1
\$\begingroup\$

I am unsure what is wrong with my amplifier / microphone coupling.

I have read many of the posts on stackexchange including the following:

http//electronics.stackexchange.com/questions/68525/electret-microphone-with-class-a-amplifier-has-no-speaker-output

  • I am using low impedance headphones (~38Ω, Logitech)
  • I have tried bridging the pins on the jack leaving the LM386 and it helps, but not enough

http//electronics.stackexchange.com/questions/76020/linkwitz-mic-preamp-mod

  • I tried measuring the impedance on the mic - obviously incorrectly. But, initial values seem to start around the same (I think this is the problem?)

The amplifier circuit on my breadboard :

Based on this audio amplifier kit.

Datasheet here: http//www.jameco.com/Jameco/Products/ProdDS/151204.pdf

As far as I can tell, the circuit is working as designed (i.e. there is audible output from the circuit; gain adjustment seems to work, no smoke, etc)

So far, I only have one microphone to test it with. It is a "Boundary Business Microphone" from RadioShack.

As sold, there was a pendant about 18" from the condenser microphone which housed a small circuit and took a single LR44 (button cell, 1.5v) battery.

It is currently modified to take an LR6 (AA, 1.5v) battery. The diagram for the (modified) circuit is below:

http//i.imgur.com/rrswB49.jpg

I presume the reason for the inclusion of the pendant is the ~16ft long cable (which I have no desire to remove).

When I made the aforementioned modification, I was suspicious that the change from an LR44 to an LR6 (AA, 1.5v) would alter the behavior of the microphone. However, this has not been a problem when using this mic with a PC audio card (SB Audigy) or a handheld voice recorder. With this circuit, I can hear only faint input even when blowing hard onto the mic or tapping it with my finger.

Do I have an impedance mismatch due to two/too many pre-amp circuits?

If so, how do I fix it? If not, please point me in the right direction.


My proposal:

  1. Return the microphone circuit as near to the original operational parameters as possible: (for the sake of "knowing it is right")
    • The LR44 has an internal resistance of 4.19Ω
    • The LR6 has an internal resistance of 0.28Ω
    • ^Values obtained here: http//www.g-holbeck.com/english/elteknik/indexri.htm
    • The resistor in series with the battery on the circuit board is 2.2kΩ
    • I should replace the 2.2kΩ resistor with a 2.6kΩ in the microphone circuit? Is it that easy?

And, since I will need to use this microphone with this circuit for now:

  1. Add a bypass to the microphone to decouple the pre-amp circuit in the pendant

or:

  1. Reduce the input voltage on the amplifier circuit

    • Currently at 9v
    • The datasheet for the circuit recommends 7 - 15v input.
    • Lowering the voltage will reduce the output from the LM386?
  2. Reduce the voltage supply to the microphone

    • Seems like this will require adjustments to the bias voltage.
    • Not sure I have a clue how to do this...

or:

Something I have totally overlooked? I am a newb after all. Please forgive my ignorance.

Thanks for any help, I really want to understand this.

PS: Could only add two links, so some are intentionally broken above.

Edit://

The circuit which I built is on the right below. I included the microphone pendant circuit in this diagram on the left. I had to draw this myself because the datasheet listed above explicitly claims copyright to its contents (so, no screenshots).

^^Right click, view image (in Firefox, IE at least) to see the full size.

I tried to embed the image other ways but EE.SE considers anything other than this way a 'link' that contributes to my two link maximum.

\$\endgroup\$
  • \$\begingroup\$ How are you connecting the modified mic to the amp? That is, where do your mic's "Output (+)" and "Output (-)" terminals go? \$\endgroup\$ – EM Fields Jul 25 '14 at 8:21
  • \$\begingroup\$ Just show the precise circuit you think you are using. \$\endgroup\$ – Andy aka Jul 25 '14 at 11:27
  • \$\begingroup\$ Do you have a 'scope? You should be able to see the output from the mic on a 'scope. (~10 mV - 100 mV level.) You need to provide a bias voltage for the microphone. I would guess more than the 1.5 volts you are using... but maybe not? Can you measure the bias voltage the radio shack thing was using? \$\endgroup\$ – George Herold Jul 25 '14 at 13:24
  • \$\begingroup\$ @EMFields I have updated the question to include a diagram for the circuit. \$\endgroup\$ – Gabriel Jul 26 '14 at 0:39
  • \$\begingroup\$ @Andyaka I have updated the question to include a diagram for the circuit. \$\endgroup\$ – Gabriel Jul 26 '14 at 0:39
0
\$\begingroup\$

It is possible that the circuit may simply not have enough gain for the microphone, that's all.

