Electronerd

do you have any passive input devices at all.

If you can master designing a audio amp, using passive devices, first, than you'll have a better understanding of what is needed to couple a active device to an amp stage.

That's why I use two speakers, for input and output.

That gives enough design challenge in it's self.

Hi hobbyist,

I'm getting ahead of myself, aren't I? I will focus on passive devices from now on, then multiple stages. I'm afraid I haven't had much time to work on Attempt #3 due to school, but fear not, it's coming!

Austin

 

Don't use a speaker as a microphone because its resonance will cause a very boomy sound.

A power amplifier has an extremely low output impedance that damps the resonance so that it is barely heard.

Use an electret microphone.

 

Austin,

I am NOT negating what audioguru, posted,

But for a very basic audio amp, using 2 passive devices, input and output speaker, works very good to give a headstart in understanding the elementary design of a audio amplifier.

My designs are elementary, but they are good enough to give encouragement to proceed farthur with your own design.

So here is a audio amp, I built and tested it, if you build it to the schematic, it should work properly, this way you can get a hands on experiance and have something to work with for your own improvements.
-----------------------------------------------------------------
In both circuits, Q1 and Q2 have NO voltage amplification,
they are used ONLY to raise the input impedance high enough for the amplifier stage Q3, and Q4, to be able to work into it.

Because were taking a very low output impedance 8 ohms, and matching it to the amp, with enough gain to amplify, so the amp output needs to be high enough resistance to be able to get a proper gain with the ratio of (RC / RE) in the AC gain.



This first one has a volt. gain of around 7.
If you put your ear up to the output spkr. and speak normal voice, into input, it will soumd about as loud as a telephone conversation.



This second one is VERY sensitive, again put the output close to your ear,
but whisper into the input, because it is so sensitive, that a whisper will be heard very clearly, if you talk into it, hold the input spkr. away about 6 inches from your mouth, because anything above a whisper will overdrive it.

If you made R13 into a 1K pot, then you can adjust the sensitivity readily.



Notice the input is referenced to positive supply, that's because a PNP is at the input stage, and it's emitter is positive going.

So input to this stage base emitter ne3eds to be positive, referenced.

 

I think Hobbiest has an extremely hot transistor. The datasheet for the 2N3904 shows it working poorly at 100mA and above but here it has a current of 152mA.

This is a class-A amplifier when a class-AB pair should be used.

 

Hi,

It's a little hot, but it's within the manufacturers tolerance allowed, 200Ma. continuous.
When I designed it, I looked at the Max. allowable, continuous collector current, from my data sheet, and chose a 75% max. IC allowable, for my value of quiescent current.

 

Hi,

It's a little hot, but it's within the manufacturers tolerance allowed, 200Ma. continuous.

But its datasheet shows that it survives 200mA but performs poorly at 100mA and more.

I was talking about its heating. Your transistor is almost at its max allowed temperature.
Its output power into the speaker is very low.

 

Thanks audioguru.

Heres one I just designed, and built, and tested, see if this is any improvement.

 

Heres one I just designed, and built, and tested, see if this is any improvement.

I simulated it. I show its assymmetrical clipping when its input is too high at 7mV peak. When its input is reduced to 6mV peak then its output has fairly low distortion and the output power is a whopping 0.01W. That is 1/100th of a Watt!

If the same 3 transistors are made into a class-AB amplifier then the output power will be 45 times more (0.45W).

 

Thank you hobbyist and Audioguru for all your advice and information, I truly appreciate it!

@hobbyist: Your audio amplifier is very interesting to me, and the way you add multiple stages is what's interesting. I am remembering that The Electrician said that I could possibly approach a voltage gain of 100, which seems impossible to me considering the "Static Characteristics" graph (Pg. 2, Figure 1), but perhaps you could likewise? Imagine having multiple stages each having close to a voltage gain of 100, and then using a last Common-Collector stage to convert the impedance to a lower value. We can both work up to that goal, eh?

@Audioguru: Thank you very much for the notes you made about the 2N3904 specs. Originally I wasn't sure how to bias the 2N3904 since it didn't seem to have a "Static Characteristics" graph like the BC547B. However, upon further examination I saw the first graph on page three indicated those same conditions. Once I bias the BC547B efficiently, I will move on to the 2N3904 and so on. I also calculated the power dissipation of the new Attempt #3 and it's 130mW; far below the maximum rating (625mW).

