simple BJT audio amp experiment (need help)

Wendy

Joined Mar 24, 2008
23,415
That dynamic mic is little more that a short. If you adjust the variable resistor to zero something, maybe more than one component, will smoke. It is amazing the number of components you can take out like that, the pot, the mic, the transistor.

Since the gain of the transistor is such a variable, we're using the variable resistor to find the sweet spot where the transistor is a linear device.

Yes, you can use an 8Ω speaker as a mic. Intercoms use a simple switch, as switch the roles of the mic/speaker. The single transistor design basically sucks though, you can make it work, but you will never be happy with it.

The carbon mic sucks less, you will get recognizable results, but it still won't be good.

A dynamic mic and speaker are extremely similar, both use voice coils and magnets. A speaker has a much bigger cone, to produce more volume, while a mic has a small plastic diaphram, for fidelity. The basic physics is pretty similar too.

When it comes to mics, there are exceptions (as with speakers). The technology keeps changing, but I'll leave it at that.

If you do what I did, get a couple of cheap 8Ω speakers for your junk box.

If you want to run some real experiments with audio amps, the web abounds with plans. You can get good results with as few as two transistors, I've seen some people around here with them, including their own designs.

With chips it gets even simplier.
 
Last edited:

Thread Starter

count_volta

Joined Feb 4, 2009
435
LOL Murphy's law indeed.

I already tried umm connecting the speaker without a resistor, and I saw no smoke and perceived no burning smell, I have to assume my speaker, transistor, and etc still work until I discover otherwise. ;)

So a regular mic has almost no resistance eh. I guess if a carbon mic is more sensetive to sound I will get one.
 

Thread Starter

count_volta

Joined Feb 4, 2009
435
Ohhh thank you. Well as long as I understand the NEED for 3 transistors and capacitors and what they do to overcome real world limitations I don't mind getting complicated.

What I think I'm going to do is build the circuit that Bill Marsden posted earlier and just barely have it work and see how much the quality sucks.

Only then can I appreciate your circuit and I will try to understand and build it. So I'm going to build 2 amps then. A theoretical crappy one and your good one. I want to do this to see the difference.

On another note....

I looked online and its hard to find a carbon microphone. Would a cheap microphone also work Bill? Can you reccomend one?
 

millwood

Joined Dec 31, 1969
0
let's me just say that you put a lot of efforts into it.

a few questions for you to think about.

1) what is the input impedance of each stage and how does that impact your frequency response.

1.a) the input impedance of the 1st stage will be dominated by RB5, which is very low. That means attenuation losses.
1.a) the input impedance of the 2nd stage will be dominated by RB3, which is extremely low, and will create (ac) loading problems for the 1st stage.

there are a few problems like that.

2) you do NOT want to put a 0.1uf capacitor between any transistor's collector and base. unless you purposefully want to listen to muddy sound.

3) you do NOT want to load your last stage with an 8ohm speaker. No way.

4) if you have any respect for fidelity, do NOT ac couple multiple stages.

5) at this rate, your gains will be so significant that you can afford to introduce a little bit of negative feedback to improve quality. and you don't want to do that with ac coupled stages.

6) I do suggest that you read a few books about audio amps. Self is a good starting point, as is Sloan's book (more practical).

good luck.
 

millwood

Joined Dec 31, 1969
0
if you want, this is a jlh1969b power amp. class A operation throughout, high-fidelity, and model'd after the Williams tube amps.

it is not difficult to figure out how it works. you can read lots of about it on the net.

the 2nd one is a jlh1969m (mosfet). my design, and it can be class A or class B, depending on bias and will slide seamlessly into class B if it runs out of bias.
 

Attachments

Audioguru

Joined Dec 20, 2007
11,248
The Jameco mic is an electret type. It needs to be powered and has an output of only 10mV so needs an amplifier with a voltage gain of more than 200.
 

millwood

Joined Dec 31, 1969
0
a simple answer to your question is no. an electret mic, like the jameco one, needs DC bias, which you do NOT have in that circuit.
 

Thread Starter

count_volta

Joined Feb 4, 2009
435
Allright, then what would you guys reccomend? Link me to the place where I can buy it if possible.

Like I said before I just want to get Bill's simple circuit to work period, even if it sucks and I get a ton of noise and the sound is very weak. I just want to hear something. Only then will I appreciate a proper amplifier like the hobbyist has posted.

Here is Bill Marsden's proposed circuit that will just barely work, which is what I want.



Carbon mic's are hard to find. I looked.
 

millwood

Joined Dec 31, 1969
0
Like I said before I just want to get Bill's simple circuit to work period, even if it sucks and I get a ton of noise and the sound is very weak. I just want to hear something. Only then will I appreciate a proper amplifier like the hobbyist has posted.
I don't know if you can ever make that particular circuit work with an electret mic.

but here is one if you want. it is a typical mic preamp - not suited to drive a speaker.

if you just want to hear something, you can build a pretty power mic power amp out of a lm317, a few resistors and a couple capacitors, with your electret mic.
 

Attachments

Thread Starter

count_volta

Joined Feb 4, 2009
435
millwood what are those capacitors for? I rather look for a microphone that can work with Bill's original circuit, since its so easy to understand.
 

millwood

Joined Dec 31, 1969
0
the dc bias is provided to the elecret mic via the left most resistor in my circuit. and the capacitor following the mic is to block the dc and allow the ac signal to be amplified by the transistor.

the two resistors provide proper dc biasing for the transistor, and the resistor on the emitter is to help define DC bias and to linearize the transistor. you will hear very weak signal, if at all from such an amp because it wasn't meant to drive a speaker.
 

Thread Starter

count_volta

Joined Feb 4, 2009
435
Did you select that particular circuit only because of the electret mic? There are many other mic's out there. I will buy the one that works.

Your circuit is not very difficult to understand either. Can you come up with some random values for the resistors and capacitor and I will try to simulate it.
 

millwood

Joined Dec 31, 1969
0
try the circuit on the right that bill gives and see if it will work.

electret mics send out a voltage signal and also a current signal: the jfet inside it is essentially a follower.

the voltage out is no more than 100mv so you want to use a small pot in serial with it: 1k is ok. the rail should be no more than 5v (3v is OK). if 1k isn't OK, you can adjust it downward to increase the gain.

to use a darlington as the transistor: a normal npn is OK.

and to be frank, I am less than 50% sure that this will work.

the issue here is that you need an amplifier with a high gain - very difficult to do with one stage where the load (the speaker) is of very low impedance. what I am trying to do is to use the electret mic in an unconventional way: use it as a current drive, and use a high beta npn / darlington.

but its sucess will depend on the mic, the npn you use and the power supply. so it is hard to predict with any level of confidence if it will work.
 

millwood

Joined Dec 31, 1969
0
Did you select that particular circuit only because of the electret mic? There are many other mic's out there. I will buy the one that works.
the coil types are very expensive and has even lower output than the electret.

electret types are easier for a preamp but your issues here are quite different.
 
Top