tuning a circuit for specific defects to add "character" . With no global feedback, it is likely to have a lot of character. Besides, you did not even include what the devices are, so no way to evaluate the performance even for simulated transfer function and distortion.
Might I suggest before you copy unknown amplifiers, you get a copy of Bob Cordell and Doug Self books on amplifier design and read up on it.
There is a reason we go to engineering school to understand these things.
Plus, it is not complete. It does not show the TWO power supplies this design needs. No label the Q1-8, no value in C1
I'm sorry if I sound harsh here but I am trying to help you. You are way over your head just based on your question.
Is it "reasonable"? Well that depends on your goal. Do you intend to design and etch a board? Have you ever matched transistors? Have you done simulations for stability margins? FWIW, it is not my favorite topology. I find the "standard" LTP input, Darlington VAS, thermally compensated spreader, with global feedback using Miller or two pole Miller compensation to be most practical.
There is a reason we go to engineering school to understand these things.
Plus, it is not complete. It does not show the TWO power supplies this design needs. No label the Q1-8, no value in C1
I'm sorry if I sound harsh here but I am trying to help you. You are way over your head just based on your question.
Is it "reasonable"? Well that depends on your goal. Do you intend to design and etch a board? Have you ever matched transistors? Have you done simulations for stability margins? FWIW, it is not my favorite topology. I find the "standard" LTP input, Darlington VAS, thermally compensated spreader, with global feedback using Miller or two pole Miller compensation to be most practical.
Your diode symbols are ambiguous - is that one 1N4148 or 3 in series? Its normal to draw 3 diodes as 3 diodes.
Hello, mark. Nice to see you again
D1, D2, D3 and D4 on the circuit diagram are respectively composed of three 1N4148 in series
D5, D6, D7 and D8 are four 1N4148 in series respectively
Is there any defect in this circuit
Dear Jacks,
Please do a small homework. Enter the circuit in LTSpice. Then, there is a procedure to feed music signal and the output can be heard on a loud speaker. There is a thread on that subject. You can easily search for it ! You will know immediately how the circuit sounds ! Voila !
Please do a small homework. Enter the circuit in LTSpice. Then, there is a procedure to feed music signal and the output can be heard on a loud speaker. There is a thread on that subject. You can easily search for it ! You will know immediately how the circuit sounds ! Voila !
I only know that LTSpice can simulate whether the circuit diagram works normally.
You mean LTSpice can simulate the sound of this circuit? I'll look up the subject. thank you very much.
To avoid global NFB, I would rework the output stage to be more like:
V+
|
Signal-->NPN
|
-[res]-
|
Vgnd-->PNP
|
DC servo<--|--> load
|
Vgnd-->NPN
|
-[res]-
|
Signal-->NPN
|
-[res]-
|
GND
Hopefully that's at least somewhat readable.
The top transistor can be NPN if it is used as a constant current source, or if the input signals are differential instead of common mode. Or it can be changed to PNP.
The middle 2 transistors stabilize Vce across the top and bottom transistors, greatly reducing IMD and HD.
Because this output stage has gain, the input stage/s can be greatly simplified.
V+
|
Signal-->NPN
|
-[res]-
|
Vgnd-->PNP
|
DC servo<--|--> load
|
Vgnd-->NPN
|
-[res]-
|
Signal-->NPN
|
-[res]-
|
GND
Hopefully that's at least somewhat readable.
The top transistor can be NPN if it is used as a constant current source, or if the input signals are differential instead of common mode. Or it can be changed to PNP.
The middle 2 transistors stabilize Vce across the top and bottom transistors, greatly reducing IMD and HD.
Because this output stage has gain, the input stage/s can be greatly simplified.
My English is terrible. I apologize. I don't quite understand your plan. But thank you very much!!
these seems more like a simplified circuit
for basic understanding or circuit analysis.
would not be very feasible amplifier
in real life.
for basic understanding or circuit analysis.
would not be very feasible amplifier
in real life.
I Am Lil Bit Confused With Power Supply Input Of Amplifier 58v Or 50v At Same Time???
This circuit diagram is of class a design. There's nothing wrong with the power supply voltage
Class A? No, I don't think that's feasible with 50V power supplies and only one pair of output transistors because the bias current and heat dissipation would be far too high for a practical amplifier. For safety, you would probably require at least four parallel pairs of output transistors and another pair as the driver transistors (plus a huge heatsink and power suppy).
It's important to understand that simulations may work fine, as they often do when they are only a schematic diagram under test. However, sims generally ignore real world conditions like heat dissipation and maximum power ratings of the actual components. Just because a transistor has a label of say, MJL3281, it may not affect the simulation at all if I substitute it with a small signal transistor labelled 2N5551, for example. In the real world, the result will be an instantly dead transistor and a bad start to any design. Sim. programs are usually blind to many real-world conditions and they're only useful to a tech or amateur who already knows how to go about design, component types and ratings to begin with. That's right, you need to understand how real components work and how to design within their ratings before you play with simulation programs. It's not all that difficult but it won't happen without extensive learning aided by respected textbooks. A bit of formal study would help a lot more than forum chat and unqualified schematics posted on the net.
Personally, I think it's great to doodle and experiment with simulated designs but its smarter to begin by building original or perhaps only slightly modified, existing and proven designs then build and test them to know what design elements should work and how it is that they work reliably up to their design ratings. Then you can confidently tackle your own versions, hopefully now confirmed in reality.
It's important to understand that simulations may work fine, as they often do when they are only a schematic diagram under test. However, sims generally ignore real world conditions like heat dissipation and maximum power ratings of the actual components. Just because a transistor has a label of say, MJL3281, it may not affect the simulation at all if I substitute it with a small signal transistor labelled 2N5551, for example. In the real world, the result will be an instantly dead transistor and a bad start to any design. Sim. programs are usually blind to many real-world conditions and they're only useful to a tech or amateur who already knows how to go about design, component types and ratings to begin with. That's right, you need to understand how real components work and how to design within their ratings before you play with simulation programs. It's not all that difficult but it won't happen without extensive learning aided by respected textbooks. A bit of formal study would help a lot more than forum chat and unqualified schematics posted on the net.
Personally, I think it's great to doodle and experiment with simulated designs but its smarter to begin by building original or perhaps only slightly modified, existing and proven designs then build and test them to know what design elements should work and how it is that they work reliably up to their design ratings. Then you can confidently tackle your own versions, hopefully now confirmed in reality.
Last edited:
The power design of this circuit is relatively small, so only a pair of output tubes are used