@Everyone: I've been working on Attempt #3, but I'm having trouble with achieving a voltage gain of 100. I've improved my design to a gain of 4 but that isn't a significant change. The only solution I can think of without transforming the BC547B transistors biasing conditions would be to add a emitter capacitor across Re. I have a question concerning the proper resistor value in series with the cap; firstly, by using the following equation:

\(C=\frac{1}{2 \pi f X_c\)

hobbyist said that to find the proper value for the resistor in series with Ce, I do \(X_c=\frac{R_1}{A_v}\) which would be \(X_c=\frac{300k}{4}=75k\). That seems far from correct, since the value is so high. I thank everyone for their contributions regarding the matter and hope you post in the future!

Austin

 

Austin, this will probably complicate your life, but my intent is the opposite.
Datasheets are good, and you should always review the datasheet before you use a part for the first time.

IMHO, to design a hobbyist's audio amp using a general purpose, small-signal transistor, you don't need a datasheet. All you need to know is that Vbe≈0.7V, Ic≈Ie, and Ib<<Ic (β>100). It is important to make sure that the breakdown voltage Vceo is comfortably higher than Vcc. If you design an amp using BC547B, and another using 2N3904, and another using 2N2222, and..., ad nauseum, you will find that nothing has to change except the transistor part number.

When you get to designing power output stages, the datasheet becomes much more important. If you are designing amps with bandwidth much greater than 20kHz, the datasheet becomes important. If you're designing high-speed switching circuits, the datasheet is important. For what you're doing, generally not.
I'm just trying to save you some time.

 

Austin, this will probably complicate your life, but my intent is the opposite.
Datasheets are good, and you should always review the datasheet before you use a part for the first time.

IMHO, to design a hobbyist's audio amp using a general purpose, small-signal transistor, you don't need a datasheet. All you need to know is that Vbe≈0.7V, Ic≈Ie, and Ib<<Ic (β>100). It is important to make sure that the breakdown voltage Vceo is comfortably higher than Vcc. If you design an amp using BC547B, and another using 2N3904, and another using 2N2222, and..., ad nauseum, you will find that nothing has to change except the transistor part number.

When you get to designing power output stages, the datasheet becomes much more important. If you are designing amps with bandwidth much greater than 20kHz, the datasheet becomes important. If you're designing high-speed switching circuits, the datasheet is important. For what you're doing, generally not.
I'm just trying to save you some time.

Thanks Ron, I understand and appreciate the advice, but I do have time; plenty in fact (I'm a youngster). I like the fact of meeting the proper needs for each transistor I design with, and I am getting faster at biasing them now that I know more about the concept. Once I move on to the other classes of amplification (Class-AB, Class-B), I will most likely be needing the datasheet. Furthermore, I know that in the future I'll be designing projects that require more bandwidth and high-speed switching applications, so I think it's best to get in the habit of analyzing the datasheet now.

I "elmer" or teach kids to obtain their amateur radio license at my local radio club and in the future, once I master Class-A amps, I'm going to teach the kids how to design their own Class-A amp and provide materials for them. There are many kids that are interested in just obtaining their license but I want to provoke an interest in electronics as well.

Ron, thank you very much for all your contributions, you have certainly quenched some of my curiosity. I'm sure everyone out there will get tired of me because after transistors come op-amps!

 

Hello Everyone,

I've been reading more about transistors, and in a recent Amateur Radio Handbook (2007) they say that a transistor should have no more than 25dB of gain per stage (Av=18). Otherwise, the Class-A amp will become more like an oscillator rather than an amplifier. Nevertheless, I assume I can still add the cap in parallel with Re to increase the gain? This seems as though it would increase the gain without degrading the performance of the amplifier. In Ron's transistor circuit with 40dB of gain, is that high of gain feasible without making the transistor act like an oscillator?

Thanks,

Austin

 

Kids who make circuits on breadboards make oscillators instead of amplifiers because the breadboard has such a high capacitance between rows of contacts that makes positive feedback and oscillation.

If you use a compact design of a pcb then the circuit should work perfectly.
Like all circuits, the supply needs a supply bypass capacitor.