5
\$\begingroup\$

I am trying to design an amplifier for a 1 W 8 ohm speaker for educational purposes. The idea is to start from a common emitter amplifier to amplify voltage and then add a common collector (emitter follower) amplifier to amplify power and deal with impedance matching. I am a mechanical engineer and I am just learning about electronics so please bear with me.

The first step seems to work fine. I have sized the various components following this tutorial. I assumed a required gain of -10.

Enter image description here

Enter image description here

The model for the 2N2222 transistor is the following

.model Q2N2222 NPN(IS=1E-14 VAF=100
+   BF=250 IKF=0.3 XTB=1.5 BR=3
+   CJC=8E-12 CJE=25E-12 TR=100E-9 TF=400E-12
+   ITF=1 VTF=2 XTF=3 RB=10 RC=.3 RE=.2 Vceo=30 Icrating=800m  mfg=Philips)

Now I would like to add the second stage (common collector) and I came up with this:

Enter image description here

The model for the power transistor is (downloaded)

.MODEL D44H11_HD NPN(
+IS=2.14e-10 NF=1.271265 BF=208.89 RB=2 RBM=0.1 IRB=10
+VAF=342 NE=2.7349 ISE=1e-8 IKF=30 NK=0.9687
+BR=4 IKR=1.05 VAR=35
+XTF=1800 TF=1.9e-9 ITF=200 VTF=40
+CJE=1.4e-9 MJE=0.3092662 VJE=0.4723539
+CJC=175.527e-12 MJC=0.383595 VJC=0.479488
+TNOM=25 Vceo=80 Icrating=8 mfg=ON)

The choice of the 16 ohm resistor was dictated by the the input impedance for the common collector Zin = beta * RE_2 (is this correct?)

If now I add a load RL = 8 ohm after the Cout capacitor I have the following Vout and current across it:

Enter image description here

It is clearly not enough to drive the speaker. Of course I am missing a few things, but the purpose of this exercise was to start small and then build my way up.

  1. Can I actually build an (educational) amplifier with this approach?
  2. Do I need to increase the gain of the first stage and change the transistor?

I could buy a pre-designed kit of course, but what's the fun with that...

Updated circuit 1

Thanks to the suggestions I have updated the circuit replacing the Q2 transistor with a Darlington pair and reducing the common collector emitter resistance to 2 ohm. Maybe it is unrealistic, but I assume I can get it by using multiple resistances in parallel.

Enter image description here

And a much better outcome!!

Enter image description here

Updated circuit 2

After reading the various comments and answers, reasearching online and a lot of experimentation, I think I have made some progress. (Or got the same results with a much more complicated circuit :-))

I have made the following changes:

  1. Replaced second stage transistors with what I had available (2n3904 & bd139)
  2. Reduced the amplitude of the input signal to 100mV
  3. Increased output decoupling capacitor to 1000uF
  4. Increase RE2 resistor to 20 Ohm
  5. Added a bypass capacitor to the emitter of the first stage to increase the gain. This was causing oscillations in the power across the speakers
  6. Added bootstraps to both stages. I still need to understand how bootstrap works to better define the values of the resistors and capacitors, but it seems to increase the overall gain, and the oscillations of the output power
  7. Added feedback resistor Rfb. Based on this article.

enter image description here

enter image description here

enter image description here

Transistor models (from Onsemi)

.MODEL Qbd139 npn
+IS=1e-09 BF=222.664 NF=0.85 VAF=36.4079
+IKF=0.166126 ISE=5.03418e-09 NE=1.45313 BR=1.35467
+NR=1.33751 VAR=142.931 IKR=1.66126 ISC=5.02557e-09
+NC=3.10227 RB=26.9143 IRB=0.1 RBM=0.1
+RE=0.000472454 RC=1.04109 XTB=0.727762 XTI=1.04311
+EG=1.05 CJE=1e-11 VJE=0.75 MJE=0.33
+TF=1e-09 XTF=1 VTF=10 ITF=0.01
+CJC=1e-11 VJC=0.75 MJC=0.33 XCJC=0.9
+FC=0.5 CJS=0 VJS=0.75 MJS=0.5
+TR=1e-07 PTF=0 KF=0 AF=1
* Model generated on Feb 14, 2004
* Model format: PSpice

