This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

Which INA needed? Up to 24 KHz, input: 2 x 1 kOhm impedance coils, output: Audio mic. or line signal

Other Parts Discussed in Thread: INA1651, INA1620, INA1650EVM, INA1620EVM, INA821, INA103, INA163, INA849, INA149, INA851, INA826, INA826EVM, INAEVM, INA848

Hi!


I need an instrumentation amplifier circuit (INA) for scientific precision measurements of weak EMF-signals.


My input will come from 2 coils that will pick up weak EMF signals in an environment with possibly a lot of noise. One of the coils will pick up the wanted signal + the noise and the other will only pick up the noise. That’s why I want to subtract signal no. 2 (the reference coil) from signal no. 1 (the signal coil).

My immediate plan is to use ordinary telephone pick-up coils. I have bought 2 Monacor coils that supposedly have an impedance of about 1 kOhm and a resistance about 280 Ohm.

The output of the INA will serve as input to a sound card (external, connected to a PC by USB). The sound card cuts off signals over 24 KHz.

The external sound card is also a mixer, with 2 amplifier steps with adjustable gain, but I need the INA to amplify my signals significantly, because of the noise in the sound card.

The possible inputs to the sound card are described as:
1 x Mic, mono: 0.5 mV (with an optional 20V phantom power)
1 x Line In, stereo: 2 mV, 150 mV, RCA L/R
1 x Tape In, stereo: 420 mV, RCA L/R

I also need to build a power supply for the INA. The general power lines here in Denmark are 230V, 50Hz.

The sound card/mixer microphone input has an optional 20V phantom power. Could that be useful?

So which INAs would serve my purpose?

And how should I build a power supply that doesn’t add noise or other distortion to my weak signals?

I look forward to hear from you. Thank you.


Best regards,

Henrik R.

  • Hi Henrik,

    So which INAs would serve my purpose?

    INA1620 is a differential instrumentation amplifier for audio application, which it meets the requirements. Also INA1651/2 may work for the application as well. 

    https://www.ti.com/lit/ds/symlink/ina1620.pdf?ts=1696253738945&ref_url=https%253A%252F%252Fwww.ti.com%252Fproduct%252FINA1620

    https://www.ti.com/lit/ds/symlink/ina1650.pdf?ts=1696260840378&ref_url=https%253A%252F%252Fwww.ti.com%252Fproduct%252FINA1650

    The sound card/mixer microphone input has an optional 20V phantom power. Could that be useful?
    And how should I build a power supply that doesn’t add noise or other distortion to my weak signals?

    The phantom power is able to source a few mA for the microphone application, and it may not have the power to drive the INA1620. Also, it will be good ideas to implement separate power sources to drive the microphone and preamplifier for the microphone. 

    With 230Vrms and 50Hz, you may need build a separate power supply for this. I would recommend to use switching power supply to step down the AC line voltage to DC voltage, and use LDO to further regulate the switching PS's output voltage and generate the required voltage for INA1620 application. 

    You mentioned sound card in the application, you may tap in PC's AC/DC power lines as the power sources, but you will need to further filter out the DC output and reduce the switching noises for the audio application by implement additional LDO and/or LC filters. In other words, it will be the best option to use linear power supply to drive the audio application. The front end of of the AC/DC switching power supply will provide you with the high power efficiency to step down the AC to working DC voltage range, and back-end of the working DC voltage may be further regulated with LDO and/or LC filters to attenuate the switching noises from the switching power supply.  

    FYI, below is the link of our Audio op amp selection.

    https://www.ti.com/amplifier-circuit/op-amps/audio/overview.html

    If you have other questions, please let us know. 

    Best,

    Raymond

  • Great. Thank you!

    But what's the difference between INA1620RTWT and INA1620RTWR ?

    it will be good ideas to implement separate power sources to drive the microphone and preamplifier for the microphone.

    I am not sure what you mean by 'microphone' or 'drive the microphone'?

    There will be no ordinary microphone(s). There will be 2 coils, where one is picking up the wanted signal + noise and the other is only picking up the noise. That's why I need to subtract one signal from the other to filter out the noise.

  • Hi Henrik, 

    The sound card/mixer microphone input has an optional 20V phantom power. Could that be useful?

    I missed phrase the input signal. When the phantom power is mentioned above, I was thinking about the microphone.

    You have to check the sound card's output current specification. Generally, the phantom power does not design for high current application. So it may not have the power to drive the instrumentation amplifier.  Also, you need to pay attention to the input common mode voltage, otherwise, the instrumentation amplifier may not be operate linearly. 

    Enclosed is the link for INA1620EVM and INA1650EVM. 

    https://www.ti.com/lit/ug/sbou205/sbou205.pdf?ts=1696283394091&ref_url=https%253A%252F%252Fwww.ti.com%252Ftool%252FINA1620EVM

    https://www.ti.com/lit/ug/sbou178/sbou178.pdf?ts=1696284148231&ref_url=https%253A%252F%252Fwww.ti.com%252Ftool%252FINA1650EVM

    The input Vcm voltage examples are shown in Figure 55 and Figure 60 in INA1620's datasheet,  but it does not explain in details. The Vcm input voltage requirements are presented better in Figure 33 to Figure 38 in the INA165X's datasheet.  

