Go Back   Home > Forums > >
Home Forums Rules Articles diyAudio Store Blogs Gallery Wiki Register Donations FAQ Calendar Search Today's Posts Mark Forums Read

Instruments and Amps Everything that makes music, Especially including instrument amps.

New old-school electret mic preamp
New old-school electret mic preamp
Please consider donating to help us continue to serve you.

Ads on/off / Custom Title / More PMs / More album space / Advanced printing & mass image saving
Reply
 
Thread Tools Search this Thread
Old 30th January 2018, 01:56 PM   #1
Rufinus is offline Rufinus  Germany
diyAudio Member
 
Join Date: Nov 2004
Location: Stuttgart
Question New old-school electret mic preamp

Hello colleagues,

i was recently trying few lavalier mics in combination with Zoom H1 recorder and came to conclusion, that the simple electret mics, which are running from 1.5V LR44 battery or similar are just having too low gain. Therefore best mic - is DIY mic! I hope at least.

Of course OP-amp design would be an option, but i would like to have low noise, low current consumption plus simplicity of old good BJT common emitter. Therefore following schematics was found:

Click the image to open in full size.
Source: http:// 404 Not Found


But it is designed for 6-30V! And i plan to run my mic from 3V silver-oxide battery.

Therefore i tried to recalculate it for 3V:


Click the image to open in full size.

What i'm not sure is about the value of R3, as i don't quite understand how it works. For me it looks like Q1 base biasing voltage with negative feedback from Q2 (emitter follower).

Is my recalculation correct? What is confusing me a bit are the voltages at Q1 collector and Q2 emitter - they are off Vcc/2.

Your constructive critics and advise would be appreciated!

Last edited by Rufinus; 30th January 2018 at 02:01 PM. Reason: Pic resize
  Reply With Quote
Old 31st January 2018, 12:07 AM   #2
Gnobuddy is offline Gnobuddy  Canada
diyAudio Member
 
Join Date: Mar 2016
Quote:
Originally Posted by Rufinus View Post
What i'm not sure is about the value of R3, as i don't quite understand how it works. For me it looks like Q1 base biasing voltage with negative feedback from Q2 (emitter follower).
The easy way is to make R3 low enough to have negligible DC voltage drop across it. (Easier to do with high-beta transistors, and you can also raise R5 a lot, reducing Ic and therefore Ib of Q1. Maybe try making R5 ten times bigger, say 56k.)

Do the same with R4 as you did with R3 - make it small enough to have negligible DC voltage drop across it (probably already true in your circuit, particularly if you raise R5 as I suggested.)

With those conditions met, Q1's base is basically at 0.6 - 0.7 volts, and that is also the voltage across R7 (and sets emitter current of Q2).

With R6 having the same value as R7, it will also have the same 0.6 - 0.7 volts across it, so the emitter of Q2 should be sitting somewhere around 1.2 - 1.4 volts DC. If you want to change the DC voltage at the emitter of Q2, change R6 appropriately.

Keep in mind, you've created a high-gain circuit. I would guess a voltage gain of 40 - 50 times in Q2; AC voltage gain of a simple common-emitter stage like this is approximately 40 times the DC volts across the collector load resistor.

You are also running this high-gain circuit off a very small DC supply voltage. This means the microphone can only handle very quiet sounds - anything louder will cause heavy clipping at the output.

It looks like you've simulated this circuit in LTSpice, so you probably already have more accurate simulated numbers than my quick guesstimates.

-Gnobuddy
  Reply With Quote
Old 31st January 2018, 12:10 AM   #3
PRR is offline PRR  United States
diyAudio Member
 
PRR's Avatar
 
Join Date: Jun 2003
Location: Maine USA
Quote:
Originally Posted by Rufinus View Post
...simple electret mics... are just having too low gain....
....i tried to recalculate it for 3V...
The H1 input sensitivity should be ample to bring input stage hiss up into the ADC's (24-bit!) working zone. You may have to up the level more to get fully listenable final tracks; but if it is "on the tape" (file) then you can fix it later.

OK, you want a preamp anyway. You need to specify a gain: a little or a lot? You need to specify a maximum output needed (and if you have high gain, the output may be high) *before* you think about a battery. (If you specify nominal 1.24Vrms, you can't directly get that with a 3.0V battery.)

You need to specify the mike impedance to get a good hiss-fit. You need to know the H1 input Z (a low 2K!) to figure the output requirement.

What IS the gain of your plan? Without counting on thumbs, I note that neither plan has any NFB. So it will probably overload with 20mV-60mV inputs and approach 25% THD just before clipping. If your signals are truly tiny, this may be acceptable. Nearly any time I have had to use such a mike input, overload and distortion were issues.

