All OpAmp mic design (no FET at first stage)

[homero.leal]

Hi Guys,

I have been looking at some mic designs which use OpAmps, but all of them use a FET for the first input stage.

My question is, does anybody knows or can point to a mic design which works using OpAmps instead of a FET for the first stage, (I mean, receiving signal from the capsule) and which also could work with 48V phantom power?

May be this will be asking too much, since i have read that OpAmps have a higher current draw than FETs, but anyway... if some of you guys could help, It will be greatly appreciated!

Thank you in advance, and best regards!

HL

 

[Khron]

Blasphemy!... https://youtu.be/tSajE72Tpjs?t=127

Jokes aside, if memory serves, i think some CAD mics (Ex00's i think), and i've read some Neumann TLM's are also fully "chipped".

Some poking around recordinghacks.com confirmed i'm not (yet fully) senile
http://recordinghacks.com/images/mic_extras/cad/E300-schematic.png

 

[homero.leal]

Thank you Khron... It looks massive and pretty complex...

May be that's the reason FET's are prefered for the first stage...

Anyway... I'll take a look to see if I can understand something of it..

Thank you again, and best regards!

HL

 

[Khron]

No sweat

The bottom part's the oscillator + voltage multiplier for the capsule bias voltage.

U1A/B seem to buffer/amplify each side of the capsule separately, U2 being DC-servos for each half of U1.

U3A is the output mixer-buffers for "blending" the two sides of the capsules, and U3B "mirrors" the signal.

I'm a bit confused about those 1uF caps in the feedback loops of U1 - that would mean a high-shelf attenuation that's VERY low in frequency. For some reason, they wanted lots of feedback only at really low frequencies?

 

[Matt Nolan]

I'm a bit confused about those 1uF caps in the feedback loops of U1 - that would mean a high-shelf attenuation that's VERY low in frequency. For some reason, they wanted lots of feedback only at really low frequencies?

No. That would be true if the source was resistive. But the source is capacitive. So, it is just like a flat amplifier with a resistive input and resistive feedback, but instead of resistors you have capacitors. The 1uF provide flat feedback. This is called a "charge amplifier"

 

[Khron]

Great point, the capacitive nature of the source slipped my mind

 

[homero.leal]

Thank you Matt and Khron...

I found another blasphemous on the family... It's E300 little brother.. CAD E100, and Recording Hacks also has the schematic:

http://cdn.recordinghacks.com/images//mic_extras/cad/E100-schematic.png

Looks simpler than E300, since E100 it is not Multi-pattern. Also, capsule looks to be an electret.

Removing power, roll off, and HP Filter switches and components, schematic should be more manageable.

Hope so!

Best regards...

HL

 

[ln76d]

eee...
Opamps...
Go for microprocessor, make some DSP, put sh*tty capsule and make switchable emulations of U47, U67, C12, M251 etc.
This would be microphone of the XXI era - or maybe it should be i-XX

 

[homero.leal]

Also... why not train a neural network to learn vintage mic signal contitioning, and reproduce them from plain sources? Just thinking...

The real reason I think, is that analog components are easier to assembly for DIYers...  Trust me, programing fourier equations to deal with audio signals, in custom DSPs is out  of reach for mere mortals. 

And by the way... some guys did somting very intesting similar to what you just mentioned...
http://vintageking.com/slate-digital-vms-virtual-microphone-system

 

[Khron]

That Slate mic is exactly what i thought of too, when reading ln76d's message

https://www.youtube.com/watch?v=pdLGRlqiW60

 

[ln76d]

Blaaah...
I don't buy that 01110101011100 johny bravo emulator microphone.
What next?
Vst plugin with Jackson or Sinatra voice emulation?
We need to wait for apfel i-mic!
Probably after every sung verse you will have to pay in apfel shop for possibility to sing another.
But who cares?
Most important is that it will be made from white plastic!

If i wouldn't be so leazy i would use tape for recordings...

BTW. i was learning programming DSP and i hate that sh*t - it isn't so complicated but i hate all kind of programming

 

[Pip]

Check this thread out for some things in the ball park.

The reason it is always a FET or a Tube has to do with the high input impedance requirements of the system if you will. So Opamp used will probably be a FET based input. There is the high frequency option to get around this. See PDF attached!



http://groupdiy.com/index.php?topic=61037.msg773076#msg773076

 

[homero.leal]

Thank you very much Pip...

Great references... I'm having a read of them. Thread discussion is very interesting.

I wonder if CAD's Ex00 schematics are trustable... I mean, I haven't found a reference of some DIYer who has built a mic with this schematic. It will be great if somebody could confirm that...

