Made a preamp but have issue with noise

[inklen]

I made this pleamp - https://www.ti.com/lit/ug/tidu765/tidu765.pdf - and it kind of works. But I made it with another mic model - https://mouser.com/datasheet/2/670/cmc-6027-24t-1777130.pdf. The difference is sensitivity - the original design mic has -42 dB and the one I used -26 dB. Everything else seems to be the same.

Now, since I just wanted to make it work as a prototype (I'm still learning, slowly), the mic on my pcb is soldered with a pair of wires ~25 cm / 10" long. And when I connect 9V battery I start hearing only some kind of noise. But when I hold the mic capsule with my fingers I hear everything clearly. Once I leave it everything stops working again. The sound level is the same.

I honestly want to know the reason for this. So thank you in advance.

Alex

 

[schmitt trigger]

As you can see from the second page of the datasheet, the mic's case is a shield. The electret itself is very high impedance and will pick up any stray noise. And even though it is buffered by a JFET, its output will still pick up noise with such long leads that you got..
As you can see in TI's app note, the microphone is soldered directly to the board.

 

[inklen]

Thank you! Let me try that.

BTW, does it mean that if I want to leave the wire long I should use some kind of shielded pair? Will it work? ... I need to read something about it.

 

[BradtheRad]

mic on my pcb is soldered with a pair of wires ~25 cm / 10" long

This length of wire unshielded is likely to pick up hum and noise. So yes indeed, you should use shielded cable.

 

[inklen]

Made it short and now there's no noise. But now here's another issue - I use in-ear simple headphones for testing it, and when I just turn it on the sound level is very good. Then, after ~10 min it fades away. Is this because of headphones (and not line-in), or could it be something else?

Thank you for taking your time answering my ... questions.

 

[barry]

Sounds like something is ooverheating or your battery is dying

 

[inklen]

I just checked - battery is good, no heat anywhere...

 

[inklen]

I just realized I used 10V tantalum capacitors. Could this cause this fading issue?

 

[Audioguru]

The schematics are missing an important supply bypass capacitor.

I think all electret mics that size have a sensitivity almost the same.

The schematics do not show the polarity of the capacitors. They assume that you know the correct polarities.
Most of the capacitors should have the + wire at the IC except C3 which I think should have its + wire at the mic.

 

[inklen]

The schematics is on p.21.

 

[Audioguru]

I wish you posted the schematic. I copied and pasted it here:

 

[inklen]

oh, sorry, I misunderstood.

--- Updated Aug 31, 2020 ---

I think I'll make it again but with capacitors that have right voltage. Maybe that's the reason of that fading.
Building an experience here

 

[Audioguru]

With the 9V supply, most capacitors are fine with only 4.5V across them.
The polarity of the capacitors is extremely important. Do they have the correct polarity?

 

[inklen]

I will double check but I'm pretty sure they are. Otherwise it wouldn't work at all I guess.
I'm going to make this pcb again Make it bigger and simpler. I made it now double sided but very small with 0402's.

 

[barry]

Building a new Pcb without knowing why the original didn’t work is not a good idea. What makes you think the size of the PCB Matters?

 

[inklen]

What I'm thinking is maybe mic sensitivity plays some role. In the doc - 2.1 Gain Calculation - the sensitivity is used for R2 value calculation. In my case the sensitivity is -26dB (not -35dB). This will give an R2 = 27 kOhm. But I think even with 75k it should work, maybe having some distortion... not sure.

Also, the C2 is 15 pF in calculations but it's 10 pF in the schematics. Not much of a difference but C2 calculation uses R2 value. So C2 may be different for my mic.

No other ideas for now except to make this PCB again but "slightly more right"...

 

[barry]

The sensitivity has nothing to do with it.

It works, in your words, “very good”. And then it doesn’t. So some circuit parameter is changing over time or temperature. Your 10 V capacitors might actually want to be higher voltage, not lower. Maybe you can post a picture of your layout?

 

[inklen]

This is the best I can do I think.

Front:

Back:

I turned off filled areas.

 

[barry]

Is there a ground plane? And, just to reiterate what someone else said, verify the tantalum polarity. Tantalums really hate reverse voltage.

 

[inklen]

Yes, the ground plane on both sides. I will double check tantalums tomorrow.