Adafruit MAX4466 mic board connection to audio in on mini DVR

[hoghunter]

I am supplying the electret mic board with 3.3v and it's OUT pin to the READYTOSKY DVR's audio IN pin. The DVR is powered by a separate 5v. Grounds are isolated from each other. I am getting no recording of my voice during testing. Just a bunch of noise (I tried to attach my AVI file)!! I have adjusted the gain twice, no change. I added a 22uf capacitor in line with the audio out/in, no change.... The data sheet is attached. What am I doing wrong? Please help.

 

[hoghunter]

I believe I found the problem. I had the audio IN ground wire attached to the 5v side (DVR). I moved it to the 3.3v ground along with the mic ground and have good audio now.

 

[hoghunter]

New problem....... I have both 3.3v & 5v on the breadboard, but my final circuit will only have 12v and 5v. The mic board likes 3.3v. When I feed the DVR and the mic board with 5v and both go to the same 5v ground, I am back at square one.............just a bunch of noise. Now what? How can I clean up this audio signal using 5v?

I am also trying to 'trigger' recording on the DVR with a signal from a digispark controller. I've been messing with the audio and triggering issues for a few weeks now and I am still stumped!
Any help would be appreciated.

 

[hoghunter]

UPDATE: I placed both mic and DVR on 5v (+/-). I then placed a 4.7k resistor on the (+) lead to the mic to drop voltage down to approx 3v. I also placed a 100uf capacitor on the out wire from the mic.
I am getting good audio now, but am still getting electrical 'noise' in the audio. Could this be as simple as adjusting the gain? Or, do I need to adjust what I already have? Or, maybe it is this adafruit mic board I am using? I'm sooooo close!

 

[audioguru]

You are the second person to complain on a forum about the noise from the cheap Chinese mic module. Simply look at the MAX4466 preamp used and compare its very high noise level to an audio opamp like an OPA134. It has 200 times more noise!

 

[hoghunter]

You are the second person to complain on a forum about the noise from the cheap Chinese mic module. Simply look at the MAX4466 preamp used and compare its very high noise level to an audio opamp like an OPA134. It has 200 times more noise!

Fair enough. I will step away from the adafruit audio breakout board. I asked previously in another thread specifically which electret mic I should get as digikey sells dozens of different types with dozens of different parameters. So my question is, which electret mic do I need to order and should I also order the OPA134 'op-amp' ? In the meantime, I will try to scavenge a mic from something around the house and use that instead. The mic will be used to pick up voice/sounds from 12" to 10' away and send the signal to a small 5v DVR to be recorded onto a micro SD card.
Your input is very much appreciated. Thanks.

 

[hoghunter]

My circuit will be supplied by a 12v battery. Is the circuit below more or less what I need to put together? I assume the signal (+) from the electret mic would go to 'L in'? And, 'AL' (audio left in this schematic) would take the clean mono signal to the DVR's Ain ? If this OPA2134 is what I need, where should I get it from? Digikey?

After more research, I am guessing this is a 'preamp' circuit that is incorporating an OPA2134 'op amp'? The THAT 1510P also sounds like an ideal op amp to use also. I have looked at many videos and instructionals online and have seen a lot of voice recording or to speaker circuits that do not use an IC of any kind and they are getting very clean results. Do I need the op amp for my intended use as I am not outputting to a speaker or amplifier, but just recording to a dvr?


**broken link removed**

 

[audioguru]

I have used electret mics from toys, telephones and cheap intercoms. When they have the same size then their sensitivity are almost the same and their frequency response sound the same.

Your preamp with an opamp has a gain of only 2 times, a little more than a piece of wire. You probably need a gain of 200 which can have a volume control to reduce loud sounds.
One transistor might not have enough gain and its negative feedback would be almost zero so its distortion will be very high. Two transistors were used as a fairly good preamp 50 years ago. An audio opamp used as a preamp has distortion extremely low which cannot be heard and is hard to measure. Some old people are used to hearing old circuits with high distortion (fuzz or overdrive) and they like it.

