TLV8811: The output of the non inverting amplifier does not come out.

Hi JiHoon,

for a bipolar input voltage with the input voltage partially going negative you would need a bipolar supply voltage as well. Or at least a small negative supply voltage at the ground pin coming from a LM7705 or similar:

jihoon_tlv8811.TSC

What supply voltage do you use?

Kai

Hi JiHoon,

Please follow Kai's suggestion. Below is the input common mode voltage range for TLV8811 per a single supply voltage rail from 1.8Vdc to 5Vdc. The datasheet is shown Vs of 3.3Vdc as an example, which min. of Vcm has to be above 0V or GND.   

If you have additional questions, please let us know. 

Best,

Raymond

R1 and R2 are replaced. R3 and R4 also changed. R5 and R6 also changed.

The dc offset of the input voltage is 80mV, and the waveform comes out from +100mV to +60mV.

There is no negative voltage coming out of the input voltage.

The supply voltage is 3.3V to V+ and GND to V-.

When I test by adding 1uF as in the circuit diagram above, the waveform appears at first, but as time passes, the waveform does not come out.

How do I get a continuous output waveform?

Hi Jihoon,

yes, the circuit works until the input bias current has fully charged up C1. After that the input goes into saturation.

A path for the input bias current is missing behind C1. And you should eventually create a pseudo ground to allow negative going input voltages, for instance like that:

jihoon_tlv8811_1.TSC

Kai

As a result of testing as per the circuit diagram you posted, the output waveform is displayed, but it is not amplified.

Shouldn't R1 and R2 be swapped?

And V+ is 3.3V and V- is GND.

Also, Input Voltage is a waveform of 60mV ~ 100mV.

I cannot use 5V.

I need to use 3.3V.

As a result of testing according to the circuit diagram you posted, the output value comes out like the graph you posted.

But I need to amplify, how do I do that?

If you look at the circuit, you used 1.65V, but I can only use 3.3V.

Also, the input frequency range is about 100Hz ~ 300Hz, can you make a circuit diagram for this?

It's okay if the gain drops a little.

Hi Jihoon,

the TLV8811 has a too low gain-bandwidth to allow the gain to be implemented by only one gain stage. You will need two stages and distribute the gain evenly on both stages. Like this for instance:

jihoon_tlv8811_3.TSC

Kai

Hi JiHoon,

Thanks Kai for the simulation and circuit improvement. As Kai pointed out, the unity gain BW of TLV8811 is too low at 6kHz. If the input range is from 100-300Hz, the circuit's gain has to be << 20 V/V (6kHz/300Hz at -3dB point).  If you are going to use a single stage amplification, you may have to pick a higher BW op amp, say OPA391 or OPA396 or similar. 

The reasons are: your requirements needs to have gains at approx. 75-100 V/V. At 300Hz input signal, the cutoff frequency prefers to be >3kHz at minimum (without phase shift). the op amp for the application should have BW > 300kHz (3kHz*100) minimum (Use Gain Bandwidth Product equation to predict the requirements).  

With the two staged amplifications, the total gain is 8.5*8.5 = 72.25 V/V (or 37.18dB as shown). If you are not concerned about the phase shift at the output, this may be adequate.

If you have additional questions, please let us know. 

Best,

Raymond 

There is not enough space for the part to fit.

So I have to use only one opamp to amplify it.

Can't it be amplified using only one opamp?

The picture above is the circuit I tested.

The Vout waveform is 130mV to 230mV.

The dc offset of Vout is 180mV.

VG1 is also 60mV to 100mV.

The dc offset of VG1 is 80mV.

The above waveform value is the result of measurement with an oscilloscope.

Calculating the actual gain is 50mV/20mV=2.5.

Theoretically, the gain should be 8, but only 2.5.

How do we actually get a gain of 8?

Hi Jihoon,

what is the input signal during the test (waveform, frequency)?

Also, do not directly touch the output of TLV8811 with the scope probe. The capacitive load of scope probe can result in instability. Add an isolation resistor according to section 7.4.4 of datasheet.

Kai

Hi Jihoon,

The circuit you tested should have gain of 8. Please insert 100Ω resistor at the output of the op amp and the input frequency is 100Hz, and you should get the following result. 

Without AC input signal, TLV8811's output should be measured at approx. 1.65Vdc.

With AC input of ±20mV (block the DC), the TLV8811's output should be measured ± 8*20mV =  ± 160mV, which resulted 1.81Vpk (1.65+0.160V) and 1.49V_valley (1.65V-0.160V). 

