Author Topic: EEVblog #611 - Electret Microphone Design  (Read 32795 times)

0 Members and 3 Guests are viewing this topic.

Online EEVblogTopic starter

  • Administrator
  • *****
  • Posts: 37661
  • Country: au
    • EEVblog
EEVblog #611 - Electret Microphone Design
« on: May 03, 2014, 12:12:50 pm »
Part 5 - Doug Ford explains the design of Electret Microphone circuits, and all the associated traps:
Transconductance, correct biasing, and temperature effects, leaky JFETS, non-linear clipping & headroom, current noise and voltage noise, and pseudo bias resistors.
Also, the non-intuitive concept of higher resistance = less noise, and how this applies to electret biasing.
And what is a soggy circuit board and why is it important?

 

Offline electronics man

  • Frequent Contributor
  • **
  • Posts: 686
  • Country: gb
Re: EEVblog #611 - Electret Microphone Design
« Reply #1 on: May 03, 2014, 01:30:18 pm »
Very nice video very useful  :-+
follow me on twitter @get_your_byte
 

Offline mikeselectricstuff

  • Super Contributor
  • ***
  • Posts: 13695
  • Country: gb
    • Mike's Electric Stuff
Re: EEVblog #611 - Electret Microphone Design
« Reply #2 on: May 03, 2014, 02:34:22 pm »
When I've taken electret mics apart, I've wonderd about the lack of a bias resistor - I wonder if they use a special FET with some resistance built in?
Youtube channel:Taking wierd stuff apart. Very apart.
Mike's Electric Stuff: High voltage, vintage electronics etc.
Day Job: Mostly LEDs
 

Offline Legion

  • Frequent Contributor
  • **
  • Posts: 360
Re: EEVblog #611 - Electret Microphone Design
« Reply #3 on: May 03, 2014, 03:07:32 pm »
When you guys are talking about how they don't use bias resistors anymore and then about how a high value gate resistor reduces noise it got me wondering, when does "high resistance" become "open circuit"? If the answer is "when current stops flowing", what constitutes a stoppage of current flow? Does 1 electron/second still count as "current"?

I guess what I'm getting at is, is it even meaningful to talk about a 100 teraohm resistor? Or is it an open circuit at that point?
 

Online ejeffrey

  • Super Contributor
  • ***
  • Posts: 3685
  • Country: us
Re: EEVblog #611 - Electret Microphone Design
« Reply #4 on: May 03, 2014, 03:37:06 pm »
You can have a resistance as high as you want as long as it is 'ohmic.' i.e., the current is proportional to voltage down to arbitrarily small voltages.  This depends on the semiconductor physics of the material, not the value of the resistance.  If you take a bunch of megaohm resistors in series for instance, you can get as large a resistance as you want.  Insulators on the other hand, are simply not ohmic, at low voltages zero current flows.  At high voltages current flows due to various mechanisms like field emission and electrical breakdown.  Also, while the bulk of a material may not conduct, you can still get conduction along the surface due to water films.  So when you see a specification for 'insulation resistance' it doesn't mean that the material acts as an ohmic conductor, but is rather a specification for maximum leakage current at some specific voltage due to those secondary mechanism.
 

Online NiHaoMike

  • Super Contributor
  • ***
  • Posts: 8973
  • Country: us
  • "Don't turn it on - Take it apart!"
    • Facebook Page
Re: EEVblog #611 - Electret Microphone Design
« Reply #5 on: May 03, 2014, 03:52:34 pm »
Those cheap microphones can easily be hacked for better performance.
http://sound.westhost.com/project93.htm
Cryptocurrency has taught me to love math and at the same time be baffled by it.

Cryptocurrency lesson 0: Altcoins and Bitcoin are not the same thing.
 

Offline jaxbird

  • Frequent Contributor
  • **
  • Posts: 778
  • Country: 00
Re: EEVblog #611 - Electret Microphone Design
« Reply #6 on: May 03, 2014, 03:58:51 pm »
Big thanks to Doug Ford for this excellent series  :-+

Very much enjoyed it. Doug is clearly very knowledgeable, experienced and a great communicator.