Why the value of your R5 so high? Usually it is not needed unless the preamp drives a shielded cable (100 ohms isolates the capacitance of the cable to prevent high frequency oscillation).
I have never used R2 on all the preamps I have built.
All ICs need a supply bypass capacitor or two as close to the IC as possible. A 0.1uF ceramic and a 10uF electrolytic.

 

[hoghunter]

I have used electret mics from toys, telephones and cheap intercoms. When they have the same size then their sensitivity are almost the same and their frequency response sound the same.

Your preamp with an opamp has a gain of only 2 times, a little more than a piece of wire. You probably need a gain of 200 which can have a volume control to reduce loud sounds.
One transistor might not have enough gain and its negative feedback would be almost zero so its distortion will be very high. Two transistors were used as a fairly good preamp 50 years ago. An audio opamp used as a preamp has distortion extremely low which cannot be heard and is hard to measure. Some old people are used to hearing old circuits with high distortion (fuzz or overdrive) and they like it.

Why the value of your R5 so high? Usually it is not needed unless the preamp drives a shielded cable (100 ohms isolates the capacitance of the cable to prevent high frequency oscillation).
I have never used R2 on all the preamps I have built.
All ICs need a supply bypass capacitor or two as close to the IC as possible. A 0.1uF ceramic and a 10uF electrolytic.

The circuit is just copied from a search I did. It is not mine. I just posted it to verify the type of circuit using an OPA2134. I have a 4.7kΩ resistor in the 5v lead going to the microphone (+) which pulls it down to approx 3v. I also have a 100μf capacitor in the signal lead going to the DVR's audio in. That got me audio finally. My recorded voice sounds good, but I have a very noticeable and steady 'hiss' that I need to get rid of. The mic I am using is one of these cheap amazon's for now:

 

[audioguru]

Most electret mics have only 2 wires, the audio wire and the ground wire. Your Amazon mic has 3 wires and might have very poor performance (hiss and distortion).

 

[hoghunter]

I won't be using that junk after all. I took the electret mic off of that board (in shrink tube) and soldered a pos and neg lead to the mic. I then connected it directly to an audio jack to my PC (red to red & black to black). I used my PC's 'sound recorder' to test it. The recording of my voice sounded great. Since doing that, I have created a circuit on the breadboard (see below) and it is complete, except for adding a 1 uf or .1uf capacitor as I am waiting for some capacitors to come in the mail. I have the output signal going to a mini DVR for recording. I tested the circuit without the capacitor and got little or nothing. I am powering the mic with 5v. The DVR is powered by the same 5v. I assume the .1uf capacitor is a critical item in this circuit and will make all the difference?
Some of the circuits I see after doing a search online, show a capacitor on the 'output' and some do not have it. Should I add that one also?

 

[hoghunter]

Most electret mics have only 2 wires, the audio wire and the ground wire. Your Amazon mic has 3 wires and might have very poor performance (hiss and distortion).

I guess my main question to you is, will that circuit work if sending the signal to a DVR like the one pictured in post #1?

 

[audioguru]

I do not know if your DVR audio input is from a microphone (very low level) or from an MP3 player or radio (fairly high level).
The mic preamp circuit you found uses a 23904 transistor that has a wide range of current gain. Therefore some transistors will be cutoff and produce distortion, some transistors will work fine and some transistors will be saturated and produce distortion. The preamp will produce no low frequencies if its input capacitor is only 0.1uF, use 1μF to 2.2μF instead. The negative wire of the capacitor should be at the base of the transistor.
A preamp should always have an output coupling capacitor but its value must be calculated with the impedance it drives and the lowest frequency you want.

