Help please - mic preamp/mixer.

[picbuck]

Would anybody please like to take a look at the attached schematic and tell me whether I'm going to get mild smoke, or a real explosion?

The big mystery for me centers around R3 and R4, feeding the output buffer IC2B. I'm not sure what their values should be, or the values of pots Rmic or Rline, or the value of resistor R2d.

The purpose of this preamp-mixer is to feed an electret microphone's amplified output to a computer's line-in. The auxillary line-in is so that a line-in will still be available when the mic is in use.

As you see, the whole thing is built of nearly identical TL072 modules. I kinda think I have these correct, possibly.

I'm using only the +12 volts from the computer power supply, even though there's also about 1/2 watt of -12 volts available. This is because...well, I wanted to see how that would work.

Any help or comment will be greatly appreciated.

 

[audioguru]

Your electret mic will not work because it is not powered. It needs to be fed about 0.5mA from a filtered voltage of about +9V through a 10k resistor.
Your first opamp IC1A is inverting with a low input resistance of only 2.4k ohms which cuts the mic signal in half. IC1A should be a non-inverting opamp with an input rersistance of about 20k to 30k ohms so that it has the full mic signal at its input.

The mixer is OK but R2d can have its value increased if you need more gain.

 

[picbuck]

Well, so much for thinking I had those TL072s down.

Thanks for your reply, Uncle $crooge. I see what you mean and I'm on it.

 

[Nigel Goodwin]

Check the 'sticky' at the top of the General Electronics Chat forum, it contains AudioGuru's Electret mike preamp circuit.

I would agree with AG, it's a BAD idea using an inverting opamp for the first stage - and if you do, you need to seriously consider the impedance.

 

[sbarro]

I'm curious why you want so much gain. With two stages and amplification of 20 for each, your total gain is 400. For a pre-amp, and unless you want to amplify very soft sounds like those from some insects, this is way more gain than you need for the human voice. A single electret mic with built in FET, hooked up in typical fashion can put out around a volt of signal. I believe the actual level depends upon the power supply voltage and resistance of the power resistor. When you mix in several mics, you need about 50k resistors to prevent signal interference, and that resistance can cut down the signal level several times, to around a quarter volt. At least this is my understanding of how it works. Anyway, a single electret doesn't really need a pre-amp for normally audible sounds close to the mic, and even several of them in tandem might even work with a passive mixer, though I'm not recommending that without experimentation. I'm not sure what will then happen if your mic mixer is connected to a power amp with a long cord, but maybe someone more knowledgeable than me can explain further.

Best regards,
Tom

 

[audioguru]

I'm curious why you want so much gain. With two stages and amplification of 20 for each, your total gain is 400.

For over 5 years, I have my "VU meter" project showing sound levels in my family room. Its max voltage gain is 1820 but it has AGC that reduces the gain to 182 when sounds are loud. It picks up me breathing quietly 3m away or a pin dropped on the floor in the next room.

 

[Nigel Goodwin]

I'm curious why you want so much gain. With two stages and amplification of 20 for each, your total gain is 400. For a pre-amp, and unless you want to amplify very soft sounds like those from some insects, this is way more gain than you need for the human voice. A single electret mic with built in FET, hooked up in typical fashion can put out around a volt of signal. I believe the actual level depends upon the power supply voltage and resistance of the power resistor. When you mix in several mics, you need about 50k resistors to prevent signal interference, and that resistance can cut down the signal level several times, to around a quarter volt. At least this is my understanding of how it works. Anyway, a single electret doesn't really need a pre-amp for normally audible sounds close to the mic, and even several of them in tandem might even work with a passive mixer, though I'm not recommending that without experimentation. I'm not sure what will then happen if your mic mixer is connected to a power amp with a long cord, but maybe someone more knowledgeable than me can explain further.

You 'might' get a volt out of an electret mike - if you placed it next to a jumbo jets engine!.

In normal use the levels are similar to other types of mikes, low mV.

I suspect you are woefully hopefull about passive mixers

One opamp for each mike, as per the sticky I mentioned above (you can easily tweak the gain of each one to what you want), and one opamp as the mixer.

 

[sbarro]

You 'might' get a volt out of an electret mike - if you placed it next to a jumbo jets engine!.

In normal use the levels are similar to other types of mikes, low mV.

I suspect you are woefully hopefull about passive mixers

One opamp for each mike, as per the sticky I mentioned above (you can easily tweak the gain of each one to what you want), and one opamp as the mixer.