Hope we get to hear much more from Doug :)

For those interested in the complex design aspects of audio, I can recommend having a good look at the www.linkwitzlab.com website.

Although I'm sure Dave would instantly label Mr. Linkwitz as an audiophool :P He has actually been designing instruments for HP for about 40 years or so. And he has received a lot of recognition together with Mr. Riley for their work in filters.
Analog Discovery Projects: http://www.thestuffmade.com
Youtube random project videos: https://www.youtube.com/user/TheStuffMade
 

Offline Richard Crowley

  • Super Contributor
  • ***
  • Posts: 4317
  • Country: us
  • KJ7YLK
Re: EEVblog #611 - Electret Microphone Design
« Reply #7 on: May 03, 2014, 05:07:13 pm »
When I've taken electret mics apart, I've wonderd about the lack of a bias resistor - I wonder if they use a special FET with some resistance built in?
A VERY common transistor used in cheap electret mic capsules is the 2SK596.  It has built-in bias resistor AND protective diode as you can see within the dotted line portion of the circuit below...



http://www.openmusiclabs.com/wp/wp-content/uploads/2011/03/2SK5961.pdf
 

Offline calexanian

  • Super Contributor
  • ***
  • Posts: 1881
  • Country: us
    • Alex-Tronix
Re: EEVblog #611 - Electret Microphone Design
« Reply #8 on: May 03, 2014, 06:01:21 pm »
I think we need an extended series of how things really work for the wanker musician and audiofool. It will end the ever present questions from said people when they learn I am in electronics, only to tell me how wrong my explanation is! Evidently these people think physics can lie!
Charles Alexanian
Alex-Tronix Control Systems
 

Offline jaxbird

  • Frequent Contributor
  • **
  • Posts: 778
  • Country: 00
Re: EEVblog #611 - Electret Microphone Design
« Reply #9 on: May 03, 2014, 09:37:14 pm »
I think we need an extended series of how things really work for the wanker musician and audiofool. It will end the ever present questions from said people when they learn I am in electronics, only to tell me how wrong my explanation is! Evidently these people think physics can lie!

Maybe you can be a bit more specific on what you have in mind? is the Doug/Dave presentation bogus?

Analog Discovery Projects: http://www.thestuffmade.com
Youtube random project videos: https://www.youtube.com/user/TheStuffMade
 

Offline Richard Crowley

  • Super Contributor
  • ***
  • Posts: 4317
  • Country: us
  • KJ7YLK
Re: EEVblog #611 - Electret Microphone Design
« Reply #10 on: May 03, 2014, 09:48:57 pm »
The characteristics of the capsule itself are an important factor, of course. But the mechanical design of the complete microphone head, even the pattern of the grille, etc. have major effects on the performance of any microphone. 

And then there are decades of "tradition" and "superstition" in the entertainment biz about what microphone models are good for various applications. And there is a HUGE influence where wanna-be musos use a particular microphone (or guitar or whatever)  because their favorite rock star uses that product. 

Rode is a relative newcomer on the scene of pro microphones. But they have made some significant contributions to the body of the world's better microphones. I am still waiting for Dave's promise of the discussion of designing very low-noise circuits.

And I found the discussion of selecting the load resistance and supply voltage very interesting.  But I'd like to know why some of the microphones at the very tip of the pinnacle of the very best microphones on the planet (DPA high-voltage mics) take 140V power supplies, even though they are electret and presumably don't even use such a remarkably high voltage to bias the capsule, or even to operate a firebotle (as they are solid-state).
 

Offline jaxbird

  • Frequent Contributor
  • **
  • Posts: 778
  • Country: 00
Re: EEVblog #611 - Electret Microphone Design
« Reply #11 on: May 03, 2014, 09:57:46 pm »
In my experience, all my quality microphones need the 48v phantom voltage from my mixer.

That has not really covered by Doug in this series so far.

