DIY Budget Microphone

joodias

Member
2014-11-16 1:26 pm
Hello there,

I'm building a low cost mmicrophone, using an electret capsule. I intend to use Schoeps style circuit, similiar to the Alice mike. Here's the schematic:

electret_xlr2.gif


I plan to also include 2 22pF caps paralel to the 100K resistors. Also, I changed the leftmost cap to 100uF. The trannies used are going to be 5087 in smd package.

I have some questions:

1) Where the two 100K resistors connect togheter, in some designs (e.g. Alice Mic) it's added a 6K8 resistor from there to the 12V rail. As I understand, it is used to limit the current provided by phantom power. Yet, it seems like most phantom powering devices include this resistor internally. So, should I include it in the circuit or not?

2) The two biasing resistors are usually 2K2 Ohms, yet, some designs show different, higher values, like 5K6 (in both cases it is used a 2K2 Out Impedence capsule). What value should I use for a 2K2 IMpedance, 9.7mm capsule?

3) There are two variants of the capsule I intend to use. One has a -46dB sensistivity and the other -36dB, if I''m not mistaken. I'm going to be using the mic mainly for dialog and ambient sound recording (for films) and, ocasionally, music recording (as ambient mics, also for films). What would be the best fit?

It would be most helpful if you could answer these quesitons.

JDias


P.S. If there is interest in these mikes, once they're finished I'll post some photos, and the BOM and board if anyone desires to build them.
 

epicyclic

Member
2012-01-06 10:21 am
Hi
1)..yes there are 6K8 resistors in both legs of 48V phantom powering systems but this is an additional 6K8 to limit the current by virtue of the volt drop across it . If you powered your circuit from a 48V phantom system I have calculated it would draw near the maximum allowed 10mA , with the added 6K8 this is reduced to nearer 6mA . In summary if its 48V include the 6K8 and bypass it to the -ve rail with a 47uF cap..as the Schoeps circuit , if the phantom supply voltage is less then leave it out .

2)..I believe increasing the value of the 2K2s would increase the gain of the circuit and however reduce the voltage across the capsule . There is a balance to be had here between the two . The min volts across the capsule should be 1.2V and the max 10V. I would leave the 2K2s as they are to allow for variations in the capsules characteristics .
3)... I cant help you with this ..no experience.
 

joodias

Member
2014-11-16 1:26 pm
Thank you for the reply!

1) In summary if its 48V include the 6K8 and bypass it to the -ve rail with a 47uF cap..as the Schoeps circuit

Yes I intend to use 48V. So, an 100uF cap before the 6k8 resistor, one 47uF after it and a 100uF after the 330R would be a good coice or a bit overkill?
Also, I'm using a simplified circuit, but I see that in the original Schoeps they use diode after the 6k8 resistor to the emiter of each trannie. Would the circuit favor from adding them?
 

PRR

Member
Paid Member
2003-06-12 7:04 pm
Maine USA
> two biasing resistors are usually 2K2 Ohms

Exact value not important. 2K is usually a fine value.

> One has a -46dB sensistivity and the other -36dB

If they are low-price, buy both. I used to have a drawer full of $2 mike capsules. The specs will "often" show that a -46 capsule has wider bandwidth than a -36 capsule-- trade-off sensitivity to get more "flat" at each end. Also the -46 capsule will take louder sound (nominally 10dB more) before overload; the -36 may have lower hiss in very quiet situations. OTOH there are many internal details and sensitivity is NOT a good guide.

The mike as shown sucks nearly the maximum current allowed for Phantom (9ma against 10mA limit for true P48). Another 6.8K in there would drop total draw to ~~3mA. Either is fine for full-P48 sources. The added-6K8 may be a better choice for battery power P48 interface. Either way the output impedance is low-enough for any situation that does not "demand" transformer output.

> Schoeps they use diode

Phantom is not the only way. I have used Simplex power, and IIRC Schoeps sold a lot of gear that way. A Simplex power interface into this Phantom load might blow it up. The diode would block that. As I think it is very unlikely you will ever face a Simplex mike input, I would skip the diode.
 

epicyclic

Member
2012-01-06 10:21 am
Yes I intend to use 48V. So, an 100uF cap before the 6k8 resistor, one 47uF after it and a 100uF after the 330R would be a good coice or a bit overkill?
Also, I'm using a simplified circuit, but I see that in the original Schoeps they use diode after the 6k8 resistor to the emiter of each trannie. Would the circuit favor from adding them?

