Unity-Gain Bandwidth OpAmp vs group of Transistors to amplify small audio signal

[HMS]

Hi everyone,

I have a mic output signal that is too small and I can amplify it using transistors for a bandwidth of (20-20kHz). the question is:

Do i get the same result when using Unity-Gain Bandwidth Op-Amp? or should I combine both (Transistors + Op-Amp) for better results and lower cost?

I am confused, because i need to amplify mic signal without affecting phase (very important) and have all output frequencies amplified the same level.

Any advice is highly appreciated. thanks!

[CaptDon]

Well, first off, 'unity gain bandwidth' must be specified. U.G.B. is the frequency where there is no additional gain available. All opamps have a unity gain at some point. 1Mhz and 4Mhz was common for opamps used in audio work. Sometimes U.G.B. was specified in an open loop mode where the gain at low frequencies was 10K or even 100K being common. Then as the test frequency rose the gain falls off until it reaches unity. Almost any opamp of the common variety used in audio will have a usable gain of at least 20Db out to about 100KHz. What you want for your Mic is a 'low noise' opamp.

[srb1954]

I have a mic output signal that is too small and I can amplify it using transistors for a bandwidth of (20-20kHz). the question is:

Do i get the same result when using Unity-Gain Bandwidth Op-Amp? or should I combine both (Transistors + Op-Amp) for better results and lower cost?

I am confused, because i need to amplify mic signal without affecting phase (very important) and have all output frequencies amplified the same level.

Why is it so important that you don't affect the phase?

To minimise phase shift (you can't completely eliminate it) you will need the bandwidth of your circuit to greatly exceed the bandwidth of the incoming signal or apply complex phase correction networks to equalise the time delay variation through the amplifier.

Of course, the slightest shift of position of your microphone will undo all your hard work designing an amplifier with minimal phase shift.

[Konkedout]

I think that the advice offered by the two contributors before me looks good.  But I would like to add that you should stick with op amps which are recommended for use with audio.  I guess that the gain-bandwidth of an LM324 is too low anyway, but additionally..it and many other non-audio op amps produce significant crossover distortion.   Wikipedia has an explanation of crossover distortion which looks reasonable.

I will repeat the advice of the above, that unity gain bandwidth is the frequency at which the amplifier open loop gain falls to 1.  Generally you need to use an amplifier with unity gain bandwidth that is far above any frequency that you want to amplify with good fidelity.

[magic]

An operational amplifier (or integrated instrumentiation amp) is probably easier to apply for a beginner...

You need to have some idea about output impedance of the microhpone, its signal level, and desired amplified output level.
Otherwise, just try NE5534 in standard noninverting gain circuit, see what happens and continue from there...

[Kleinstein]

Using an OP-amp is easy and can give good linearity.
With discrete transistors one may get lower costs, if one accepts some distortion. The design is a bit more tricky, though not to bad with a check in the simulation.

A big question is also what supply is available.

[RandallMcRee]

No phase shift? That was not exactly what OP said:  without affecting phase.  In audio shorthand this usually means avoiding the many topologies that invert the output.

So don’t use opamp in inverting mode.

[TimFox]

Simplified approximation for phase shift of op-amp circuit:
1.  Assume that unity-gain frequency f_U is well above the operating frequency f, and that the op-amp is "unity-gain stable" (excess phase shift well below 180^o at f_U ).
2.  At the operating frequency, the open-loop phase shift of the amplifier will be approximately 90^o (without feedback).
3.  At the operating frequency f, the open-loop voltage gain will be approximately A_OL = f_U / f .
4.  If the design value for closed-loop gain (with negative feedback) is A_CL , then the "loop gain" A_L = A_OL / A_CL.
5.  The approximate phase shift at f will be (90^o) / A_L .

For a more accurate computation, SPICE (with a good model for the op amp) will give gain and phase for the complete circuit with feedback.

[CaptDon]

For the Original Poster, a couple of things. If the microphone is a low impedance variety such as 150 ohms or 500 or 600 then to maintain the published frequency response characteristic you will need to have a correct value load resistor across the Mic prior to amplification. The next problem comes from how your signal is getting to the mixing/recording equipment. Is it balanced or unbalanced input? In either case are you driving a long snake cable, typically 100 feet? All of this will determine your buffer stage after amplification. Also, does the mixer/recording equipment provide phantom power? Phantom power usually comes from +48 V.D.C. referenced to pin 1 (XLR) and is fed through a pair of 6.8K resistors one going to pin 2 and one going to pin 3 of the XLR connector. This is how FET preamplified condensor Mic's get their power. As for phase shift, a huge phase shift at low frequencies is not tolerable because during mixdown it could make instruments and voices with low frequency content nearly disappear from the mix, for instance a Mic 180 degrees out of phase in relation to other Mic's is a real problem! A tiny bit of phase shift at the higher frequencies will mimic the absolute Mic physical position relative to the source and will not be detectable except for a steady tone in a stereo mix so tiny amounts of phase shift really won't matter. Sounds like you are a beginner in the field of audio and are worried about all of the 'stories' you have heard about phase. You need to think about the questions above to determine your most optimal amplification specification and final design.

[BeBuLamar]

Oh if the OP meant flat frequecy response from 20-20kHz and not inverting the phase then an opamp chip designed for audio should be fine. It's not the cheapest but may cost $5-$15 per chip. Typically the mic would have low output impedance but need very high gain like 40 or so. So perhaps using inverting amp in 2 stages.

[TimFox]

Another type of IC suitable for microphone preamplifiers:
https://www.thatcorp.com/datashts/THAT_1510-1512_Datasheet.pdf

[shapirus]

Any advice is highly appreciated. thanks!

In addition to everything above, check this out: http://nwavguy.blogspot.com/2011/08/op-amp-measurements.html
It has some practical recommendations and comparisons of specific op amps for audio systems.

[magic]

Only practical if all you care about is THD+N at low gain and with source impedance equal to whatever it happens to be in the O2

Also, going by the datasheet LM4562 should have lower noise than NE5532 in the O2, so I suspect that a counterfeit chip had been tested.

[shapirus]

Only practical if all you care about is THD+N at low gain and with source impedance equal to whatever it happens to be in the O2

Also, going by the datasheet LM4562 should have lower noise than NE5532 in the O2, so I suspect that a counterfeit chip had been tested.

That's fine and expected. Everything except pure math always serves only as approximate guidelines that suggest a general idea and save your time by reducing the size of the data set that you need to process. Ultimately you have to verify and confirm things yourself.

[MathWizard]

There's the 8-pin LM386 audio amp IC, it must be one of the oldest and most common audio IC's  still in use.

For an "electret" mic,  here's a pic of a simple 3 BJT mic amp I made for a radio TX running off 5V, it's easy to change the center frequency and bandwidth by changing the LC-low pass filters. Mine is too close to 10KHz center freq. R63/64 is a 10k pot