Analog Discovery Projects: http://www.thestuffmade.com
Youtube random project videos: https://www.youtube.com/user/TheStuffMade
 

Offline Richard Crowley

  • Super Contributor
  • ***
  • Posts: 4317
  • Country: us
  • KJ7YLK
Re: EEVblog #611 - Electret Microphone Design
« Reply #12 on: May 03, 2014, 10:12:11 pm »
48V was the "standard" established back in the early 1960s by Neuman when they made the first solid-state mics for Norwegian Broadcasting Corp.  Apparently NRK already had a good 48V supply (for their emergency lighting) and they just used it to power the microphones.

Relatively few modern microphones need the full 48V to meet their published specs. A great many (especially electret) will operate perfectly fine on 24V or even 12V. Of course, the non-electret mics that operate on phantom voltage typically use the full voltage (nearly 48V after bypass/filtering) as the bias voltage across the condenser capsule.  Here is the classic "Schoeps circuit"...



This is Scott Dorsey's modification for the budget-priced true condenser mics, notably the Oktava products from Russia.

The 1/2/3 pins at the right are the XLR output connector. R5 and R7 collect the phantom voltage directly, and filter/bypass it across C5, then persent it directly to the capsule through the R9, 1 gigohm resistor. The two terminals at the left go to the mic capsule, of course.  Q3 provides the necessary impedance conversion, as well as the "phase-splitting" so that proper anti-phase outputs are presented through emitter-follower transistors Q1/Q2 to the balanced mic line on XLR pins 2 and 3.
 

Offline jaxbird

  • Frequent Contributor
  • **
  • Posts: 778
  • Country: 00
Re: EEVblog #611 - Electret Microphone Design
« Reply #13 on: May 03, 2014, 10:28:33 pm »
48V was the "standard" established back in the early 1960s by Neuman when they made the first solid-state mics for Norwegian Broadcasting Corp.  Apparently NRK already had a good 48V supply (for their emergency lighting) and they just used it to power the microphones.

Relatively few modern microphones need the full 48V to meet their published specs. A great many (especially electret) will operate perfectly fine on 24V or even 12V. Of course, the non-electret mics that operate on phantom voltage typically use the full voltage (nearly 48V after bypass/filtering) as the bias voltage across the condenser capsule.  Here is the classic "Schoeps circuit"...



This is Scott Dorsey's modification for the budget-priced true condenser mics, notably the Oktava products from Russia.

The 1/2/3 pins at the right are the XLR output connector. R5 and R7 collect the phantom voltage directly, and filter/bypass it across C5, then persent it directly to the capsule through the R9, 1 gigohm resistor. The two terminals at the left go to the mic capsule, of course.  Q3 provides the necessary impedance conversion, as well as the "phase-splitting" so that proper anti-phase outputs are presented through emitter-follower transistors Q1/Q2 to the balanced mic line on XLR pins 2 and 3.

Thanks for the nice schematic  :-+

I'm sure my mics would work at a lower voltage with minor modifications, I was mostly wondering where the 48v came from. I'm surprised that we must blame the Norwegians for it :D

Well then I guess the other Scandinavian Bruel and Kjaer have some blame as well.

Cheers

Analog Discovery Projects: http://www.thestuffmade.com
Youtube random project videos: https://www.youtube.com/user/TheStuffMade
 

Offline Richard Crowley

  • Super Contributor
  • ***
  • Posts: 4317
  • Country: us
  • KJ7YLK
Re: EEVblog #611 - Electret Microphone Design
« Reply #14 on: May 03, 2014, 10:34:37 pm »
I was mostly wondering where the 48v came from. I'm surprised that we must blame the Norwegians for it :D
Well then I guess the other Scandinavian Bruel and Kjaer have some blame as well.

Well DPA (Danish Pro Audio) are an offshoot from B&K because many recording engineers (particularly those of us who do classical music recording) started using B&K lab measurement mics, and DPA took over and made a range of products more suitable for audio recording (vs. laboratory measurements).
 