No I don't think that's overkill .
As PRR you don't need the 2 additional zeners . The voltage across each trannie will be defined by the base emitter voltage over the 10Ks and the multiplication factor of the 100Ks . I calculate the voltage to be approx. 7V across each.
 
Last edited:

joodias

Member
2014-11-16 1:26 pm
Thank you for your answer.

>
If they are low-price, buy both. I used to have a drawer full of $2 mike capsules.
.

Yes, they are indeed 2$ capsules. The only thing is that i won't have much time for testing, but I'll try to work it out. But, without speaking of the sensitivity, do you believe these capsules can make a decent mic? I don't expect it to be a U87, but I was hoping for it at least be near enough to a under 100$ rode video mic. Is it too much o ask for? :D

These are the capsules I plan to use:

AOM-6738L-R - PROJECTS UNLIMITED - MICROPHONE, OMNI LEADS, 9.7MM | Farnell element14 Portugal

AOM-6746P-R - PROJECTS UNLIMITED - MICROPHONE, OMNI PCB, 9.7MM | Farnell element14 Portugal
 

Lojzek

Member
2012-02-10 12:12 pm
Croatia
Those are not precision measurement mic capsules. Find the ones of 6 mm in diameter, something like Monacor MCE 4000. I have built one with a Panasonic WM61 a long time ago and this was a perfect measurement mic, which I compared to an Audiomatica Clio mic and these two performed equally well.
 

joodias

Member
2014-11-16 1:26 pm
Those are not precision measurement mic capsules. Find the ones of 6 mm in diameter, something like Monacor MCE 4000. I have built one with a Panasonic WM61 a long time ago and this was a perfect measurement mic, which I compared to an Audiomatica Clio mic and these two performed equally well.

I'm not exactly looking for a measurement mic. The only requisites I have for the capsule is that it has to be as quiet as possible and not to have any major frequency response spikes that are not easily fixed in a DAW and that could ruin the whole recording. I choose the 9.7mm one because I though it would be better for picking up medium/low end compared to a smaller capsule, to add some warmth.

My original choice was going to be the Primo EM172. I've heard amazing clips recorded with these, but for me to get them where I live they'd cost about 30$ or more. That would already defy the purpose of this build, which is to build a mic under 20$.

Do you believe the ones I chose are not suited for the job?
 

PRR

Member
Paid Member
2003-06-12 7:04 pm
Maine USA
The $2 Omni capsules can be quite excellent.

I used such mikes for years as high overheads for recording in a music recital hall. (With far less line-interface than you have.) Those recordings stood up well. The main flaw was the room was oddly shaped and loud ventilation. But still better than many public venues.

If you stuff them IN a saxophone, they distort.

There are two main "flavors". Some are flat well below 100Hz. Others roll-off below 300Hz, because that is what you want in an answering machine or other speech-only use. Panasonic's data sheets show this clearly. (Psonic split the business some years back, but the current maker uses the same part numbers.)

The top end has a mild rise then fall, like any condenser mike. I think the high-sensitivity ones tune near 8KHz, the low-sens ones more like 16KHz.

$2 cardioid mikes mostly have real bass fall-off, inherent in their small size and simple assembly. Not just fall-off, they go omni down low. If you need directionality, Primo is the type I favored.
 

PRR

Member
Paid Member
2003-06-12 7:04 pm
Maine USA
> add to Hot and Cold a 47R resistor in series and a 22nF cap to ground
> instead of using a zener, adding a regulator for loer noise


Why?
Why?

Are you expecting heavy RF fields?

The plan as shown has *worked* for many people. Even with some fancier capsules I used a Zener bypassed with a cap. The main hiss-source is still the teeny JFET inside the capsule.

Keep it simple. Stick with the known good plan.

AND- buy a lot of capsules and parts and try all different ways. "i won't have much time for testing" is not a good path to success. Good recording is constant trial and experience.