There are IMHO better plans. But this "1 inch" snow is burying the dog and today is not good for an extended essay.

EDIT- Gno beat me by 3 minutes on some of these points.
  Reply With Quote
Old 31st January 2018, 02:43 AM   #4
Gnobuddy is offline Gnobuddy  Canada
diyAudio Member
 
Join Date: Mar 2016
Quote:
Originally Posted by PRR View Post
EDIT- Gno beat me by 3 minutes on some of these points.
Been home sick with the 'flu, got on diyAudio during a brief spell between fever in chills, happened to be at the right time to respond to this thread.

FWIW, I am always glad to see your viewpoint, which is always informative, and frequently thought-provoking.

-Gnobuddy
  Reply With Quote
Old 31st January 2018, 12:58 PM   #5
Rufinus is offline Rufinus  Germany
diyAudio Member
 
Join Date: Nov 2004
Location: Stuttgart
Thanks for your reply, guys!

Of course, i should have specified technical requirements in more detail.

Here they are: level of my electret mics, connected to H1 is in the region of -30 dB (one is even below -40) with a normal speech. The target is to bring it somewhere between -20 and -12 --> needed gain is about 30 (15 dB), please correct me if i'm wrong.
Noise level and distortion should be kept to the reasonable minimum.

So as the average output of electret mic is only about 2 mVpp then i'm targeting for 60 mVpp output of the amp. Let it be even 100 mVpp, i'm still safe with 3V battery.

Unfortunately i do not have exact data about different mics i have, but the average electret mic output impedance is about 2K (it is defined by FET load resistor, which is typically 2K).

Input impedance of H1 is low, about 2K, therefore emitter follower stage based on Q2.

Optimum input level of H1 concerning noise is about 40 - 55, it produces minimum hiss possible. Of course i could increase the H1 gain to 60-90 to make recording level acceptable, but it introduces a lot of hissing, therefore idea of own preamp.

Here is the the same schematics, but with added values:

Click the image to open in full size.

And here is the circuit with adjustment, as Gnobuddy suggestet:

Click the image to open in full size.

Here emitter of Q1 is sitting at about 1.5mV and base is at 505mV, i think it is a bit too low, or?

Here is also gain, simulated with my "original" values - looks like 30 to me, or?

Click the image to open in full size.

I would greatly appreciate advises regarding improvements, negative feedback in particular, or better circuits, as PRR proposes.

P.S. Somehow i sought that R3 if providing negative feedback..
  Reply With Quote
Old 31st January 2018, 04:00 PM   #6
Gnobuddy is offline Gnobuddy  Canada
diyAudio Member
 
Join Date: Mar 2016
Quote:
Originally Posted by Rufinus View Post
Here they are: level of my electret mics, connected to H1 is in the region of -30 dB (one is even below -40) with a normal speech.
I'm still confused by this. I owned a Zoom H1 for a while, and I don't remember having any problem getting adequate (speech) recording levels with the built-in electret microphones.

Are you at a great distance from the sounds you are trying to record?

Is the MOSFET in the electret mic properly biased with the external drain resistor to a small positive voltage? (It looks okay in your schematic.)

Quote:
Originally Posted by Rufinus View Post
the average electret mic output impedance is about 2K (it is defined by FET load resistor, which is typically 2K).
Sounds good, you want R3 to be much bigger than 2k, so it doesn't heavily load the electret output. I suggest R3=22k as a starting point.

Quote:
Originally Posted by Rufinus View Post
Here emitter of Q1 is sitting at about 1.5mV and base is at 505mV, i think it is a bit too low, or?
Your Q1 bias voltages look fine to me. C3 is doing nothing, and can be removed. Actually, I suggest removing R4 as well (replace with a wire, emitter of Q1 direct to ground. Then remove C5 as well.

With C5 removed, you will actually have negative feedback through R3. Your circuit now acts as a current-to-voltage converter, the signal current from the electret microphone flowing through R3 will set the AC output voltage. (And the signal voltage at Q2 emitter will be double that.)

Quote:
Originally Posted by Rufinus View Post
P.S. Somehow i sought that R3 if providing negative feedback..
In your circuit as shown so far, R3 is providing only DC feedback, which sets the operating point of Q1 and Q2.

There is no AC (signal) feedback through R3, because all AC signals have been removed by C5, which makes sure there is no signal voltage at the junction of R6 & R7.

Why not try the current-to-voltage version? Remove C3 and C5, short R4, set the AC gain with R3. Since you have actual negative feedback, distortion should be lower, and you can set the gain you want with the value of R3. (Some experimenting may be needed to get the right gain.)