Offline jaxbird

  • Frequent Contributor
  • **
  • Posts: 778
  • Country: 00
Re: EEVblog #611 - Electret Microphone Design
« Reply #15 on: May 03, 2014, 10:47:14 pm »

Well DPA (Danish Pro Audio) are an offshoot from B&K because many recording engineers (particularly those of us who do classical music recording) started using B&K lab measurement mics, and DPA took over and made a range of products more suitable for audio recording (vs. laboratory measurements).

Have to admit I have no knowledge of DPA, but I do remember buying a B&K mic with psu and certificate at some obscene amount many years ago.

I can however appreciate a good classical recording with a full ~110dB dynamic range any time of the day :)

Analog Discovery Projects: http://www.thestuffmade.com
Youtube random project videos: https://www.youtube.com/user/TheStuffMade
 

Offline calexanian

  • Super Contributor
  • ***
  • Posts: 1881
  • Country: us
    • Alex-Tronix
Re: EEVblog #611 - Electret Microphone Design
« Reply #16 on: May 03, 2014, 11:14:35 pm »
I think we need an extended series of how things really work for the wanker musician and audiofool. It will end the ever present questions from said people when they learn I am in electronics, only to tell me how wrong my explanation is! Evidently these people think physics can lie!

Maybe you can be a bit more specific on what you have in mind? is the Doug/Dave presentation bogus?

Oh no. Quite the opposite! This is how things need to be explained. Doug is dead on all of these things. His few minutes of explaining how things work is worth a hundred "Expert" opinions of how things "Sound" Objective, not subjective. Musicians and audiophiles are more into the psychological value of a piece of equipment than any quantitative measurements or performance. I find that if you explain how something works rather than explain how it "sounds" the person will be able to understand why things are the way they are and can figure out on their own what is quality and legitimate and what is bullshit.
Charles Alexanian
Alex-Tronix Control Systems
 

Offline jaxbird

  • Frequent Contributor
  • **
  • Posts: 778
  • Country: 00
Re: EEVblog #611 - Electret Microphone Design
« Reply #17 on: May 03, 2014, 11:44:34 pm »
Oh no. Quite the opposite! This is how things need to be explained. Doug is dead on all of these things. His few minutes of explaining how things work is worth a hundred "Expert" opinions of how things "Sound" Objective, not subjective. Musicians and audiophiles are more into the psychological value of a piece of equipment than any quantitative measurements or performance. I find that if you explain how something works rather than explain how it "sounds" the person will be able to understand why things are the way they are and can figure out on their own what is quality and legitimate and what is bullshit.

Nice, can't disagree with that :D

It's often a fine line between bullshit and legitimate, the bullshitters are unfortunately often clever enough these days to include some truth in their bullshit.

But personally my stance is that, hey if anyone feel their gold/silver/super pure copper whatever cable etc. makes them enjoy their music more, then who am I to argue? it's not like there are not enough warnings/ridicule already online, so let them have their fun, it's their money to spend on whatever they wish.

Analog Discovery Projects: http://www.thestuffmade.com
Youtube random project videos: https://www.youtube.com/user/TheStuffMade
 

Online EEVblogTopic starter

  • Administrator
  • *****
  • Posts: 37661
  • Country: au
    • EEVblog
Re: EEVblog #611 - Electret Microphone Design
« Reply #18 on: May 03, 2014, 11:55:02 pm »
In my experience, all my quality microphones need the 48v phantom voltage from my mixer.
That has not really covered by Doug in this series so far.

I thought we did that already?
If not, it's covered in material I haven't released yet. I definitely remember Doug covering that.
The problem with this series is that all the raw content is all a bit higgledy-piggledy. I have to edit together what segments I can in a somewhat organised fashion. I hope I don't miss anything because of that. 48V phatom power is in there somewhere...
 