You have a gain of 10X in the first stage. A LM386 without any connection between pin 1 and 8 has a gain of 20X. Together, these multiply out to 200X which may seem like like a lot. But, expressed in decibels, it is only:

$$20\times\log_{10}{200} \approx 46\text{dB}$$

It could simply be that this is not quite enough gain for the given microphone.

The datasheet for your RadioShack mic gives a nominal sensitivity of -65 dB, which means that for that given reference sound pressure level, the signal would have to be amplified by 65 dB just to reach 1 Volt RMS.

If this sensitivity figure is referenced to the standard 1 Pa = 94 dBSPL, it is rather low in comparison to some other microphones. For instance, the widely used dynamic microphone Shure SM58 is quoted as -54.5 dB, which is more than 10 dB "louder".

Because mic sensitivities can be so low, microphone preamplifiers usually have gains that begin at around 60 dB and can get as high as 80 dB or more. And note that these gains are just for the microphone pre-amp to bring a signal to "line level"; they do not include any additional gain in power amplifiers for driving speakers or headphones. In your circuit, we are including the power amp's gain.

\$\endgroup\$
  • \$\begingroup\$ I cannot upvote your answer or I would. Also, I would like to add that by adding a 10uF capacitor across 1 & 8 and a 0.1uF cap between 7 & gnd on the LM386 did add quite a bit of amplification. However, there is still some quirky behavior going on. For example, turning the 10K trim pot 'down' increases amplification in a certain range (roughly between 1/2 & 1/3 way to left). Also, when certain volume levels are induced on the mic, the amplification is apparently too much and leaves the output as a high pitched squeal. Too much output from/to LM386 or biasing circuit too latent? \$\endgroup\$ – Gabriel Jul 28 '14 at 20:53
  • \$\begingroup\$ Volume swells in the potentiometer could be a flaky ground connection, or intermittence of the wiper. Substitute a different pot, or try cleaning that one. Since the pot is capacitively coupled, to the previous circuit, it should not disturb its bias. It's direct coupled to the LM386 (so moving the pot affects bias current coming from the + input), but the datasheet examples do exactly the same thing. \$\endgroup\$ – Kaz Jul 28 '14 at 21:42
  • \$\begingroup\$ Thank you so much for helping me out with this question. Although you have provided a thorough answer; you really should have asked me if I had read the datasheet on the LM386!!! Imagine my red face when I realized that I had not... Anyway, I think I will move this project from breadboard to perfboard and see how it behaves then. Thanks again! \$\endgroup\$ – Gabriel Jul 28 '14 at 21:46
  • \$\begingroup\$ Squealing, hmm. Hard to say, but it could point to some oscillation, though it's hard to explain why that would only be present when you have input. Grasping at straws: distortion creates non-linear mixing of signals, so RF interference gets demodulated into audio range. See if meticulous grounding makes a difference: make the LM386 and the pre-amp have separate ground returns all the way to the power supply. \$\endgroup\$ – Kaz Jul 28 '14 at 21:48
  • \$\begingroup\$ Re: meticulous grounding - will do. \$\endgroup\$ – Gabriel Jul 28 '14 at 21:49
0
\$\begingroup\$

That circuit helps a lot. So a few things. I'd move R4 (10k ohm power supply splitter) down so that it didn't block the 9V from the 3.3k mic bias resistor.

Second I don't think the 100nF coupling cap is big enough. You're loosing signal through the 3.3k bais resistor and then through the 4.7uF cap on the non-inverting input. (or on 9V rail once R4 is moved.) Calculate the corner frequency of 3.3k and 100nF. There's a similar voltage divider on the output, with a 10k resistor.

Oops! Well I don't know the LM386 either! One thing I will observe looking at the LM386 spec sheet. All the amp circuits show 10 ohm in series with 0.05uF on the output. I couldn't find a reason for this in the rather thin section on "Application Hints". But you might try adding it.

\$\endgroup\$
  • \$\begingroup\$ I agree, the two diagrams seem to have conflicting information. For example, the circuit I built has a 47uF cap on pin 5 to the output. But, the LM386 datasheet shows (like you said) a resistor and cap (much lower farad) in series on the same pin out. \$\endgroup\$ – Gabriel Jul 28 '14 at 21:03
  • \$\begingroup\$ Yeah the 10 ohms 0.05 uf has corner somewhere a bit below 1MHz, the IC may need a bit of loading up there. You should also do the 100nF/ 3.3k ohm corner freq. (I get ~500 Hz).. where is middle C again? \$\endgroup\$ – George Herold Jul 28 '14 at 23:09

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.