If it is a dynamic speaker (remember what I said about the physics and how many types there are.) it will work better than most mikes, but the frequency response won't be as good.

A condenser microphone is much harder to use. We are trying for simplicity (I thought). A carbon is much easier to use (which is why it was the standard for the phone company from before tubes were invented until LSI, large scale integration). A carbon mic really doesn't need much of an amp, while a condensor mic generates much smaller signals.

So did you get any readings with a DVM off this circuit? If you want I'll throw something together and test it along side, but electrit and condenser mics are close cousins, don't use them until you have something better to drive a speaker with.

The quickest way to tell if it is working is to put the mic next to the speaker. If you have acoustic squeal then it's happening.

I am using an 8 ohm speaker for the mic remember.

I am just really not sure how good of a microphone a speaker is, because think about it. The diaphragm of a microphone is made to respond to vibrations in the air (relatively weak force) while the diaphragm of a speaker is made to respond to a magnetic force (much stronger) .

Speakers were not made to be used as microphones.

I will try the circuit again and if it doesn't work we can worry about feedback and whatever else improves the gain of an amplifier. I should really finish reading the BJT transistor section of AAC.

 

I am using an 8 ohm speaker for the mic remember.

I am just really not sure how good of a microphone a speaker is, because think about it. The diaphragm of a microphone is made to respond to vibrations in the air (relatively weak force) while the diaphragm of a speaker is made to respond to a magnetic force (much stronger) .

Speakers were not made to be used as microphones.

I will try the circuit again and if it doesn't work we can worry about feedback and whatever else improves the gain of an amplifier. I should really finish reading the BJT transistor section of AAC.

You are starting from a false premise though. I've used speakers for microphones, both for practical and experimental use, they are good as or better than most microphones in terms of sensitivity. I keep referring to fidelity, that is where the big difference lies. A cone will tend to reject high frequencies, on a dynamic mic they handle this by not having one. That same cone offers a larger surface for air waves to interact with though, so the sensitivity is increased. Even with the reduced fidelity, a cheap speaker is better than telephone, and good as or better than AM radio.

The basic construction on both speakers and mics is virtually identical. There is a floating coil with a fixed magnet. It can be the other way around, but this is the norm. This is mainly due to the mass of the magnet.

The other thing is you need to adjust the base potietiometer for a value of 100ma or so through the collector. This puts the transistor in its analog region.

Do you have a protoboard?

Like I said I'm going to perform the same experment. Are you familar with the terms Darlington pair or Sziklai pair?

 

Nope, never heard of Darlington pair or Sziklai pair.

And yes of course I have a breadboard. What do you think I am building my circuit on? I have pretty much everything, soldering iron, multimeter, just no oscilloscope sadly. Expensive. But some day I will have one.

I built your circuit exactly in post #8 Bill. Well except my source was 4.5 volts instead of 3 volts. It didn't fry anything. I noticed that the speaker on the collector side does crackle at the moment when I make the last connection of the microphone at the base, i.e. there is a current in the collector.

I will try my circuit again and take it to the lab in my department to do some accurate measurements with multimeter and even an oscilloscope. I will come back with a full report of what is going on and all the measurements. Aye Aye captain.

 

Do you have a meter? You can adjust the collector as I suggested.

You probably notice I post only on certain times of day. My work schedule has gone to 6-7 days a week, so I'm here ususally before and after work (I work 10PM to 7AM).

The reason I mention the two pairs is they are methods to boost the gain, and yet they still look like a single transistor.

 

I didn't get much in the way of results either. I tried a Darlington for it's higher gain. In all cases I could hear scratching on the speakers, a lot clearer with the Darlington. I can make this work, but I'm going to have to tinker a bit.

I apologize for the work you've already done. I got the carbon mic version to work very well, it was a lot more important in the design than I knew.

My other thought is to use a capacitor to allow the dynamic mic more current flow. I suspect a condensor mic is worse, since it's signal is a lot lower, and requires conditioning. I tried a variation of the bias for the Darlington pair.

 

Hehe Bill I wish I had a carbon mic. The point of this experiment is for me to learn the limitations of theory vs real life. I'm still going to take my circuit and make measurements later today to see just how weak the signal is.

P.S. Does a carbon mic need an amplifier at all? Maybe the transistor is pointless with it?

 

Never tried it, some amplication is always good, but thinking about it, we're talking early telephone, so maybe not.

 

P.S. I tried the above schematic by using 8 ohm speakers for both the speaker and mic, and nothing, silence. A speaker diaphragm is not designed for responding to vibrations in the air, while a microphone diaphragm is.

examples like that will work only on paper and not in reality. speakers generate very low signal, and to power another speak, you need tremendous amount of gain, which is next to impossible given the low impedance the load speaker has.

yes, your condenser mic is an electret mic.

 

yes, your condenser mic is an electret mic.

No.
A condenser mic requires an external 48V supply to polarize its diphragm.
An electret mic has the 48V "built-into" its electret material.

 

An electret microphone is a condenser microphone.

http://en.wikipedia.org/wiki/Conden...ser.2C_capacitor_or_electrostatic_microphones

 

OK, I used a TIP101 and got a half way result. I used the 3rd schematic I drew. I could hear the mic if my ear was to the speaker very plainly, but never got any accoustic squeal. This transistor has a DC gain of 20000, so it has plenty of boost.

I set the collector current for 100ma, then replace the meter with the speaker.

I may try a triplett Darlington to see if that makes a difference, then give up.

 

Hey,
I've been following this thread and I think there's a problem with the current configuration of the circuit.

count_volta has said that when he speaks into the microphone, it outputs 0.1Vpeak ac signal.
Since the microphone is connected directly to the BJT's base, this 0.1Vpeak changes dramatically the operating point of the BJT, and it can make it swing between cut-off and saturation regions.

So, shouldnt the mic be connected before R1 and R2, and not directly to the base?
(but i dont know if its OK to connect the mic directly to the 3V source).
But at least that way, the operating point of the circuit will remain much more stable since 0.1Vpeak << 3Vdc (not exactly ' << ' but its better comparing to the relation between 0.1Vpeak to 0.7Vdc).

What do you think?

 

What you are referring to is overdrive. At this point we aren't even close (I wish!). The mic generates both current and voltage, in the last configuration I'm basically shorting the mic though a low impedance and using the current. With the transistor firmly biased on it is a very low input resistance.

I was wondering what would happen if I added a 3rd transistor to increase the gain even more.

I think you may be looking at the older schematics. I think the reason we had such a weak signal is the mic was trying to feed its signal through the base bias resistors. I've gotten rid of that problem by using a capacitor to allow a direct current path through the transistors BE junction, more or less.

The one conclusion I've come to is while speakers make great mics, I've done it too many times to think otherwise, the signal is small and very low impedance.

Another option is to go with the flow, and make a impedance matching transistor stage. There is a configuration called a common collector (AKA emitter follower) that might do the trick.

Understand this will NEVER be a practical amp. The design is extremely vulnerable to thermal runaway, which can take out the transistor and the speaker. It is a proof of concept circuit only. If I can get it to work very well I may turn it into an experiment, assuming something like that is not already in the AAC book.

I did make some minor changes. I used the resistors shown because that was what was in the box, and in the ballpark. I also wired the speaker to ground instead of Vcc, and rotated the 100µF cap accordingly.

There is no reason other people can't have fun with this, try something out an post it here or in your own thread.

 

Hey,
Thanks for explaining it to me.

I was speaking about the left circuit.
The (DC) operating point of this circuit should be:
VBE = ~0.7Vdc.
VCE = ~1.5Vdc.
That is when you dont speak into the microphone.

When you start speaking into the microphone, then the mic's output, which is an ac signal, overrides the DC level.
So correct me if i'm wrong please, wouldnt VBE will turn to be:
VBE = ~0.7Vdc + 0.1Vsin(ωt) ?

 

The problem with that circuit is a speaker isn't 8Ω at DC. You will never get ½ VCE. Also, you are assuming the voltage from the mic will stay intact, which isn't the case. The BE junction, which is forward biased, is a very low impedance contact with ground, and the diode junction will only change a little in voltage in response to the mic. At this point the active element of the mic is current, which is pretty low. The transistor is entirely dependent on the current generated by the mic.

In practical terms, I could barely hear the mic out the speaker. I figured out the biasing was not as good as it could be. It worked, but for a high gain device like a Darlington I could do better, which would also translate backwards. The new arrangement will allow me to adjust the collector current to 100ma, no matter what the transistor.

When I have more time I'll diagram some of the concepts out.

 

Hello,
Sorry then i must be misreading here something.

count_volta has said that when he speaks into the mic, then it outputs an ac signal of 0.1V amplitude.

Now you say that "the diode junction will only change a little in voltage in response to the mic".

Dont these two versions contradict each other?

I guess i just dont understand a basic thing, why when you start speaking into the mic (assuming continuously), the vBE voltage doesnt turn to be: vBE = VBE + vbe = ~0.7V + 0.1Vsin(ωt) ?

 

I don't know how he made the measurement. I'd be tempted to feed a mic into an oscope, which is a high impedance measuremet. It's what I've done in the past, but there is a difference in how a signal source like this would respond.

A diode junction like this isn't going to be as straight forward as you think. I'm not sure where the forumula came from, but you are trying to oversimplify the case, the signal source may or may not generate that much voltage with a simple sound, we are in the realm of analog, which is entirely variable by definition.

To cause a swing of 0.1 V on a PN junction is pretty extreme. It can be done, but not with microwatts of power. For really low power cases like this I start with 0.6V, transistors rarely exceed 0.7V (though regular power doides do). Transistors respond to current changes, and amplify current, as a general rule. These designs are first and formost current amplifiers, and respond to current changes. A 0.1VDC change on a junction would not be from a linear source, it would take a large change of current to cause this.

There are other configurations other than common emitter that would follow the input voltage, and not load the mic down. This is not one of them.

 

The problem with that circuit is a speaker isn't 8Ω at DC.

What do you measure for the DC resistance of your speaker?

By the way, a good place to get small speakers to experiment with is your local Goodwill or Saint Vinny's. Get a pair of computer speakers and break them open to get out the raw speakers.

My 8 ohm rated speaker obtained from Goodwill measures 7 ohms DC.

On another note, connecting the speaker to a scope and having a look at the output gave me this result:

Mouth nearly touching the cone, speaking in a normal voice gives about 100 mV.

Mouth 6 inches away with a normal voice gives a few millivolts.

 

A speaker makes a lousy microphone because of its undamped resonance.
A speaker with DC in it does not make a good mic and does not make a good speaker.
A transistor or darlington with a very low impedance collector load has a low voltage gain.

I would simulate it but don't have the time right now.

 

What do you measure for the DC resistance of your speaker?

*blush* A good question. Because sooner or later I'll be using one for one of the articles I bought a Radio Shack version, and have used a little 2" unit from an old computer (I have tons of old cases). I've been thinking about doing what you've suggested. Another little 1" unit, similar to what Radio Shack sell, seems to be pretty deaf, which was a distraction.

If I sound a little odd now and again I'm working heavy hours, but being a confirmed ACC addict can't stay away.

Count Volta asks interesting questions, the kind I like to step back and rethink what I think I know, and this one caught my interest. Like I said earlier, I was able to do this with one transistor, one speaker, a carbon mike, a 9V battery, and a clorox bottle. Too bad the mics aren't that available anymore.

A word to the other folks. The AAC book shows this as an example of a Common Emitter amplifier.

It is pretty typical for this kind of illustration. However, no one in their right mind uses a separate power supply to bias a part like this. So this is why I went with this biasing scheme I am currently using.

Add a resistor as a homage to Murphy (and to keep the smoke in) and you have my circuit.

A speaker makes a lousy microphone because of its undamped resonance.
A speaker with DC in it does not make a good mic and does not make a good speaker.
A transistor or darlington with a very low impedance collector load has a low voltage gain.

I would simulate it but don't have the time right now.

I've mentioned the fidelity issue, but it works well enough for commercial intercoms (remember school?).

I've addressed the DC issue.

This is a demo circuit. Adding circuits is a non starter, but it may explain some of my other problems with the gain.

But honest input is always appreciated.

I measured my 3 speakers, 7.9Ω, 8.9Ω, and 7.2Ω. My lesson for the day it seems.