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.

INA149: differential to single ended and offset

Part Number: INA149
Other Parts Discussed in Thread: AMC1350, , INA146, AMC3330, TLV6001, OPA391, OPA396

Dear experts,

I am working on an analog circuit to measure AC 220V 400Hz. I am using AMC1350 to convert +/-5V to +/-2V differential voltage signal. I need to convert  this differential signal to single ended and add an offset to achieve 0-3V. This 0-3V signal is connected to micro-controller ADC. I referred TI's document (LINK), in which it explains several configurations to achieve above requirement. Initially I wanted to use the 1st circuit shown in the document. However, I would like to use the 4th circuit for its simplicity and usage of minimum number of OPAMPs. The problem now is the IC used- INA146, which is not available and it is going to be discontinued. So I want to use INA149 that does the half of the job-converting differential to single ended.

Could you please suggest a circuit that converts differential signal to single ended, at the same time giving an offset to the signal. Kindly suggest superior alternative to INA149 if any, as I see INA149/146 can be operated at +/-18V. which is way greater than 3.3V-the voltage of my interest.

Regards,

Rajesh BN.

  • Hi Rajesh,

    The INA146 is still active and will not be discontinued any time soon. We are only discontinuing the orderables that comes in the small tube carrier.

    Notice that the large tape and reel carrier is still listed as active and will continue to be produced.

    As you pointed out, the INA146 can be operated up to +/18V but it can also be operated on supplies as low as +/-2.25V.

    See TINA simulation below for operation on a typical +/-5V supply.

    INA146_4Vdiff_3Vout.TSC

    Please let me know if you have any more questions.

    Regards,

    Zach

  • I see the negative of the differential signal is connected to the ground. is it mandatory?

  • Hi Rajesh,

    No this is not required, it is just there to give the simulation a common-mode input voltage.

    The typical output common-mode voltage of the AMC1350 is 1.44V, therefore it is more accurate to show the circuit as updated below.

    This should be more representative of your actual input signal to the INA146.

    INA146_4Vdiff_3Vout_1.44Vcm.TSC

    Regards,

  • Dear Olson,

    Thank you for your prompt solutions. I will consider the circuits u mentioned. But, I would like to use INA149 for this application, for the INA146 is not available in stock. Could you please suggest a circuit to shift the signal. In addition to it, I would like to use single voltage, i.e, 3.3V as shown in Figure-7 in the TI document

    (LINK). Using +/-5V complicates my PCB design. Please find the TINA file attached.AMC3330_INA149.TSC

    Regards,

    Rajesh BN.

  • Hi Olson,

    I have recently learned from a TI document available on the AMC3330 page (LINK) about a circuit that process a differnetial signal for the ADC input(0-3V). It uses TLV6001 circuit to interface with ADC. I have a query regarding this interface.

    Since, TLV6001 is CMOS based, and CMOS based OPAMP is recommended to be used before the ADC pin of the microcontroller, can I connect a RC filter at the output of TLV6001 before connecting to the micro-controller?

    Secondly, there are comparators that are used to derive over voltage alarms/trips. Can I use the output of the above mentioned OPAMP, or would you recommend to use another opamp for voltage follower?

    Please find the TINA simulation attached.

    diff_to_sing.TSC

    Regards,

    Rajesh BN.

  • Hi Rajesh,

    I cleaned up your AMC3330+INA149 simulation and was able to make it converge a lot faster. TINA really prefers to use the "ground" component as ground instead of a terminal whenever possible. The main issue with this circuit is that the INA149 does not have enough output swing when operating on a single 3.3V supply. See from the simulation below that the output is clamped between 2.3 and 0.992V.

    AMC3330_INA149_Vout_Limit.TSC

    Because of the output swing limitations, it is appropriate to use a rail-to-rail output amplifier such as the TLV6001. However, you must consider that even rail-to-rail amplifiers have output limitations. We call these "rail-to-rail" because they can swing very close to the rail (usually 100s of mV) but you can never swing all the way to ground on a single supply.

    Notice in the article you mentioned that the author did not drive the TLV6001 output all the way to ground, but instead allowed 500mV of headroom to either rail.

    Simulating your circuit again using the TLV6001 shows that the output is clamped at 36mV. Overloading the output into the negative rail is undesirable as this distorts your output signal and degrades the signal integrity.

    The TLV6001 datasheet shows that the maximum output swing limitation could be as much as 100mV from the rail. However, this describes the voltage in which the output has completely collapsed to the rail and performance is degraded. For best linearity, we must consider the open-loop gain (AOL) specification. The AOL specification tells us that we should allow 300mV of headroom to either rail for best performance.

    Below, I have modified the circuit again to allow the proper headroom at the TLV output which provides a linear output to the ADC.

    AMC3330_TLV6001.TSC

    can I connect a RC filter at the output of TLV6001 before connecting to the micro-controller?

    Yes, you can connect a proper anti-aliasing RC filter at the output. The exact RC values depend on the ADC being used and the sample rate. You may also need to include some feedback capacitors in parallel with the 2.71kΩ resistors to limit the bandwidth and ensure that the amplifier is stable.

    there are comparators that are used to derive over voltage alarms/trips. Can I use the output of the above mentioned OPAMP, or would you recommend to use another opamp for voltage follower?

    This really depends on the circuit being driven. If the comparator input is already high impedance, I believe you do not require an additional buffer stage. 

    Regards,

    Zach

  • Dear Zach,

    Would you recommend IC which are superior to TLV6001.

  • Hi Rajesh,

    I can certainly recommend other op amps to replace the TLV6001 in your circuit. Can you confirm what metric you would like to improve? For example: offset voltage, bandwidth, power, etc?

    What is the resolution of the ADC being driven? What is the sample rate?

    Regards,

    Zach

  • the ADC is 12bit, and the range is 0 to 3V. The sample rate is 15KHz. The ripple frequency of the signal is 5KHz. I want to use the range as close as possible. Let's say, a 300mv room to be allowed. Then I have to operate between 0.3 and 2.7. u can recommend opamp in other aspects of any 

  • Hi Rajesh,

    You could use the OPA391 or cost-down option OPA396. The AOL is specified for output swing 100mV from either rail for light loads.

    In this case, you can swing from 0.1V to 3V for your ADC input voltage.

    See updated circuit below.

    Regards,

    Zach

  • Hi Rajesh,

    See attached TINA simulation file for your reference.

    AMC3330_OPA391.TSC

    Regards,

    Zach