    But what's the difference between INA1620RTWT and INA1620RTWR ?

    The ICs are identical, but INA1620RTWT comes with 250 units/package, while INA1620RTWR comes with 3000 units/package. 

    If you have other questions, please let me know. 

    Best,

    Raymond

  • I guess it would be much easier to simply use a number of batteries as power supply?

  • Hi Henrik, 

    I guess it would be much easier to simply use a number of batteries as power supply?

    Yes, it would be easier to place series of Li+ batteries or battery pack together (Use 18650,  26650 type of batteries or similar), if you just want to check out the performance. You probably can get the EVM module and modify the circuit per your input configuration + filters (differential and common mode LPFs, if any). 

    Please let us know if you have other questions. 

    Best,

    Raymond

  • Now I have looked at the documentation for INA1620 and INA1651 (and INA1620EVM and INA1650EVM).

    None of them really seem optimal. INA1620 only has 1 op amp in each channel instead of the 3 op amps of a 'real' instrumentation amplifier. And INA1651 (having 3 op amps per channel) has a fixed unity gain, which very much disqualifies it for me.

    It seems to me that the ideal amplifier circuit for me would have 3 op amps (like a 'real' instrumentation amplifier) and the possibility of a large gain/amplification.

    I was imagining something like what is described here: https://www.electronics-tutorials.ws/opamp/opamp_5.html - in the section under the headline "Instrumentation Amplifier".

    Doesn't TI have something like that, for frequencies at least in the audio range - possibly also higher?

    Maybe it could be achieved using INA1651 and INA1620 in series? But that seems like clumsy overkill to me...

    I look forward to hear from you. Thank you.

  • HI Henrik,

    Raymond is currently  OOO, but we will look into your query soon; please allow a business day.

    Thank you and Regards,

    Luis

  • Hi Henrik,

    Doesn't TI have something like that, for frequencies at least in the audio range - possibly also higher?

    Below is the selection of instrumentation amplifier. The list consists 2 and 3 -op amp instrumentation amplifiers(IAs). The latest IAs seem to be designed in 3-op amp types, but these are not optimized for audio grade applications, see the link below.  INA849, INA163, INA103, INA821 and many other IAs should work to cancel the noises from the voice coils' front end. 

    https://www.ti.com/amplifier-circuit/instrumentation/products.html

    Alternatively, you may use difference amplifiers or built your own difference amplifier from the list below. 

    https://www.ti.com/audio-ic/amplifiers/overview.html

    If you are able to provide us the input signal swing from the voice coils, gain requirements and available supply rails, I can simulate an input condition per the application (it looks like that you'd like to have variable gains of the IAs). Since you mentioned that the design will be used for audio application, I am suggesting the audio related products. Is 1kohm impedance in voice coil characterized at 1kHz or other frequency?  

    If you have other questions, please let us know.

    Best,

    Raymond

  • Why do you mention INA149, which has a fixed unity gain. I need a high gain.

    When I try to compare INA1620 with some of the others, like INA821, many of the parameters are missing for the INA1620. Is that an error? See here:

    www.ti.com/.../

  • Hi Henrik, 

    It was a typo, I meant to write as in INA849, which is listed from the parametric search. 

    Is that an error?

    INA1620 and INA1651 are optimized for CMRR and audio applications (tailored for human factor consideration in the audio application), and INA821 and INA849 are optimized for industrial application, where the input differential signals are very weak within high noisy environment. Surely, these have similarities in the configuration. 

    The possible inputs to the sound card are described as:
    1 x Mic, mono: 0.5 mV (with an optional 20V phantom power)
    1 x Line In, stereo: 2 mV, 150 mV, RCA L/R
    1 x Tape In, stereo: 420 mV, RCA L/R

    I need your design requirements or the differential input signal amplitude ranges in order to simulate the analog front end, say with INA849 or INA821. Since the above output settings are wide, it is possible that the application may require both attenuation and amplification configuration to accommodate the above requirements. My speculation is that the input signals for Mic application may need to be attenuated, while the input signals to the stereo and tape deck ports may need to be amplified. Therefore, I need to have the input signal ranges of the telephone pick-up coils in order to understand what are needed for the application. I am not certain that if one IA circuit will meet all the above requirements.

    If you have other questions, please let us know. 

    Best,

    Raymond 

  • Thank you!

    I don't actually know how weak my signals will be, but a high CMRR and a high gain will be of the greatest importance.

    I have not yet had a chance to measure the impedance of the coils. :-(

    I am not sure why you think I would need attenuation? For me the CMRR is the "attenuation". But I will have to experiment to determine the optimal gain, that gives the sound card a sufficient signal without overloading it.