Sorry Nigel, but I just did the experiment and confirmed that, with a mic supply of 3 Volts and a 2.2k power resistor, with a 10 uF isolation cap, on a scope, I measure one Volt AC signal amplitude (two Volts pk-pk), with a moderately loud yell into the mic at about 1 inch distance. And I assure you, I cannot yell nearly as loud as a jumbo jet. The electret is the off-the-shelf Panasonic WM-61A model, which I believe has a built in FET. I haven't checked yet how signal strength changes with supply voltage and power resistor, though the values given closely approach the standard operational conditions given in the mic specs, and nowhere near the max voltage specified. Have you also done experimentation with this? If so, I'm curious why your experience and mine are so different. Perhaps someone else would like to add data of their own. As an aside, I believe lapel mics are electrets run with phantom power and with the signal going directly into PA systems, without the need for pre-amplification.

But let me ask, if it's indeed true that these little buggers can put out so much, wouldn't passive mixing seem like a good idea? The suggestion to stick op-amps everywhere feasible or imaginable is contrary to my own idea of good circuit design, which emphases simplicity and efficient component utilization, even apart from other mundane issues such as cost.

Best regards,
Tom

 

[Nigel Goodwin]

Sorry Nigel, but I just did the experiment and confirmed that, with a mic supply of 3 Volts and a 2.2k power resistor, with a 10 uF isolation cap, on a scope, I measure one Volt AC signal amplitude (two Volts pk-pk), with a moderately loud yell into the mic at about 1 inch distance. And I assure you, I cannot yell nearly as loud as a jumbo jet. The electret is the off-the-shelf Panasonic WM-61A model, which I believe has a built in FET. I haven't checked yet how signal strength changes with supply voltage and power resistor, though the values given closely approach the standard operational conditions given in the mic specs, and nowhere near the max voltage specified. Have you also done experimentation with this? If so, I'm curious why your experience and mine are so different. Perhaps someone else would like to add data of their own. As an aside, I believe lapel mics are electrets run with phantom power and with the signal going directly into PA systems, without the need for pre-amplification.

But let me ask, if it's indeed true that these little buggers can put out so much, wouldn't passive mixing seem like a good idea? The suggestion to stick op-amps everywhere feasible or imaginable is contrary to my own idea of good circuit design, which emphases simplicity and efficient component utilization, even apart from other mundane issues such as cost.

A moderately loud yell from an inch away is incredibly loud - hardly normal useage.

The internal FET isn't an amplifier, it's just a buffer - needed to preserve the stored charge on the electret element.

You can't get more efficient than an opamp mixer, one opamp, one resistor, and an extra resistor for each channel, resulting in a perfect mixer. Any sort of passive design loses MASSIVE amounts of signal, and is no where near perfect.

What are you wanting to do with the mixer anyway?.

 

[picbuck]

<< I'm curious why you want so much gain. >>

Thanks for your interest, sbarro. I'm a babe in the woods here, so after very much surfing around I decided an electret cartridge (Radio Shack, in this case) puts out a 2 millivolt signal voltage. I then settled on a gain of 400 to approximate a 1 volt line/preamp (as far as I know) signal voltage.

Additionally, this all comes about because of wanting to record into a computer using Audacity audio editor. My present setup is a powered electret feeding into the computer's mic input, and this does work fine. (the batteries for the mic, along with a switch, are in an Altoids box, as is customary)

It works fine, but not as fine as I'd like--or I think I'd like, maybe. Judging from the visual plot in Audacity, the recording level is perhaps some 20% of maximum. I'd like to raise this because...well, I kinda think it might be a good idea. I have some notions involving noise removal (by Audacity) and background noise.

Thus a preamp, feeding the mic into the computer's line-in jack. And then that uses up the line-in jack so I wanted to replace that in case of need, so all of this is what I came up with...so far.

 

[picbuck]

<< I haven't checked yet how signal strength changes with supply voltage and power resistor >>

Just to mention it, I'm running my electret with 3 volts (batteries checked good). I did substitute a 9 volt battery (checked good), and this made no difference in sound levels that I could detect. But my "detection" was by ear, not with any kind of instruments.

 

[picbuck]

<< Check the 'sticky' at the top of the General Electronics Chat forum, it contains AudioGuru's Electret mike preamp circuit. >>

Thanks Nigel, I did check that schematic. Pored over it religiously, in fact. But then I started adding stages of amplification and, more importantly, added a mixer, so I got lost in the valley of what size resistor goes where? But thanks to the help I received here things are improving.

 

[picbuck]

<< What are you wanting to do with the mixer anyway? >>

The original idea was/is to put a mic signal through a preamp, to be fed to a computer's line-in jack.

This uses up the line-in jack, of course, so I wanted to replace that in case of need.

Thus a preamp-mixer with two inputs, mic and line-in. The mic input amplified, the line-in input fed straight through a buffer, the two signals mixed, and the output still feeding the computer's line-in jack.

I envision all of this installed into a blank cover on the front of the computer, along with a line-out (to speakers) jack. I'm sure some would be horrified by all those wires snaking around the front of the computer, but to my eye the whole thing will look pretty snazzy.

