2 stage amplifier - any advice welcome

[boylesg]

I have implemented the amplifier in the short circuits magazine below but with modifications. The first stage is as per my previous successful class A amplifier.

If I put the output of this amplifier through the set of PC speakers I am using, then I have to turn the volume knob right down to just above zero.

It runs of 12V rather than 6V - I forgot to change the DC source in the circuit below.


With my multimeter I seem to have confirmed that I can vary the output impedance of the second stage over quite a wide range - ~0.2mA - 1A going through the speaker.

So it means that I can solder up this general purpose amplifier and adjust the output impedance up to suit the control pin of a 555 but still be able to use the amplifier for any other purposes requiring different out put impedances.

Are there any improvements people suggest I make or any problems that I might not be anticipating?

How critical are the capacitor values with amplifiers? The software TransistorAmp specified that I should use a 470uF on the output of the class A amplifier. But since I have added a second stage, I removed the output capacitor for this first stage, followed the example in the circuit below and used a 100uF capacitor on the output of the second class AB stage. It sill seems to work just fine.

 

[BradtheRad]

How critical are the capacitor values with amplifiers? The software TransistorAmp specified that I should use a 470uF on the output of the class A amplifier. But since I have added a second stage, I removed the output capacitor for this first stage, followed the example in the circuit below and used a 100uF capacitor on the output of the second class AB stage. It sill seems to work just fine.

Yes, that is called direct coupling between the pre-amp and the power amp. It works because you are alternating an NPN with a PNP.

The large capacitor going to the speaker is for DC blocking. It will acquire a charge, while passing AC to the speaker. It should be large enough that its charge does not vary much, even down to 20 Hz (or whatever frequency you want as your lowest bass response).

Looks as though Q4 is used as a variable resistor. Or a current source. Right? Is it your plan to adjust it so it brings Q2 and Q3 to just the right turn-on point, for minimum distortion?

Does R4 figure into the same plan? Haven't seen that method of pulling biasing current through the speaker.

 

[chuckey]

"e output impedance of the second stage over quite a wide range - ~0.2mA - 1A", no this is wrong, what you are doing is changing the QUIESCENT current through the output tyransistors, set it to 5mA and it will work fine. If you leave it at 1A, you will over heat the transistors and flatten the batteries (if used). The output impedance is the effect when you are driving a low impedance load with a high impedance amplifier, the output voltage seems low, because it is all being dropped inside the amplifier.
This is what I do not understand, the input of a working NE555 trigger is .1mA @ 5-15V, so it is at the minimum 50K, so you do not need a power amp to drive it. You need a reasonable amount of volts, going from a good earth potential to the rail voltage. So if you want to use the amplifier in your circuit, replace the loudspeaker with a 1k resistor, and feed the live signal (not rail ) end of it through a 2mf capacitor to a 10K resistor to earth. Connect the junction of the 2MF cap and 10K to the trigger.
Frank

 

[FvM]

...but with modifications

No change for the better.

 

[boylesg]

Yes, that is called direct coupling between the pre-amp and the power amp. It works because you are alternating an NPN with a PNP.

The large capacitor going to the speaker is for DC blocking. It will acquire a charge, while passing AC to the speaker. It should be large enough that its charge does not vary much, even down to 20 Hz (or whatever frequency you want as your lowest bass response).

Looks as though Q4 is used as a variable resistor. Or a current source. Right? Is it your plan to adjust it so it brings Q2 and Q3 to just the right turn-on point, for minimum distortion?

Does R4 figure into the same plan? Haven't seen that method of pulling biasing current through the speaker.

Q4 and R3 are meant to be a rubber diode so that I can vary the voltage bias on the complementary pair. When I implemented the original short circuits version with just the diode the sound coming from the speakers was very poor. But as soon as I replaced it with the rubber diode and played around with the pot, the sound coming from the speakers dramatically improved. So from now one I will stick with 'rubber' diodes for flexibility.

If you look at the description of the original 'short circuits' magazine version of the amplifier, R4 = 820R and is meant to bias the complementary pair. So replacing it with the 10K pot means that I can vary the output current of the secondary stage AB and perhaps drive larger speakers or a larger load of some other description if I wish. And I confirmed that this works quite well with my multimeter.

- - - Updated - - -

"e output impedance of the second stage over quite a wide range - ~0.2mA - 1A", no this is wrong, what you are doing is changing the QUIESCENT current through the output tyransistors, set it to 5mA and it will work fine. If you leave it at 1A, you will over heat the transistors and flatten the batteries (if used). The output impedance is the effect when you are driving a low impedance load with a high impedance amplifier, the output voltage seems low, because it is all being dropped inside the amplifier.
This is what I do not understand, the input of a working NE555 trigger is .1mA @ 5-15V, so it is at the minimum 50K, so you do not need a power amp to drive it. You need a reasonable amount of volts, going from a good earth potential to the rail voltage. So if you want to use the amplifier in your circuit, replace the loudspeaker with a 1k resistor, and feed the live signal (not rail ) end of it through a 2mf capacitor to a 10K resistor to earth. Connect the junction of the 2MF cap and 10K to the trigger.
Frank

Well I don't need a power amp for the 555. But my idea was to try and create a generalised adjustable amp that I can use for other purposes as well if I wish.

So is the following correct: I am 'amplifying' the quiescent current but not really acheiving anything with respect to the audio signal?

If so how would you go about making the output impedance of the second stage adjustable then?

Like this?



It is my current understanding that high impedance = high resistance = low signal current = lower power/dB and low impedance = low resistance = high current = high power / dB

 

[boylesg]

So if you want to use the amplifier in your circuit, replace the loudspeaker with a 1k resistor, and feed the live signal (not rail ) end of it through a 2mf capacitor to a 10K resistor to earth. Connect the junction of the 2MF cap and 10K to the trigger.
Frank

After reading this again you mean to do this:



If I was to replace the 10k resistor with a pot as shown, would this work as a variable output impedance amplifier that I could use for other experiments apart from the 555?
Efficiency and similar issues aside. E.G. Could I reduce the 10k pot and hook this back up to the PC speakers? Perhaps I would need to jumper the 1k resistor

 

[boylesg]

I think the following would be a far easier and more practical way for me to implement a flexible amplifier for use with my 555 or with any other experiment later on:

So there would be two connectors on the board - one from stage 2 class AB and one directly from stage 1 class A. The coupling between stage 1 and 2 would then have a jumper of some desription so that I can choose whether or not to use stage 2.



With the stage 1 + stage 2 amplifier, with my mulitimeter, I am still only measuring current peaks of the order 20-30 uA.

Am I only measuring the quiescent current through stage 2 here and the varying signal current is invisible in DC mode?

 

[FvM]

Unfortunately, every modification is introducing different faults and changes for the worse related to the original "short circuits magazine" design.

In your initial version, R6 is dwarting the bootstrap load concept and constant class AB bias.

In post #5, the class AB bias current is removed by connecting R5 to ground, in post #7 R5 is omitted, giving distorted class B operation.

In post #6, you are trying to adjust gain by shorting the amplifier output.

I don't want to further comment about using a class AB power amplifier to drive a high impedance 555 pin. I already suggested a class A amplifier in a previous thread.