I read an article which stated that OpAmps with Bootstraping could be a great replacement for FET at first stage (low distortion, low noise and clarity), but the schematic on the article used external, plus and minus power sources.

I have nothing against externally powering mics at a studio room, but I would like a simpler and easier to use build, and that's the reason for looking about a 48v Phantom Powered, all OpAmp mic design.

Thank you again for your kind help, and best regards!

HL

 

[Khron]

For a starting point(?), i've filtered TI's available dual opamps with JFET input, under 1.5mA quiescent current, in DIP or SOIC packages.

You can sort them in order of quiescent current in the table, but i think you'd have to read through the datasheet for noise specs etc

http://www.ti.com/lsds/ti/amplifiers-linear/operational-amplifier-op-amp-products.page#p480=2;2&p1typ=0.000351;1.5&p2954=PDIP;SOIC&p89=FET

The at least theoretical max current (short-circuit) you can draw from standard phantom power is about 14mA, if my math isn't too rusty. Drawing even 2-3mA through the pair of 6.8k resistors should drop the voltage, in the mic, down towards the 30v area, so most opamps should be quite "safe", i'd say

 

[Pip]

Thank you very much Pip...

Great references... I'm having a read of them. Thread discussion is very interesting.

I wonder if CAD's Ex00 schematics are trustable... I mean, I haven't found a reference of some DIYer who has built a mic with this schematic. It will be great if somebody could confirm that...

I read an article which stated that OpAmps with Bootstraping could be a great replacement for FET at first stage (low distortion, low noise and clarity), but the schematic on the article used external, plus and minus power sources.

I have nothing against externally powering mics at a studio room, but I would like a simpler and easier to use build, and that's the reason for looking about a 48v Phantom Powered, all OpAmp mic design.

Thank you again for your kind help, and best regards!

HL

The CAD X00 solution is actually on board rechargeable NiCad batteries. That is how the get the Power scheme. The P48 just charges the batteries. As described in that thread I believe by Mr. Bock states the CAD mics are overly ornate in design. Besides the opamp based concept there are also no transformers and no coupling caps between the impedance amp and the output amp due to the servo design for removing DC offset. There are caps on the output however.

As you can see this is a topic that I am very interested in. But keep in mind that it is more to do with getting the capsule polarization voltages and amplifier current where they need to be then doing away with FET's.

 

[Pip]

For a starting point(?), i've filtered TI's available dual opamps with JFET input, under 1.5mA quiescent current, in DIP or SOIC packages.

You can sort them in order of quiescent current in the table, but i think you'd have to read through the datasheet for noise specs etc

http://www.ti.com/lsds/ti/amplifiers-linear/operational-amplifier-op-amp-products.page#p480=2;2&p1typ=0.000351;1.5&p2954=PDIP;SOIC&p89=FET

The at least theoretical max current (short-circuit) you can draw from standard phantom power is about 14mA, if my math isn't too rusty. Drawing even 2-3mA through the pair of 6.8k resistors should drop the voltage, in the mic, down towards the 30v area, so most opamps should be quite "safe", i'd say

The P48 Standard is 10MA and that is only if the rails can support every input having a mic of the P48 need. The rails tend to droop after a fashion.

https://en.wikipedia.org/wiki/Phantom_power

 

[Khron]

I just divided 48v to 3.4k, and my calculator said 14,11mA - and as i mentioned, that's theoretical (at least) short-circuit, so 0v at the load

 

[homero.leal]

Thank you again Pip and Khron!

Pip... I didn't know this models have rechargable batteries... that's a interesting twist on their design. I'm not a mic guru, but I haven't seen this before, It's clever...  Also,  when you say "servo", are you refering to the negative feedback approach on the design? Haven't heard this term before. I thought it was associated to motors.  ???

Khron...  I was thinking on the ADA4841-2 from Analog Devices. I read an article from Scott Wurcer, and he uses this OpAmp in his designs. I'll take a look at the TI list you kindly shared to see if there is a similar option from them.

Best regards!

HL

 

[Khron]

Some details on CAD mics (the ones already discussed are at the bottom, in the "discontinued" section) - http://recordinghacks.com/microphones/cad

http://waynestegall.com/audio/simpledcservos.htm - DC servos in audio circuits are basically for nulling out the DC offset on their output

Well, like i said, the TI site was just a starting point, i just knew / remembered they had a half-decent table-sorting thing on their site

 

[ln76d]

If you guys thinking about opamps in microphones, maybe some project of universal simple and usefull opamp DC converter?
Like for example rode nt1a?
I could rip off their circuit - is simple, but maybe there's a place for diy on this?