If you have additional question, please let us know. 

Best,

Raymond

The figure below is the input signal waveform.

Time DIV : 0.1ms

Volts DIV : 5mV

Pictured below is the Roll Off Frequency found in the datasheet.

The chip from which the input signal comes out is the microphone sensor.

The part name of the microphone sensor is PMM-3738-VM1000-R.

We don't have space for parts, so can't we get an amplified signal with only the parts in the circuit diagram below?

I didn't have 100Ohm, so I tested it with 10kOhm.

Input Signal is 60mV ~ 100mV and Output Signal is 100mV ~ 240mV.

The gain is 70mV/20mV=3.5.

These are the results measured with an oscilloscope.

Hi Jihoon,

as Raymond mentioned

Raymond Zhang1 said:

Without AC input signal, TLV8811's output should be measured at approx. 1.65Vdc.

you should have 1.65V at the output of OPAmp when having no input signal (input terminal at the left side of 1µ cap grounded). So there's something wrong.

Another issue I see is, that the input signal looks like having high reaching harmonics. This is not a true 100Hz signal. When we talk about a 100Hz signal, we usually mean a sine of 100Hz. A 100Hz sine does not have harmonics.

So whatever you try to do, it will not work with a nanopower OPAmp. The bandwidth of a nanopower OPAmp is just way too low. Consider taking a micropower OPAmp or even a standard OPAmp showing a supply current in the 100µA...1mA range. This would be able to provide the amplification with a single gain stage.

Is the supply current critical in your application? Or by other words, what supply current can you allow?

Kai

Hi Jihoon,

As Kai suggested, you may consider to get a functional generator and configure a sinewave from 60mV to 100mV at 100Hz at the output of the generator and check your gain settings. If you use a functional generator with input specified from the above, you should get the exact output pattern from the simulation (your time axis will be different, but the signal period should be identical). 

Another thing is not clear to me. PMM-3738-VM1000-R  is a microphone transducer with input BW response from 100Hz to 10kHz. The designed op amp circuit has BW only up to 300 Hz. I am not sure what you are trying to do. How do you control audio frequency say at 100Hz or  between 100Hz to 300Hz? 

Please let us know what you are trying to do. If the microphone has to operate between 100Hz to 10kHz or higher, TLV8811 op amp is a wrong part for the application. 

Best,

Raymond

As shown in the circuit diagram below, the input signal was removed and the output waveform was measured.

Time DIV : 0.1ms

Volts DIV : 50mV

The output waveform is 180mV.

I don't know how to check the supply current.

Could you please tell me how to check it?

Can you recommend an opamp that can amplify the input frequency from 100Hz to 10kHz?

I am aiming to amplify the microphone output signal with an opamp and put it into the ADC input terminal of the MCU.

The MCU part name is MK64FN1M0VDC12.

Hi Jihoon,

I think your TLV8811 is damaged. Take a fresh one.

Again, what supply current of OPAmp can you allow? Why have you choosen a nanopower OPAmp?

Kai

Hi Jihoon,

Can you recommend an opamp that can amplify the input frequency from 100Hz to 10kHz?

Could you tell us what you are trying to do with the audio application? I think that you only need a single channel for the audio application. 

I did some search for low voltage and cost audio amplifier. The following list is suggested parts. For instance, OPA376, TLV376, OPA1671 may work for you, which both have SOT-23 (5pin) package and can operate up to 3.3Vdc. 

There may be other op amp options if we have clear requirements. My understanding is that your audio frequency range is from 100Hz-300Hz, but the microphone is able to pick up any audio information up to 10kHz or higher. With the current design, the op amp is unable to filter out any audio frequency  beyond 300Hz. Say that the current TLV8811 has the cutoff frequency at 300Hz. At 3kHz, the op amp is only able to attenuate approx. -20dB. 20mV input signal at 3kHz is only able to attenuate to approx. 0.1*20mV = 2mV. If the microphone picked up 200mV's noise at 3kHz, the op amp's output still have 20mV of signal at 3kHz, which it may be what is going on. 

I will follow Kai's recommendation and replace TLV8811 and see if the output signal is improved. You need to check the op amp's performance with a functional generator from 100-300Hz before the microphone is connected. With a functional generator, you should get a perfect input_sinewave*Gain in amplitude, say at 100Hz. If you don't, replace TVL8811 with a new one (assumed other electrical connections are proper and working).   

https://www.ti.com/amplifier-circuit/op-amps/audio/products.html?pqs=paqs&familyid=585#p480=1;1&p1261min=1.7;3.3&sort=p1261min;asc

If you have additional questions, please let us know.