 

[hoghunter]

I do not know if your DVR audio input is from a microphone (very low level) or from an MP3 player or radio (fairly high level).
The mic preamp circuit you found uses a 23904 transistor that has a wide range of current gain. Therefore some transistors will be cutoff and produce distortion, some transistors will work fine and some transistors will be saturated and produce distortion. The preamp will produce no low frequencies if its input capacitor is only 0.1uF, use 1μF to 2.2μF instead. The negative wire of the capacitor should be at the base of the transistor.
A preamp should always have an output coupling capacitor but its value must be calculated with the impedance it drives and the lowest frequency you want.

Fantastic! That is the information I have been looking for. You have helped me a lot, thank you. I will be using the 2n3904 as I have twenty more to use. The DVR audio "in" has a positive white lead and a black negative lead and the signal will be from an electret mic. I am not entirely sure where to connect the black NEG lead from the DVR's audio 'in' to...... and, of course, the white lead will connect to the POS signal 'out from the mic circuit.

I put a trim capacitor at the coupling capacitor location for now. I will mostly be recording soft voice (not loud) and sounds of nature around a deer feeder so high sensitivity is required. That said, what size capacitor should I start with for the coupling at the output? Thanks again, this really helps. By the way, I am 'old', but I want to reduce the background noise as much as possible...


This is my circuit:

 

[audioguru]

You show R6 that powers the mic as 10k which is a value too high when the supply is 5VDC. The datasheet for most electret mics show that they draw a current of 0.5mA. Then for half the supply voltage to be at the mic so it can swing the signal up and down equally, the resistor value must be (5v/2)/0.5mA= 5k ohms. Use a standard value of 5.1k.

The black input wire of the DVR audio input is the ground of the preamp and it is the shield of the connecting cable.
Since you don't know the input impedance of the DVR then you cannot calculate the value of the output capacitor. If the value is low then there will be no low frequencies. If the value is too high then it will take a long time to charge and it might pickup frequencies too low (from earthquakes?).

How will you measure the current gain of your 2N3904 transistors so that you might select one that works in this extremely simple circuit? You want the collector voltage to be at half the supply voltage so that its can swing up and down equally. If the transistor's current gain is too low then the transistor does not conduct enough and its output voltage will be clipping when it swings up. If the transistor's current gain is too high then it conducts too much and its output voltage will be clipping when it swings down. The collector current should be 2.5V/10k= 0.25mA. The base voltage will be about 0.62V then the 100k base bias resistor will have 2.5V - 0.62V= 1.88V across it and a current of 1.88V/100k= 18.8uA. The current gain of the transistor must be 0.25mA/18.8uA= 13.3 times which is too low. The value of the 100k bias resistor should be higher so that the circuit works with transistors that have "normal" high gain. The transistor's datasheet shows a minimum current gain of 55 and a typical current gain of 230 times. Some might have a current gain of 350 times or more.

 

[hoghunter]

You show R6 that powers the mic as 10k which is a value too high when the supply is 5VDC. The datasheet for most electret mics show that they draw a current of 0.5mA. Then for half the supply voltage to be at the mic so it can swing the signal up and down equally, the resistor value must be (5v/2)/0.5mA= 5k ohms. Use a standard value of 5.1k.

The black input wire of the DVR audio input is the ground of the preamp and it is the shield of the connecting cable.
Since you don't know the input impedance of the DVR then you cannot calculate the value of the output capacitor. If the value is low then there will be no low frequencies. If the value is too high then it will take a long time to charge and it might pickup frequencies too low (from earthquakes?).

How will you measure the current gain of your 2N3904 transistors so that you might select one that works in this extremely simple circuit? You want the collector voltage to be at half the supply voltage so that its can swing up and down equally. If the transistor's current gain is too low then the transistor does not conduct enough and its output voltage will be clipping when it swings up. If the transistor's current gain is too high then it conducts too much and its output voltage will be clipping when it swings down. The collector current should be 2.5V/10k= 0.25mA. The base voltage will be about 0.62V then the 100k base bias resistor will have 2.5V - 0.62V= 1.88V across it and a current of 1.88V/100k= 18.8uA. The current gain of the transistor must be 0.25mA/18.8uA= 13.3 times which is too low. The value of the 100k bias resistor should be higher so that the circuit works with transistors that have "normal" high gain. The transistor's datasheet shows a minimum current gain of 55 and a typical current gain of 230 times. Some might have a current gain of 350 times or more.