.MODEL Q2n3904 npn
+IS=1.26532e-10 BF=206.302 NF=1.5 VAF=1000
+IKF=0.0272221 ISE=2.30771e-09 NE=3.31052 BR=20.6302
+NR=2.89609 VAR=9.39809 IKR=0.272221 ISC=2.30771e-09
+NC=1.9876 RB=5.8376 IRB=50.3624 RBM=0.634251
+RE=0.0001 RC=2.65711 XTB=0.1 XTI=1
+EG=1.05 CJE=4.64214e-12 VJE=0.4 MJE=0.256227
+TF=4.19578e-10 XTF=0.906167 VTF=8.75418 ITF=0.0105823
+CJC=3.76961e-12 VJC=0.4 MJC=0.238109 XCJC=0.8
+FC=0.512134 CJS=0 VJS=0.75 MJS=0.5
+TR=6.82023e-08 PTF=0 KF=0 AF=1

It looks better to me. The power chart is not that smooth. I am not sure if it is an LTSpice issue or a problem.

A couple of notes:

  • The efficiency is also a bit better
  • The maximum power across RE2 has been reduced from 2.2W (all heat) to 1.5 W.
  • I had to tweak a few parameters and I want to understand a bit more how to calculate them.

Do you think I am on the right track? Now I am starting to look into current sources.

\$\endgroup\$
8
  • 2
    \$\begingroup\$ You will want your output stage to be 2-quadrant for the speaker. An emitter follower won't cut the mustard. (Not without crazy-stupid wasteful values for the emitter pull-down of your emitter follower stage, anyway.) Your emitter follower can pull up, actively. But the only thing you have to pull back down is that resistor. And it's not really strong enough for the job. (It's just a passive device, after all.) Finally, I'd recommend learning about bootstrapping -- for your 1st stage as well as the output stage (which won't be an emitter follower, I hope.) Good work, despite some issues! \$\endgroup\$ – jonk Nov 13 '20 at 8:24
  • \$\begingroup\$ The new darlington version seems to be able to feed 2 Ohm resistor. Everything it gets is dissipated and the massive idle DC through Q2 will make it easily hot, too hot without cooling. I quess your speaker - if not some miniature type - can stand few hundred mA DC current. Reduce dissipation by having the speaker in place of RE2. \$\endgroup\$ – user287001 Nov 13 '20 at 9:38
  • \$\begingroup\$ @user287001 How would I remove the DC component in that case? \$\endgroup\$ – Rojj Nov 13 '20 at 18:07
  • 1
    \$\begingroup\$ @Rojj Also look here and here. About \$1\:\text{W}\$ into \$8\:\Omega\$. Class-A, which is slightly less complicated. The LM380 is about 2.5 watts. (Also see here.) That should be enough for now? \$\endgroup\$ – jonk Nov 15 '20 at 7:07
  • 1
    \$\begingroup\$ @Rojj The 2nd link in my comment probably has most of what you need for class-A. The LM380 provides you with a class-AB arrangement (lower wasted dissipation) that is fairly typical. Best wishes!! \$\endgroup\$ – jonk Nov 15 '20 at 8:22
9
\$\begingroup\$

Yes, this is absolutely the right approach to take when learning about electronics.

In software, they have a mantra that it's easier to get correct code to run fast, than it is to get fast code correct.

Similarly with hardware, walk, before misguided people who have been there, done that, know it all, and have forgotten how hard it was at the beginning, suggest you run.

Your circuit will work. It won't work very well, it won't be very power efficient, but it will work, and the performance of the hardware should pretty much match that of the simulation.

It's by building a circuit like this that you learn why no commercial amplifiers use an RE_2. Once you understand what everything is doing, then replacing it first with a current source, and then the bottom half of push-pull output, will become meaningful. The progression to a push-pull output stage is a whole can of worms (biassing, cross-over distortion, thermal stability), so I'd save that for a lot later).

But before you do either of those to the output stage, there's a lot you can, and should, learn about basic amplifiers with what you have there. For instance, bypassing part of RE with a capacitor to change the first stage gain.

Happy learning. Take it slow. Don't try to run before you can walk, which may lead to discouragement.

\$\endgroup\$
1
  • \$\begingroup\$ Thanks! Very encouraging message. My next step was actually to learn and test the partial by-pass. \$\endgroup\$ – Rojj Nov 13 '20 at 9:00
3
\$\begingroup\$

You can implement an excellent class A amplifier (with no crossover distortion) this way.

The second stage (your buffer) must be a Darlington transistor, so its Rin becomes quite high (>> 10 kohm) and thus your gain of stage_1 is rather accurate at Rc/Re.

Another challenge is biasing.

I've often used DC_feedback from output emitter of the stage_2, back to the base of stage_1.

To do this, directly operate stage_2 base from collector of stage_1.

The key is a 2_resistor 1_capacitor DC_feedback from output emitter to input base.

Use two series resistors and a midpoint shunting capacitor.

  • Ground the capacitor (this will be polarized capacitor), (-) to ground; call the (+) the Vbias; use 100 µF as a starting value

  • remove your existing (two) biasing resistors on base of Q1

  • run a 10 kohm resistor from base of Q1 to Vbias node

  • run a 100 kohm resistor from the Vbias node to the output emitter

With the DC_feedback capacitor (100 µF) initially having no charge, the collector of Q1 will be initially at VDD, which causes a brief burst of high current through your Darlington transistor, and a loud pop from the speaker.

With Q1 needing about 1 milliampere collector current, and about 10 µA base current, there will be 1.1 volts drop across the (10 kΩ + 100 kΩ) feedback network. Thus the input base will be 1.1 volts DC lower than the output emitter.

You can adjust this as (if) you wish.

\$\endgroup\$
1
  • \$\begingroup\$ Thanks ! See first update of my question. Now I need a bit more time to digest the second part of your answer :-) \$\endgroup\$ – Rojj Nov 13 '20 at 8:53
2
\$\begingroup\$

As mentioned in the comments, indeed a single-transistor output stage to drive an 8 ohm speaker isn't such a good idea, it can be done and it will work it will give you disappointing performance (low volume and distortion when you raise the volume).

Instead I suggest that you build a "Class AB push pull" stage as described in many tutorials. Here's one that should be helpful.

..and deal with impedance matching.

That's a common misunderstanding among beginners. Even though audio amps often state that they have for example an "8 ohm output" that doesn't mean that they are impedance matched. What that means is that the stage is designed to drive an 8 ohm load. The voltage and current the stage can deliver is optimized for an 8 ohm load.

In reality an audio amplifier output is designed to have a very low output impedance so that it behaves like a voltage source. The Class AB push pull stage comes close to that behavior.

Impedance matching where input and output impedance need to match is usually only needed in high frequency circuits, not in audio circuits as the signal frequencies are much lower.

\$\endgroup\$
2
  • \$\begingroup\$ See my edit. I completely understand that AB push pull would be a better approach, but this simple case has allowed me to understand so much and I want to see how far I can take it. \$\endgroup\$ – Rojj Nov 13 '20 at 8:55
  • 2
    \$\begingroup\$ ..this simple case has allowed me to understand so much and I want to see how far I can take it. Excellent, it seems you "did your homework" which is always good. Indeed it is a good learning exercise to see how far the circuit you have will go and see its limitations. I agree with Neil_UK that this is the right approach. Many beginners here just want to "build something" instead of understanding why such and such a circuit is needed. But when you understand why electronics become easier and much more fun so keep learning! \$\endgroup\$ – Bimpelrekkie Nov 13 '20 at 9:00

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.