  • Hi Henrik, 

    Please characterize the telephone pick-up coils for the application. I do not know how much current that the magnetic coil may be able to generate from a sound pressure. It depends on how the application is going to be configured it, including the distance of the voice coil from sound sources and amplitudes.    

    If the differential input audio signals are weak, INA849, INA821 or other IA can amplify it without issues.  

    If you have other questions, please let us know. 

    Best,

    Raymond

  • I have bought 2 Monacor telephone and guitar pick-up coils for the purpose:

    https://www.monacor.com/products/components/cables-and-plug-in-connectors/adapters/ac-71-3-5mm/

    "Impedance: approx. 1 kΩ"

    I have measured the resistance to 280 ohm.

    I have seen top scientists use a similar telephone and guitar pick-up coil for the same purpose. As far as I could see, their measuring system was more primitive, because they didn't have the extra coil as 'noise reference'.

    Thanks for pointing out INA849 and INA821. They, and INA851, are the ones I am considering at the moment.

  • I still haven't heard from Monacor and I haven't had a chance to measure the impedance of the coils myself, but based on comparing all the parameters of INA821, INA849 and INA851 and a few others, INA849 seems like the obvious choice. BUT there doesn't seem to be an evaluation board, where it has already been soldered on?

    I am not sure how to solder such microscopic pins?

  • Hi Henrik,

    BUT there doesn't seem to be an evaluation board,

    We have INA826EVM, see the link below. The INA826 footprints in SOIC-8 is p2p compatible with INA849, where you may use the EVM to evaluate the performance of INA849. 

    https://www.ti.com/lit/ug/sbou115c/sbou115c.pdf?ts=1698242584288&ref_url=https%253A%252F%252Fwww.ti.com%252Ftool%252FINA826EVM

    I am not sure how to solder such microscopic pins?

    There are many soldering/desoldering tips over the youtube.com and you should find one per the package. Really, you just need  a heat gun to remove and solder on the package. SOIC-8 package is fairly easy to work with. A few handon practices,, you should be good at the soldering/desoldering skills.

    If you have other questions, please let us know. 

    Best,

    Raymond 

  • OK... That's a bit sad/tiresome... :-( It adds extra work tasks and uncertainties to the project. In other words: Delays and possibly extra costs.

    But what's the difference between the https://www.ti.com/tool/INAEVM-ALT-SO8 and the https://www.ti.com/tool/INA826EVM ?

    (The first one is mentioned on https://www.ti.com/product/INA849 .)

  • Hi Henrik, 

    But what's the difference between the https://www.ti.com/tool/INAEVM-ALT-SO8 and the https://www.ti.com/tool/INA826EVM ?

    The https://www.ti.com/tool/INAEVM-ALT-SO8 is the alternative solution. Please review both schematic of the EVMs. 

    Either one you have to make mods and change the gains for your application. The benefit of INA826EVM is that the kit will work as it is received. You are able to play around with your input signals and change the gains and/or filters etc.. It will be easier to troubleshoot with the User's guide, if you are new to IAs configuration. Once you are comfortable with your settings, you can swap for INA849 parts. INA826 is not what you want, but it will provide you with the general functionalities per the design requirements.  

    Alternative solution will require more work (because it is an universal INAEVM), unless you are familiar with how IA works. Of course, we can send you the simulation and assist you to achieve the design objectives. 

    If you have other questions, please let us know. 

    Best,

    Raymond

  • OK. So the INA849 comes in 2 versions: SOIC-8 and VSSOP-8. Are there any evaluation boards for the VSSOP-8 version? I guess not. :-)

    I just discovered a variant with a fixed gain of 2000: INA848. It has CMRR = 132, whereas INA849 has CMRR = 120. But INA848 has a slightly higher noise at 1 KHz. That might be a good alternative.

  • Hi Henrik,

    Are there any evaluation boards for the VSSOP-8 version?
    I am not sure how to solder such microscopic pins?

    I do not recall. SOIC-8 is the most common IA package, and it is commonly used in EVM purposes. Also, it is easier to solder/desolder vs. VSSOP-8 package. 

    I just discovered a variant with a fixed gain of 2000: INA848.

    INA849 and INA848 are identical IC design, except INA848 has the fixed gain with the internal matching gain resistors; INA849 comes with variable gains for IA application. And I do not believe that you will need 2000 fixed gains for the application. 

    Best,

    Raymond

  • OK. I will buy the INA826EVM board and the INA849 chip.

    Monacor has answered:

    The impedance of appr. 1 K ohm was measured at 1 KHz.
    
  • Hi Henrik, 

    The voice coil's inductance is approx. 159mH. Do you know what are the low and high differential signals that able to induced from the coil per your configuration?

    BTW, the PC's audio card has the cutoff frequency up to 24kHz. The voice coil's operating frequency ranges are much lower, may be up to 4 - 5kHz range. 

    Best,

    Raymond