One last comment about small-area microphone capsules: they generate an inherent noise because of statistical fluctuations in the number of air molecules bouncing off the diaphragm. It is the same effect as "rain on a tin roof", or shot noise in a circuit.

So small-area microphone capsules are inherently "hissy". If you add a lot of gain on top of that, they get even more hissy.

-Gnobuddy
  Reply With Quote
Old 31st January 2018, 07:55 PM   #7
PRR is offline PRR  United States
diyAudio Member
 
PRR's Avatar
 
Join Date: Jun 2003
Location: Maine USA
Quote:
Originally Posted by Rufinus View Post
...i could increase the H1 gain to 60-90 to make recording level acceptable, but it introduces a lot of hissing....
There's hiss everywhere.

Gain makes it louder, apparent.

System hiss level is normally dominated by input stage hiss.

So you are asserting that you (we?) can make an input stage with less hiss than the H1. In the past, this was sometimes true. Low-cost devices/designs sometimes fell short of near-ideal hiss performance. Today with modern miracle amplifier chips being promoted by several chipmakers, I am not so sure.

But *first*-- the input stage here is not the H1/DIY preamp but the FET inside the capsule. And these are often picked/biased for low current consumption, not lowest hiss. A higher current in the FET will probably be both a lower hiss voltage and a higher output level, win/win.

Above say 5V, you would not be able to influence the FET current with capsule supply voltage, it current-saturates like a pentode. But at a few Volts the FET is near or in its triode mode, and current falls with capsule supply voltage. IMHO 3V is below optimum. Recording with capsules, I always gave them like 7V. That was years ago and the FETs may be more special-purpose now. I suggest you try the capsule plus resistor on supplies from 2V to 12V (but watch the capsule terminal voltage does not go over rating, maybe 5V). Does the current increase? Does 9V give 2X the current of 3V? If so, that *may* be near 3dB lower hiss and a couple dB higher signal.

Snow to the west of me. Gotta get ready for work.
  Reply With Quote
Old 31st January 2018, 08:29 PM   #8
PRR is offline PRR  United States
diyAudio Member
 
PRR's Avatar
 
Join Date: Jun 2003
Location: Maine USA
I had assumed your SPICE FET was arbitrary. 2SK3372 is actually sold for electret heads and is used in some versions of a popular capsule. specs

Indeed the current will not rise much above the 3V level. However the gain spec gets significantly (>3dB) better. However the gain spec does not add up with the gm versus drain load (allow 80K drain resistance). So I do not know what it is telling me.

The noise spec is 4uV which is stunningly high (makes any likely next-stage hiss moot). However this is MAX and typical could be a little less or a lot less.

While pondering that I came across [url=http://www.johncon.com/john/wm61a/]a long essay on this capsule/FET and possible optimizations. Looks very worth a long read. So far I note that in Figure IV the 20K NFB resistor may be 2uV hiss and seems not-negligible to the total; few-K here may shave a little hiss.

These FET capsules distort (bottom-clip) at high level. I have used that to musical effect. I think your use is not at that level. That essay does discuss a clever linearization, as an alternative to the Linkwitz mod. Linkwitz actually moves the limits way further. Conover hardly moves the limits but greatly reduces the ~~5% THD just below clipping.
  Reply With Quote
Old 1st February 2018, 11:52 AM   #9
Rufinus is offline Rufinus  Germany
diyAudio Member
 
Join Date: Nov 2004
Location: Stuttgart
Quote:
Originally Posted by Gnobuddy View Post
I'm still confused by this. I owned a Zoom H1 for a while, and I don't remember having any problem getting adequate (speech) recording levels with the built-in electret microphones.

Are you at a great distance from the sounds you are trying to record?

Is the MOSFET in the electret mic properly biased with the external drain resistor to a small positive voltage? (It looks okay in your schematic.)


Why not try the current-to-voltage version? Remove C3 and C5, short R4, set the AC gain with R3. Since you have actual negative feedback, distortion should be lower, and you can set the gain you want with the value of R3. (Some experimenting may be needed to get the right gain.)

One last comment about small-area microphone capsules: they generate an inherent noise because of statistical fluctuations in the number of air molecules bouncing off the diaphragm. It is the same effect as "rain on a tin roof", or shot noise in a circuit.

So small-area microphone capsules are inherently "hissy". If you add a lot of gain on top of that, they get even more hissy.

-Gnobuddy
You are right, there is no issue with level of built-in mics of H1!
Problems start when i try different lavalier (tie clip) microphones. Some of them have to be powered by 1.5V LR44 cell, some are getting power from H1 (2.5V). What capsules they are actually using only God knows, therefore noise/hiss has to be qualified by ear, there is nothing much to do as to sort them and choose best sounding one.

Typical distance from source for lavalier mic is about 20-25 cm (10 inch for US guys ), which is not very far and built-in H1 mics manage it well.

So now i have adjusted the schematics per your advice, and it looks like this now:

Click the image to open in full size.

THD results: Total Harmonic Distortion: 0.020929%(0.022341%)
Looks OK to me.

With R3 at 56K i get 24mV output and Total Harmonic Distortion: 0.195695%(0.196154%), which is also OK in my opinion.

Noise performance i will have to evaluate "by ear".

Thanks a lot! The next step would be trying it on the breadboard and tweaking the R3/gain for H1!



P.S. One point, which troubles me a bit is the removed R4 - will we not get thermal instability? Or 60 ohm value of R4 was anyway not of much use?
C3 was there for AC gain of Q1 stage, or?

P.P.S. I was trying to follow the book (Art of electronics) with this design, that's why concerns and questions.
  Reply With Quote
Old 1st February 2018, 12:30 PM   #10
Rufinus is offline Rufinus  Germany
diyAudio Member
 
Join Date: Nov 2004
Location: Stuttgart
Quote:
Originally Posted by PRR View Post
I had assumed your SPICE FET was arbitrary. 2SK3372 is actually sold for electret heads and is used in some versions of a popular capsule. specs

Indeed the current will not rise much above the 3V level. However the gain spec gets significantly (>3dB) better. However the gain spec does not add up with the gm versus drain load (allow 80K drain resistance). So I do not know what it is telling me.

The noise spec is 4uV which is stunningly high (makes any likely next-stage hiss moot). However this is MAX and typical could be a little less or a lot less.

While pondering that I came across [url=http://www.johncon.com/john/wm61a/]a long essay on this capsule/FET and possible optimizations. Looks very worth a long read. So far I note that in Figure IV the 20K NFB resistor may be 2uV hiss and seems not-negligible to the total; few-K here may shave a little hiss.

These FET capsules distort (bottom-clip) at high level. I have used that to musical effect. I think your use is not at that level. That essay does discuss a clever linearization, as an alternative to the Linkwitz mod. Linkwitz actually moves the limits way further. Conover hardly moves the limits but greatly reduces the ~~5% THD just below clipping.
The part of the circuit, surrounded by the dotted line, is a close approximation of once famous Panasonic WM61-A capsule. They have used 2SK3372 FET for EU market. US capsules got some other FET.

Linwitz mod is surely worth to do, except it requires that lavalier mic actually has the WM-61A capsule and then you still have to get it out and modify it. As WM-61A is unfortunately discontinued, we have to take what is there. And what capsules and what FETs are in those lavalier mics is a very good question, as they usually have no marking. Besides, they are under powered with 1.5V LR44 (like Audiotechnica one).

But almost all of them need some more gain, therefore preamp, 3V power and then noise selection "by ear". 3V choice is based on 1/2AA Li‑MnO2 battery chemistry, which has pretty flat discharge curve, and the portability requirement. Preamp current consumption is modest enough to last for many hours of recording with over 850mAh battery capacity.

Another possibility is to use 4SR44 battery, it provides 6V and also relatively flat discharge (silver-oxide variant), but capacity is ~200mAh best case.
6V would probably improve noise problem, but battery lifetime is concern..
Still thinking..

Guys at johncon.com are interested in measurement mic, that is a bit overkill for my needs
  Reply With Quote

Reply


New old-school electret mic preampHide this!Advertise here!
Thread Tools Search this Thread
Search this Thread:

Advanced Search

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off

Forum Jump

Similar Threads
Thread Thread Starter Forum Replies Last Post
Op-Amp Preamp Circuit for Electret Mic ha9981 Analog Line Level 7 31st March 2013 04:40 AM
electret mic preamp for micro ADC hozone Analog Line Level 10 14th March 2013 01:40 PM
electret mic preamp gain emosms Analog Line Level 2 23rd September 2012 06:53 PM
electret Mic preamp teopoz Chip Amps 10 14th October 2007 09:19 AM
Electret mic preamp - will phono preamp work? darthmullet Analogue Source 6 12th October 2007 07:32 AM


New To Site? Need Help?

All times are GMT. The time now is 12:48 PM.


Search Engine Optimisation provided by DragonByte SEO (Pro) - vBulletin Mods & Addons Copyright © 2018 DragonByte Technologies Ltd.
Resources saved on this page: MySQL 15.79%
vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2018 DragonByte Technologies Ltd.
Copyright ©1999-2018 diyAudio
Wiki