 

[sbarro]

Hi Picbuck,
What is the gain of your existing mic setup? An unknown here is how much input voltage signal can your computer tolerate. My suggestion I suppose differs from that of Nigel, and I'd recommend using a single stage op amp preamp with a maximum gain of about five times the gain of your existing mic setup, and with a pot at the preamp output for manual gain control. As I recall, those cheap microphones you can buy for computers are run off phantom power from the computer, and they have no preamp. Thus, I'm curious whether your existing mic set up has unity gain.
Regards,
Tom

 

[audioguru]

The mic input on a computer powers an electret mic and has plenty of gain, maybe 200.
The line input on a computer has maybe 200 times less gain.

I have seen plenty of singers with the mic against their lips and screaming into the mic. Others sing softly at a distance of 30cm from the mic. Then the difference in level is a lot.

If the preramp has enough gain for the soft singer then it is correct. The gain can always be turned down for the screaming singer when necessary.

 

[sbarro]

Picbuck, you may find this interesting:
Powering microphones

Regards,
Tom

 

[picbuck]

Help please - mic preamp/mixer

Thanks again, audioguru and Tom for your observations and comments, and for the web page on powering electret microphones. Work proceeds.

Tom asked:

<< What is the gain of your existing mic setup? >>

No gain at all. It's just an Altoids box (as is required by law) with a switch, and two batteries making 3-VDC. It feeds directly to the computer's microphone jack. For clarity, here's a schematic.

 

[picbuck]

Help please - mic preamp/mixer

Moving along, I've made changes as advised...I think. New schematic attached.

Although it scares me, I fooled with audioguru's sticky schematic. This because I didn't know how to connect multiple amplification stages, so I went with additional information from Charles Kitchin, who seems to be known figure. Here's his article:

https://www.electro-tech-online.com/custompdfs/2011/01/avoiding.pdf

Although I didn't/don't want to take up anybody's time with unnecessary details, maybe some brief clarification is in order.

This is all an attempt on my part to solve two problems:

First: I think I need a higher audio input level to my computer to keep Audacity happy. Possibly because, by choice, I run my motherboard's onboard audio, i.e. no sound card.

Second: Actually more important, WindowsXP already has most of what's needed for making recordings, but once you start trying to do some actual work it gets to be a pain of monumental proportions to click through what seems like 1,000 windows to adjust inputs or levels.

Then add in switching between speakers and headphones...by replugging in the back of the computer. Things get irritating.

So my solution is to bypass Windows entirely. Set the computer's audio levels once and then leave them alone, passing control to the planned preamp and mixer. And also to a switched speaker/headphone output, but I think I have that figured out from surfing these forums.

And as before, the computer's microphone jack is no longer to be used. All input to the computer will be through its line-in jack.

As far as I can tell, a level of about 1-volt is expected at preamp or line-in terminals, so that's my planned input to the computer's line-in jack. If I'm anywhere near the correct levels in my planning, the 3 preamp level controls (mic, line, master-out) will take care of the details...hopefully.

Whew. Such is the plan. And thanks again for you help.

-al-

 

[audioguru]

Your new schematic does not have a gain control for the mic preamp so it might be overloaded by a jumbo jet engine nearby or somebody screaming very close to the mic. My preamp circuit in the Sticky has the 470k pot as its gain control. It adjusts the gain from 1 to 214.
Maybe the 150k negative feedback resistor for your 1st opamp should be a pot.

 

[picbuck]

<< Your new schematic does not have a gain control for the mic preamp...My preamp circuit in the Sticky has the 470k pot as its gain control. It adjusts the gain from 1 to 214...Maybe the 150k negative feedback resistor for your 1st opamp should be a pot. >>

That was the plan from day 1. The counters show that some 12,000 people have viewed your Sticky circuit. Certainly tens or them if not hundreds (thousands?) have built it, and been happy, and bragged to their girlfriends, and received their reward, so I figured what can be wrong?

But that was before I ran afoul of Charles Kitchin, waving his degrees and published papers (many). Mr. Kitchin warns in grim detail of the dangers of bias errors, but then offers a solution. And we who keep our pencils in carefully straight lines are influenced by such things.

For those who enjoy confusing themselves, here's a link to Mr. Kitchin's article. I also attach a graphic of his counsel below.

https://www.electro-tech-online.com/custompdfs/2011/02/avoiding.pdf

So bottom line, I drew audioguru's circuit with Mr. Kitchin's additions. All bases covered. When in doubt do everything. If that doesn't work do it again, but louder.

Schematic attached for those who are gluttons for punishment.

Happily, my original confusion was how to connect the stages of amplification, and this was cleared up by audiguru and Nigel. Thanks again for helping me out, and onward to order parts.

- al -