Offline calexanian

  • Super Contributor
  • ***
  • Posts: 1881
  • Country: us
    • Alex-Tronix
Re: EEVblog #611 - Electret Microphone Design
« Reply #19 on: May 03, 2014, 11:55:26 pm »
Oh no. Quite the opposite! This is how things need to be explained. Doug is dead on all of these things. His few minutes of explaining how things work is worth a hundred "Expert" opinions of how things "Sound" Objective, not subjective. Musicians and audiophiles are more into the psychological value of a piece of equipment than any quantitative measurements or performance. I find that if you explain how something works rather than explain how it "sounds" the person will be able to understand why things are the way they are and can figure out on their own what is quality and legitimate and what is bullshit.

Nice, can't disagree with that :D

It's often a fine line between bullshit and legitimate, the bullshitters are unfortunately often clever enough these days to include some truth in their bullshit.

But personally my stance is that, hey if anyone feel their gold/silver/super pure copper whatever cable etc. makes them enjoy their music more, then who am I to argue? it's not like there are not enough warnings/ridicule already online, so let them have their fun, it's their money to spend on whatever they wish.

I understand. Hey, whatever makes you happy after all!
Charles Alexanian
Alex-Tronix Control Systems
 

Offline calexanian

  • Super Contributor
  • ***
  • Posts: 1881
  • Country: us
    • Alex-Tronix
Re: EEVblog #611 - Electret Microphone Design
« Reply #20 on: May 03, 2014, 11:59:46 pm »
In my experience, all my quality microphones need the 48v phantom voltage from my mixer.
That has not really covered by Doug in this series so far.

I thought we did that already?
If not, it's covered in material I haven't released yet. I definitely remember Doug covering that.
The problem with this series is that all the raw content is all a bit higgledy-piggledy. I have to edit together what segments I can in a somewhat organised fashion. I hope I don't miss anything because of that. 48V phatom power is in there somewhere...

I don't think any equivalence of quality to having or not having phantom power was really stated, or can be.. They are just simply different methods of operation. This relates more to the virtues of dynamic vs. condenser type microphones.. Not really in the scope of this series. This is more informational by means of how they work, not how they sound or any relationship to quality or usefulness. Your mileage will vary!~
Charles Alexanian
Alex-Tronix Control Systems
 

Offline pgross

  • Contributor
  • Posts: 32
  • Country: dk
Re: EEVblog #611 - Electret Microphone Design
« Reply #21 on: May 04, 2014, 01:18:15 am »
Kudos to Dough and Dave for the mic series  :-+

Quote
And I found the discussion of selecting the load resistance and supply voltage very interesting.  But I'd like to know why some of the microphones at the very tip of the pinnacle of the very best microphones on the planet (DPA high-voltage mics) take 140V power supplies, even though they are electret and presumably don't even use such a remarkably high voltage to bias the capsule, or even to operate a firebotle (as they are solid-state).

Using 140V, (actually 130V in the DPA Mic. example) instead of 48V, will provide higher SPL handling capability of the built-in mic amplifier. For special purposes; like measurements, high SPL sound recordings in the 154-168 dB range (depending of capsule), that extra headroom might come in handy. The DPA "High Voltage" mic series must be connected to a dedicated high voltage mic. amplifier - HMA. Beside the 130V, this HMA also features a 200V supply, used for powering external polarizing condenser capsules.

Typically, it is necessary practice to attenuate the output by 20-40 dB on the HMA. Otherwise the input of a recorder or mixing console will be overloaded when the mic is exposed to these very high SPLs.

 :palm:

/ Peter
Up & running!
 

Offline Richard Crowley

  • Super Contributor
  • ***
  • Posts: 4317
  • Country: us
  • KJ7YLK
Re: EEVblog #611 - Electret Microphone Design
« Reply #22 on: May 04, 2014, 02:48:46 am »
Using 140V, (actually 130V in the DPA Mic. example) instead of 48V, will provide higher SPL handling capability of the built-in mic amplifier. For special purposes; like measurements, high SPL sound recordings in the 154-168 dB range (depending of capsule), that extra headroom might come in handy. The DPA "High Voltage" mic series must be connected to a dedicated high voltage mic. amplifier - HMA. Beside the 130V, this HMA also features a 200V supply, used for powering external polarizing condenser capsules.

