THAT1512 alternative for DIY Microphone project

Thread Starter

RagTMS

Joined Sep 2, 2022
9
Dear all,

This is a repost since the original post (https://forum.allaboutcircuits.com/...r-diy-microphone-project.188951/#post-1760251) was in the wrong section

This is the project I am trying to recreate


I'm looking for a way to make the preamp created in this video a better and more cost-efficient way. After doing some research (I am far from an expert in this field and so I may be completely wrong), the THAT1512 seems completely overkill for the diaphragm I am using and was looking for an alternative that preferably does not need the 5V to ±15V converter.

Furthermore, if you would like to help with any other part of the project, suggestions are welcome
 

ronsimpson

Joined Oct 7, 2019
2,385
If I made the microphone:
1-Because I have high end audio machines, I would run directly from mic to mixer with no electronics.
2-There is a simple one transistor amplifier that is common in powered mics. It can run from a battery or "phantom power". That will boost the signal a little.
3-There are a number of USB sound cards OR small boxes that have mic/line inputs and USB outputs.
The USB sound cards do not have the best mic inputs.
The ADC-USB boxes have good mic inputs.

Reinventing a USB-ADC is not easy. Adding a preamp inside the mic is easy.
1662136637126.png
1662136808995.png
 

Thread Starter

RagTMS

Joined Sep 2, 2022
9
Since I would like to keep costs low and I'm also primarily doing this project to increase my knowledge about electronics, I'm going to pass on buying an ADC USB. In the meantime, however, I have found the INA217. Would this be a better preamp than the THAT and would it require the 5V to +-15V converter?
 

Ian0

Joined Aug 7, 2020
6,286
As your preamp requires a single-ended input, I don't see the point of making a differential signal after the JFET. For the quality you are expecting, a JFET op-amp attached to the microphone capsule would do the job.
You could us the PCM2904 USB codec to get the USB output.
 

Thread Starter

RagTMS

Joined Sep 2, 2022
9
Would the TL072 work without outputting audible noise? If not what do you suggest as the JFEF op-amp. Also, what do the letters after TL072 (e.g TL072CPWR) mean and do they affect which op amp I should buy.
 

Ian0

Joined Aug 7, 2020
6,286
There are better JFET op-amps than the good old TL072, but it you have one it would be worth a try.
Bear in mind that it doesn't particularly like low supply voltages.
You might find some JFET op-amps that will work at 5V supplies, or you might want to try something like an MCP6021 which is rail-to-rail in and out, although its 1/f noise corner is rather high.

I know it's a bit vague, but I don't have the spec for your microphone capsule to know how much amplification it needs.
 

Thread Starter

RagTMS

Joined Sep 2, 2022
9

Ian0

Joined Aug 7, 2020
6,286
I don't have an op amp right now so I'm ready to buy one.

This is the microphone capsule https://micbooster.com/jli-microphones/262-jli2555.html
Cardioid Electret capsule without FET
Sensitivity -42±3dB (1KHz,0dB=1v/Pa)
Operating Voltage: 1.5V
Impedance: ≤2.2KΩ
Directivity: Unidirectional
Sensitivity reduction: ≤ 3dB Vs=1.5 to 1.0V
S/N ratio: ≥65dB
Maximum input sound level: 135 dB SPL


Would this 5V op amp be any good?
https://uk.rs-online.com/web/p/op-amps/3568212
That datasheet is for a capsule with a built-in preamp. There are no data for the capsule on its own.

[edit] The preamp is just a source follower so the capsule output voltage is the same as the preamp output voltage.
Try OPA1641. The input bias current 4.4nA of the TLV2462 is probably rather too high. The OPA1641 is 2.2pA.
You will need to know the capsule capacitance to be able to calculate circuit values.
 
Last edited:

Ian0

Joined Aug 7, 2020
6,286
It's a condensor microphone. Condensor is an old word for capacitor. If it didn't have capacitance, it wouldn't have any output.
From C=Q/V
IF the capacitance varies (because the distance between the plates varies due to the sound pressure waves), and the charge is fixed (because it is an electret) then the voltage varies with the sound pressure.
 

Thread Starter

RagTMS

Joined Sep 2, 2022
9
It's a condensor microphone. Condensor is an old word for capacitor. If it didn't have capacitance, it wouldn't have any output.
From C=Q/V
IF the capacitance varies (because the distance between the plates varies due to the sound pressure waves), and the charge is fixed (because it is an electret) then the voltage varies with the sound pressure.
Oh, I see. Thank you for the explanation. Should I buy the OPA1641 for this project then and would it result in a noise figure of under 1dB?
 

Ian0

Joined Aug 7, 2020
6,286
Oh, I see. Thank you for the explanation. Should I buy the OPA1641 for this project then and would it result in a noise figure of under 1dB?
I don't know - what is the self-noise of the microphone element? You can only determine noise figure if you know the SNR of the input signal.
 

Ian0

Joined Aug 7, 2020
6,286
Great! Thanks for all the help so far. One question though, TI suggests using this product https://www.ti.com/product/OPA1655 instead of the OPA1641. Do you think this is necessary?

Edit: Do you think this will make an audible difference for the better?
OPA1641 has lower bias current (2.2pA vs. 20pA) That might be more important than any distortion figures. The 1655 claims overall lower voltage noise but its 1/f noise corner is a decade higher frequency, meaning the 1641 probably has lower actual noise.
1641 has much lower current noise (0.8fA/√Hz compared to 6fA/√Hz) , which is multiplied by the output impedance of the microphone and adds to the total noise.
From the capsule capacitance, you can determine the load resistance from 1/(2πfC). The load resistance might be rather high. Some condensor microphone use 10GΩ
20pA bias current through a 10GΩ gives about 200mV offset, but that should still be OK.
 
Top