About make an audio power amplifier.

Nikolai Petrenko said:

Hmm, I have thought about use FM to transmit audio but FM receivers are too complex, so I choose AM instead. New problem raising: MPF102 are not available. Will 2N5484 or 2N5486 or 2SK170 replace MPF102 in the circuit well?

Click to expand...

Why have you got a thread 12 pages long about audio amplifiers, then want to feed it the worst possible quality via AM?.

Even if you were going to use high quality FM it would still be well below the amplifier quality we've been discussing.

Nikolai Petrenko said:

Hmm, I have thought about use FM to transmit audio but FM receivers are too complex, so I choose AM instead. New problem raising: MPF102 are not available. Will 2N5484 or 2N5486 or 2SK170 replace MPF102 in the circuit well?

Click to expand...

Hi Nikolai,

The 2N5484, 2N5485, and 2N5486 are made with the same process as the earlier MPF102 but are selections, mainly for IDSS, so they would be suitable replacements for the MPF102.
The 2SK170 would probably work but it would not be ideal because it is designed for audio frequencies rather than radio frequencies (RF).

spec

DATA SHEETS
(1) MPF102
https://www.fairchildsemi.com/datasheets/MP/MPF102.pdf
(2) 2N5484, 2N5485, and 2N5486
https://www.fairchildsemi.com/datasheets/MM/MMBF5484.pdf

Nigel Goodwin said:

Why have you got a thread 12 pages long about audio amplifiers, then want to feed it the worst possible quality via AM?.

Even if you were going to use high quality FM it would still be well below the amplifier quality we've been discussing.

Click to expand...

No, I don't think so. Shouldn't I am listening to exciting programs in radio? Or use AM to transmit guitar signal to the amp, so this will eliminate problems about I can't go anywhere in the room without move the amp (long guitar signal cable are expensive).

My subwoofer filter, sorry because of bad drawing and arranging:
NEW UPDATE! C2 SHOULD BE 220nF

An AM radio pick up all kinds of buzzing, clicks and pops (amplitude variations) interference. A half-decent FM radio ignores amplitude variations. Wireless microphones used by performers use FM.

Your transistor T2 is biased wrong so it is clipping like mad! Also it has no negative feedback so its gain will be high at about 180 times and its distortion will be at least 30%!

audioguru said:

An AM radio pick up all kinds of buzzing, clicks and pops (amplitude variations) interference. A half-decent FM radio ignores amplitude variations. Wireless microphones used by performers use FM.

Your transistor T2 is biased wrong so it is clipping like mad! Also it has no negative feedback so its gain will be high at about 180 times and its distortion will be at least 30%!

Click to expand...

Yes, I was wrong when calculate current throw R9 and the result is bias current is too hell (probably press the numpad of calculators wrong or ..... ).

Connect the output and input of T2 with 10K resistor ???

audioguru said:

An AM radio pick up all kinds of buzzing, clicks and pops (amplitude variations) interference. A half-decent FM radio ignores amplitude variations. Wireless microphones used by performers use FM.

Click to expand...

But FM receivers are too complex (except the creepy regen ones) and hard to make they quality good.

Nikolai Petrenko said:

Connect the output and input of T2 with 10K resistor ???

Click to expand...

Why? Series or parallel?

Why are you loading the emitter-follower of T1 with the very low impedance of T2? It causes T1 to produce distortion. Why use distorting transistors instead of using excellent opamps?

Or swap the T2, put it before C5?

I also have a question: A heatsink for TO-3 with size 8,5x25cm like the below

Can it handle 4 power transistors (TIP35/36) in Rod Elliott P3A amp? Less or more?

Nikolai Petrenko said:

Or swap the T2, put it before C5?

Click to expand...

That shows another problem with your lowpass filter. It must be fed from a very low impedance like the output of an opamp, not from the two 10k resistors.