Typically, it is necessary practice to attenuate the output by 20-40 dB on the HMA. Otherwise the input of a recorder or mixing console will be overloaded when the mic is exposed to these very high SPLs.

Yeah, that is the standard response. But I'm not buying it. No other part of the audio chain requires more than 30-50V max (from - rail to + rail) even for +24dBu line level.  What exactly is the point of generating perhaps "line-level" signals (which are perhaps 1~2% of 130V) and then knocking it back down to "mic-level"?  Something is just fundamentally goofy here.
 

Offline pgross

  • Contributor
  • Posts: 32
  • Country: dk
Re: EEVblog #611 - Electret Microphone Design
« Reply #23 on: May 04, 2014, 02:54:17 pm »
Quote
Yeah, that is the standard response. But I'm not buying it. No other part of the audio chain requires more than 30-50V max (from - rail to + rail) even for +24dBu line level.  What exactly is the point of generating perhaps "line-level" signals (which are perhaps 1~2% of 130V) and then knocking it back down to "mic-level"?  Something is just fundamentally goofy here.

The 130V supply is derived – I think, from the older days where the high voltage was needed as anode voltage to amplifiers with tubes.
In more contemporary semiconductor designs, this higher voltage is needed where the std. P48 Volt becomes inadequate – theoretically at least.

Let say, that we have a capsule with a sensitivity of 10 mV/Pa. If this capsule is exposed to its stated 168 dB absolute max SPL handling with THD < 1 %, the output voltage is 50.12 V peak:  94 dB – 168 dB = 74 dB. Where, inv log(74/20)10-3 = 50.12 V.

A microphone connected to a IEC 61938 compliant power supply with a current draw of 2.5 mA will sink the phantom voltage to 31V.
In other words, the voltage provided will compromise the signal handling capability of the amplifier, resulting in audio distortion.

Common practice is to attenuate (PAD) the mic output signal by 10–20 dB. Of cause it dosent help much if the amplifier is already distorted.
On some mics, a PAD is placed in the front end circuit, unfortunately this is the worst place to attenuate the signal with respect to SNR.

For most users, the “high voltage” principle is merely a marketing ploy in order to justify esoteric audio gear – you know, the average Joe equipment can’t possibly satisfy the discerning audiophile… ;)

/Peter
Up & running!
 

Offline azi

  • Contributor
  • Posts: 16
Re: EEVblog #611 - Electret Microphone Design
« Reply #24 on: May 15, 2014, 02:34:13 am »
When I've taken electret mics apart, I've wonderd about the lack of a bias resistor - I wonder if they use a special FET with some resistance built in?

I think it's very unlikely that the FET has a built-in resistor or ohmic resistance for creating gate bias. Semiconductor manufacturing is very good for making diodes and transistors, but horrible for resistors. Unless things have changed since I was doing chip design (a long time ago), the way you make a resistor is with a wire of polysilicon, which has a resistance of about 20 ohms per square. The 2SK596 datasheet (see the link in Richard Crowley's post), the input resistance is listed as 25 Mohms. That would require a very long wire.

Also on the datasheet, there are both a "protection diode" and "resistor" shown connecting the FET's source to its gate. I *suspect* that these are shown as a disguise for using the reverse leakage of the diode and/or the gate's p-n junction to bias the gate to 0 volts.  The two devices might be manufactured together into an odd kind of jFET with an extra area of p-n junction continuing outside the gate, onto the source of the device, to increase the reverse leakage sufficiently. (However, this is a guess and the manufacturer may have found another way to make a "resistor"! Unfortunately, I lack sufficient test equipment to test my hypothesis. Mike, Dave and others can try it if they want. ;-)

I got some CUI electret capsules in the mail today from Digi-Key and took one apart. What I found matched the pictures on this web page almost exactly:

http://www.openmusiclabs.com/learning/sensors/electret-microphones/

I thought that might help others who are following this thread. It is the same website that hosts the 2SK596 datasheet.

It's pretty easy to remove the transistor with intact leads if you want to play with one and explore this topic further.
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf