I have now built the circuit according to "crutschow" and replaced the two 1 ohm
resistors with 10 ohm resistors. Transistors are now running cold. The current draw
for the circuit has stabilized at 13.2mA and the voltages are as follows:

Q1 Collector 11.93 V
Q1 Base 8.36 V
Q1 Emitter 7.72 V
Q1 VCE 4.21 V

Q2 Collector 0.00 V
Q2 Base 7.00 V
Q2 Emitter 7.59 V
Q2 VCE 7.59 V

Q3 Collector 7.00 V
Q3 Base 1.35 V
Q3 Emitter 0.71 V
Q3 VCE 6.29 V

When I feed a 500mV Sine signal, the output sounds okay but the output is a
weird sine on the scope. The circuit also quits working after a while if I have
the output scoped.

I thought that the Vce's should be the same for the power transistors. Such a
large difference would account for the scope wave being weird (like 1 sine
wave superimposed on the other), wouldn't it?

John​

 

Sounds like a standing wave, but I'm not really that good at AC amps.

 

The quiescent Vce is set by the bias circuit. If you increase R3 it will lower the operating point. If you want to part things midway, then you need R3 to drop about 5.3 V. With R4 = 100 Ω, there should be about 7.1 mA flowing in it. This should be about what is flowing in R3, so you need it to be about 750 Ω.

 

The quiescent Vce is set by the bias circuit. If you increase R3 it will lower the operating point. If you want to part things midway, then you need R3 to drop about 5.3 V. With R4 = 100 Ω, there should be about 7.1 mA flowing in it. This should be about what is flowing in R3, so you need it to be about 750 Ω.

I will try changing the resistor to 750 ohms.

Thanks. You folks are such a great help.

 

Hi John, et al.

Since any significant change in the value of R3 will likely alter other aspects of the Bias circuit, it might be easier to throw in a 1K pot - initially set at ~ 750 ohms - and tweak as necessary until bias is balanced. Then measure pot resistance (removed from circuit of course) and replace with nearest fixed value available.

Just my 2 cents...

 

replaced the two 1 ohm
resistors with 10 ohm resistors

That will significantly reduce the power to your 16 ohm load.

 

That will significantly reduce the power to your 16 ohm load.

Yes, it will. He did so at my suggestion in order to demonstrate that emitter resistors could solve the thermal runaway problem. It was noted that using resistors that size would cause undesirable impacts on performance. The next step would be to find suitable lower values that are an acceptable compromise.

 

I believe gluing the diodes to the output transistors will also significantly help.

 

Hi :

One more question. How would I compute the input and the output
impedance of the two stages. Also, how would I compute the gain
of the two stages and the ouput power?

Thanks so much I have learned a lot.

John

 

Hi :

One more question. How would I compute the input and the output
impedance of the two stages. Also, how would I compute the gain
of the two stages and the ouput power?

Thanks so much I have learned a lot.

John

Uh-oh definitely starting to sound like homework
SG

 

Uh-oh definitely starting to sound like homework
SG

You are definitely full of it. I am a 63 y/o retiree trying to learn electronics
as it has always fascinated me!

 

You are definitely full of it. I am a 63 y/o retiree trying to learn electronics
as it has always fascinated me!

Geez dude lighten up didn't you see the smiley face!

 

Hi
Good for you! But I'm afraid the answer - even a simplified answer - is not easy. The first stage is an 'emitter degenerative amplifier'.The lack of a large capacitor across R4 makes all the difference and the input resistance is roughly R4 * hfe (the current gain - which will be at least 100) so we have 100 * 100 = 10k and this is in parallel with the 15k in parallel with the 100k so about 6k.
The output transistors are working as 'emitter followers'and the output resistance is very low.
You need to understand the h-parameter model really to get to grips with this, or any other ac transistor model. I'm afraid the EDUCATION section here is no help - you need a good textbook and some patience - it takes a bit to get your head round some of the concepts! Good luck though.
Colin

 

One cent from me: the 1n4148 looks rather similar to some other diodes like 1n4007 or even Zener diodes with some 2,7...3,6 V(stab). If the voltage on diodes is greater than 2x0,7V You`ve got a trouble. Begin with check up this voltage.
Next cent - impedance is more simple to measure than calculate if there is any loop.
However I didn see it at this case except the 100 Ohm.
Therefore take a 100 Ohms and multiply to average beta..... so the first cascade input impedance will be around 5 kOhm in parallel to R(b-gnd) resistor (I mean 15 k and 100 k both) or anyway around the 4....5 kOhm. Yes, its true that so prost model underestimates the transistor "inside" impact [I mean the r(e)=0,026/i(e)], but is is clear minority value here what may with an ease be neglected.

 

One cent from me: the 1n4148 looks rather similar to some other diodes like 1n4007 or even Zener diodes with some 2,7...3,6 V(stab).

Not at all correct.

A 1N4148 and a 1N4007 are only vaguely related. A 4148 is a high speed signal diode with low junction capacitance, while the 4007 is a 1 A rectifier diode with large capacitance and very poor response time. It is not recommended for 400 Hz power supplies, and is horrible in audio circuits.

And, neither of those diodes are anything like a zener diode, with very different thermal characteristics.

ak

 

Also the 1N400x is significantly physically larger than a 1N4148.

 

Why is this thread a continuation of another thread with no schematic, no parts list and no link??
The low 12V supply and 16 ohms speaker suggests puny output power (0.75W?) from a very simple circuit.

 

Why is this thread a continuation of another thread with no schematic, no parts list and no link??
The low 12V supply and 16 ohms speaker suggests puny output power (0.75W?) from a very simple circuit.

Which thread is that? There are a couple of schematics posted with component values.
SG

 

I didn't see page one.