Rod Elliot says his P3A will be fine for hi-fi use (full power only occasionally) if its heatsink is rated at 0.5 degrees C per Watt. Look at a heatsink catalog to find one like the one you have. With an output of 100W into 8 ohms, each output transistor dissipates about 27W.

audioguru said:

That shows another problem with your lowpass filter. It must be fed from a very low impedance like the output of an opamp, not from the two 10k resistors.

Click to expand...

You has said "low impedance buffer", why not?

audioguru said:

Rod Elliot says his P3A will be fine for hi-fi use (full power only occasionally) if its heatsink is rated at 0.5 degrees C per Watt. Look at a heatsink catalog to find one like the one you have. With an output of 100W into 8 ohms, each output transistor dissipates about 27W.

Click to expand...

Seem be the heatsink above can hold 4x2N3055/2955 so I think it can hold 4 transistor in P3A amp at normal hi-fi application (and can like instrument PA with full-speed cooler?)
How hot of the amp is too hot when room temp. is 25deg.C? I think above 60 deg.C is not acceptable.

Yes, the lowpass filter must be fed from a low impedance buffer unless you calculate the source impedance (the 10k stereo to mono resistors plus their driving circuit) into the filter design. I do not think you designed the filter to be a 3rd-order Butterworth since it looks like a 2nd-order Butterworth with another RC at its input which ruins the Butterworth sharpness. They should all be designed together to be a 3rd-order Butterworth.

If the thermal resistance of the heatsink is 0.5 degrees C per Watt (your heatsink might not be that good) and you feed it 27W x 4 then its surface will reach 54 degrees higher than the ambient temperature. BUT the transistors need insulators that are also 0.5 degrees C per Watt so the surface of the transistors will reach 108 degrees C above the ambient temperature. BUT the chip inside each transistor is much hotter than its surface, about 0.5 degree C per Watt so each chip will reach 216 degrees V above the ambient. Their absolute maximum allowed temperature is 150 degrees C so they will melt if the amplifier continuously produces 100W for two channels.

Luckily, music and voices produce full power only occasionally.

Please point out any error (about earthing or junctions) if they exist:

Hy Nikolai,

Pretty good attempt- at least you are thinking about the order of connections, but not quite right. Just copy the earthing and supply line connections in this schematic- which is from the original 'Transistor Equivalent' thread (post #456).

I see you are using EAGLE ECAD now.

spec

Yes, my earthing too overlap.

Now you are talking Nikolai.

Just one small point: R4 left should go to 0V separately (as R9 on my schematic). Nothing should affect the long tailed pair inputs. Q3 is a constant current generator providing the tail current (2 mA if LED VF is 1.8 V) of the long tailed pair input transistors (Q1 & Q2). It operates at DC and does not amplify the audio signal in any way, quite the reverse in fact.

The long tailed pair (Q1 & Q2) is a difference amplifier. It's job is to take the audio signal on one input (Q1) and an attenuated version of exactly what the speaker signal is at its other input (Q2) and produce a difference current to drive the voltage amplification stage (VAS) (Q4). Even a milivolt extraneous signal intruding on this process can degrade the overall distortion of the amplifier. This is especially the case for class AB amplifiers where the currents and fields are non linear.

Also, perhaps put in thick traces as on the original schematic to show heavy duty conductors.Thick conductors are vital for the high current paths of the circuit.

Only an additional minor point, but bootstrap capacitor C5+ should connect to the junction of R12 & R13 separately rather than to the output line. Just below the output line connection would be good on your schematic. The bootstrap capacitor introduces distortion and should be kept away from the output line.

Finally, almost, an independent heavy duty output line for the speaker 0V should connect to the star point at C7 lower.

'Gnd' is not a correct designation. It should be 0V throughout your schematic. Ground is a separate connection.

spec

Thank a lot about your teaching!
So Q1 and Q2 must be near perfect good (has the nearest parameter to the other)?

Ah, I have a question (may be not clear):
What kind of solder should I use? 63/37 or silver lead-free?