Thank you for the education and specifics! I will study and hopefully absorb what you have explained and will make adjustments to my schematic. I will post results as soon as I get the circuit tested as I am expecting my capacitors to arrive shortly.
Thanks again!

 

[hoghunter]

audioguru,

I updated my schematic as I interpreted your comments:
I have some 4.7k resistors, so used that instead of 5.1k for now. I indicated the GND going to the AUD in of the DVR (shielding). After studying your math related to the 100k bias resistor, I came up with a 3.3k which would give me a gain of 228 times I believe? Is a 3.3k considered 'higher' than a 100k?

Would I be in much better shape if I created a circuit using an OPA2134 opamp IC as you had mentioned previously? I have some of those coming with my capacitors.......
Would it make for a cleaner sound also?

 

[audioguru]

3.3k is crazy to bias a low current transistor. You want the collector to be at half the supply voltage (2.5V) then the collector current is 2.5V/10k= 0.25mA. The transistor has a current gain of from 55 to 230 or more so the base current should be 0.25mA/230 to 0.25mA/55 which is a very small current produced by a very high value base bias resistor. Assuming the transistor has a current gain of 230 times then the base current must be 0.25mA/230= 1.1uA and with its base at 0.62V and its collector at 2.5V then the base bias resistor value should be 1 .7 million ohms. If the current gain of the transistor is 55 then the base bias resistor value should be 414 thousand ohms. Use 1.5M then measure the collector voltage.

Transistors were used in preamps 50 years ago to make lots of distortion, use an audio opamp today so the distortion will be so low that you can barely measure it.
An OPA2134 has two opamps in it. Do you have a stereo microphone? An OPA134 has one opamp.

 

[hoghunter]

3.3k is crazy to bias a low current transistor. You want the collector to be at half the supply voltage (2.5V) then the collector current is 2.5V/10k= 0.25mA. The transistor has a current gain of from 55 to 230 or more so the base current should be 0.25mA/230 to 0.25mA/55 which is a very small current produced by a very high value base bias resistor. Assuming the transistor has a current gain of 230 times then the base current must be 0.25mA/230= 1.1uA and with its base at 0.62V and its collector at 2.5V then the base bias resistor value should be 1 .7 million ohms. If the current gain of the transistor is 55 then the base bias resistor value should be 414 thousand ohms. Use 1.5M then measure the collector voltage.

Transistors were used in preamps 50 years ago to make lots of distortion, use an audio opamp today so the distortion will be so low that you can barely measure it.
An OPA2134 has two opamps in it. Do you have a stereo microphone? An OPA134 has one opamp.

Ok, I am still working on understanding the math. I'll take another stab at it.


Oh wonderful!! I ordered these ON Ebay.......... another week or so wasted. I could 'guess' what I need till the cows come home.
No, I am using a typical electret mic. When I looked up the OPA134 I did not see anything mentioning that the OPA'2'134 was specifically for 'STEREO'.
Can I use these, or are they trash as far as my use is concerned?


5 PCS OPA2134PA TEXAS INSTRUMENTS ORIGINAL PART SHIPPED FROM USA

 

[Nigel Goodwin]

When I looked up the OPA134 I did not see anything mentioning that the OPA'2'134 was specifically for 'STEREO'.
Can I use these, or are they trash as far as my use is concerned?

It isn't specifically for stereo, and AG never suggested it was, he merely asked if your mike was stereo as you were choosing a dual opamp.

You can use just one of them if you wish (ignoring the other), but it's rather a waste - you could use the two for stereo, or use the two for different purposes.