Best,

Raymond

What I am trying to do with the audio application is a system that automatically calls 911 by receiving the words 'Help' as input into the microphone.

I want to buy the OPA1671.

Can the OPA1671 accept input signals from 100Hz to 10kHz?

Is there any problem with using the OPA1671?

If there is no problem in using the OPA1671, can I get a circuit diagram with 3.3V as the power supply?

The output signal of the opamp should be -0.3V to 3.6V.

The circuit diagram should not contain as many parts as possible.

Also, it was confirmed that 1.65V came out when no input signal was put into the TLV8811.

And the microphone input signal has a dc offset of 0.85V and a waveform range of 0.7V~1V.

Hi Jihoon,

there are zillions of OPAmps and circuits which can be used here. But why have you chosen a nanopower OPAmp? Because supply current is an issue in your application? Are there any restrictions on the supply current of microphone amplifier?

Also, where is the supply voltage of microphone amplifier coming from?

Kai

Hi Jihoon,

Can the OPA1671 accept input signals from 100Hz to 10kHz? Is there any problem with using the OPA1671?

Currently, SOT-23 (5) package is not in stock in TI store, unless you can use SC70 (5) package.  

Please reply to Kai's comments. Also, what is the BW of the microphone for the application? 100Hz - 10kHz or you'd like the op amp to  operate in the audio range 100Hz - 20kHz.  Please see OPA1671 's datasheet. 

Best,

Raymond

The reason for using the nanopower opamp was that it was chosen based on the supply voltage only.

I'm not sure how to check the mic amplifier's supply current limit.

You can refer to the circuit diagram below for the supply voltage of the microphone amplifier.

The supply voltage in the circuit diagram is 5V, is it ok to connect by changing the 5V to 3.3V?

Hi Jihoon,

I would do this way:

jihoon_tlv8811_4.TSC

VG1, V1 and R1 model the microphone. V2 is the common 3.3V supply voltage which the microcontroller, the microphone and the OPA1671 share together. This supply voltage is produced by the "SC195" which is a HF DC/DC "switcher" producing lots of HF noise and ringing. This and the fact that the microcontroller will add lots of noise on the common supply voltage line, makes it necessary to add rigorous supply voltage filtering at the microphone (R5, C4), the OPA1671 (R6, C3) and the pseudo ground "generator" (R4, C2). You can remove R4, R5 and R6, if and only if the 3.3V supply voltage is stable, clean and noise-free. But if it is not, all this noise will be amplified in the OPA1671 stage and will totally ruin the performance of your microphone amplifier. So please be warned!

The gain is chosen in order to not overload the OPA1671 with an input signal of 20mVp. R8 isolates the output from capacitive loads and is necessary to ensure stability.

R9 and C5 give a corner frequency of high pass filtering of 26Hz. R7 and C6 give a low pass filter corner frequency of about 10kHz.

In emergency call applications like yours it can be wise to further narrow the frequency response. "Old fashioned" analog telephones had a frequency response of 300Hz and 3.4kHz. The engineers who were responsible for this decision were very clever, because by limiting the frequency response in this way, breath noise and pop noise could not easily overload the microphone and microphone amplifier. And frequencies above 4kHz do not carry important voice information anyway but only unwanted noise. So you should think about this narrowing of frequency response.

Some words on the noise contribution of OPA1671 amplifier: The microphone's noise is specified to 62dB(A) in the 20Hz to 20kHz band at 94dB sound pressure. The sensitivity of microphone at 94dB sound pressure is -38dBV. So the "A"-weighted noise level is -100dBV or 10µVrms. This gives a noise density of more than roughly 70nV/SQRT(Hz). When looking into the datasheet of OPA1671 (7nV/SQRT(Hz) at 1kHz) this is very much more noise than the OPA1671 is producing. And the noise density of R9, the remaining main noise source, is only 10nV/SQRT(Hz). So, almost all of the noise at the output of OPA1671 is coming from the microphone.

Kai

I misspoke input signal.

The input signal has a dc offset of 0.85V and a waveform range of 0.7V~1V.

The output signal should be -0.3V to 3.6V.

I misspoke input signal.

The input signal has a dc offset of 0.85V and a waveform range of 0.7V~1V.

The output signal should be -0.3V to 3.6V.

Only 3.3V voltage can be used.

So, I cann't use the V1 800mV.

There is not much space for parts, so parts should be kept to a minimum.

Hi Jihoon,

V1, VG1 and R1 are parts of the microphone and the value of V1 doesn't play any role in the simulation because it is blocked by C1. And if the gain is too high or too low for your taste, then just change it to your needs by increasing and/or decreasing R7 and/or R9. You said at many placess that the signal coming from the microphone is +/-20mVp. Why is it different now?

By the way, why not doing these simulations by yourself? TINA-TI is free and can be downloaded here:

https://www.ti.com/tool/TINA-TI

As I already said, if you want to omit the filtering, you can do that, but you have to provide a 3.3V supply voltage which is clean, stable and noise-free. Otherwise you will run into disaster.

Why is the place limited for the microphone amplifier? Is it not sitting on the same printed circuit board as the microcontroller?

Kai

Hi JiHoon,

There is not much space for parts, so parts should be kept to a minimum.

I think that this may fit into the limited space.  The microphone's input is placed at the inverter input and this will improve THD performance of the OPA1671. In addition, the circuit is eliminated one coupling capacitor at the non-inverting input (not a significant space saving).   

2768.jihoon_tlv8811 microphone 02072022.TSC

I reduced the gain from 70 V/V range to 10 V/V, which is near the max. amplification that the circuit is able to handle with 3.3V supply rail. Since you may not have room to filter out higher frequency noises at the supply rail to OPA1671, you may consider to check the 3.3V ripple voltage at the output of SC195 switching regulator. Use low ESR ceramic capacitors to lower the switching ripples at the output of  the regulator.

If you have additional questions, please let us know. 

Best,

Raymond 

The circuit diagram you posted is an inverting amplifier.

Please design it as a non-inverting amplifier, not an inverting amplifier.

I need a non-inverting amplifier.

The reason the mic signal isn't +/- 20mV is that I measured the oscilloscope wrong.

The microphone sensor is on the same board as the microcontroller.

There is not much free space because there is a situation where a microphone and an amplifier must be added to the previously designed circuit board.

Then give this a try:

This is the absolute minimum you need.

Kai

Why are there two inputs VG1 and PMM-3738?

Please input only one.

The gain of 20 seems too high. I think it's appropriate to set it to 10.

Hi Jihoon,

JiHoon Hwang said:

Why are there two inputs VG1 and PMM-3738?

Please input only one.

So we don't want to power the microphone??

JiHoon Hwang said:

The gain of 20 seems too high. I think it's appropriate to set it to 10.

This is no gain of 20 but 20dB, which equals a gain of factor 10.

Kai

Does OPA1671's evaluation board exist?

I've been looking for it, but it doesn't seem to exist.

I tried to buy 250pF, but the market doesn't have the size I want for 250pF.

So, I want to buy 240pF.

Is it ok to use 240pF?

Yes.

If the dc offset is 800mV and I calculate +/-150mV at 800mV, isn't it 650mV to 950mV?
By the way, Vsig is 700mV to 1000mV. Something is wrong.

In fact, when I measured it with an oscilloscope, the dc offset was 850 mV and the waveform ranged from 700 mV to 1000 mV.

I'm trying to run a simulation with the Tina program.
How do I add OPA1671?

Hi Jihoon,

the DC level coming out of the microphone alters with the supply voltage of microphone. So it can be 800mV or 850mV or anything else. But as this DC voltage is blocked by the high pass filter cap C1, it doesn't matter at all.

Kai

Hi JiHoon,

How do I add OPA1671?

Please click on the link below and download the OPA1671 Tina-TI reference design. 

https://www.ti.com/product/OPA1671?utm_source=google&utm_medium=cpc&utm_campaign=asc-amps-null-prodfolderdynamic-cpc-pf-google-wwe&utm_content=prodfolddynamic&ds_k=DYNAMIC+SEARCH+ADS&DCM=yes&gclid=CjwKCAiAgbiQBhAHEiwAuQ6BkjmbZ18E8O4VreORvYG17XW-Qerk7D2uVlSmqk3-qFidUvZA2xlTXxoCdgQQAvD_BwE&gclsrc=aw.ds#design-tools-simulation

As Kai indicated, microphone's DC bias voltage is blocked by capacitor; only AC or audio signals pass through the capacitor C1. 

Best,

Raymond

I drew all the circuit diagrams.
By the way, how can I check the Vout graph and the Vsig graph?

Hi Jihoon,

have you watched the TINA-TI training videos?

https://training.ti.com/introduction-tina-titm-simulator?context=1127691-951878

Kai