• State
  • Locked Locked
  • Replies 5 replies
  • Subscribers 48 subscribers
  • Views 1537 views
  • Users 0 members are here
Support feedback
Options
Options
Similar topics

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.

THS4130: Spurious frequencies(even after powering with battery) after diffrential amplifer

Prakhar Jain
Prodigy 90 points
Part Number: THS4130
Other Parts Discussed in Thread: THS4303

Dear Sir,

Application : We have a sensor for counting micron size particle which comprises of two pairs of electrodes. The 2.5V sin wave modulating signal is supplied to both pair of electrode and whenever the particle is present between any two pairs of electrode there is change in the current

Methodology :. Currently we are using THS4303 for converting the current coming from the electrode into the voltage. The difference of voltage between the electrodes is sensed using the THS4130. However, the current circuit is giving some unwanted frequency when the signal is demodulated. The peaks due to the particles are quite clear but with the unknown frequency. I am attaching schematic , block diagram and out put from the board.

Isssue : As you can see there are spurious frequencies. In the video you will see that, there is some constant frequency on which on our signal( the peaks you see are the events) is riding

  Please let us know how to troubleshootTI query.rar

  • 0
    TI__Genius 15975 points

    Hi Prakhar,

    It looks like the second THS4130 is driving a capacitive load directly, which is usually responsible for reducing the amplifier phase margin and cause sustained oscillations. I would recommend isolating the second THS4130 output and the filter that it is driving with 50-ohms- R17, as shown below. Also make the R564 as 0-ohms. I think this should take care of any constant frequency seen on your capture.

    Best Regards,

    Rohit

     

  • 0 in reply to Rohit Bhat
    Prodigy 90 points

    Dear Rohit,

    Thanks for the reply

    We changed the resistor values as per your suggestion but still we are getting the noisy carrier signal, as you can see in attached video. We were expecting a flat baseline but there is no change in o/p wave form. I am even sending you the gerber file of the board.

    Our operating frequencies are upto 2Mz only . From the previous email you can see that we excite our sensor(dual channel) with frequencies of 2Mz ; the corresponding o/p in form of uA from respective channels is done trans-impedance using THS4303 and then differential amp(dual) is used leveraging THS4130. Finally the o/p is passed to FPGA based lock-in amplifier to demodulate.

    Kindly suggest!

    Thanks!

    SENSOR_CARD_GERBER.pdf

  • 0 in reply to Prakhar Jain
    TI__Genius 15975 points

    Hi Prakhar,

    Have you tried to probe the supply lines with an oscilloscope to see whether the same waveform is visible at the board output? It is possible that the 1.1MHz switching frequency of +5V DC/DC step down converter could be coupling into the signal path. If it is not the step down converter, I would recommend to probe individual stages in the signal path with an oscilloscope; starting first with the THS4303 output, followed by THS4130 in-order to isolate the spur on the board. Also, make sure the probe inputs are set to 1M-ohm. It might also help to look the board output on a spectrum analyzer to see the spur frequency.

    Is there a reason you are using only single-ended output of the THS4130 instead of the differential output? I think you can cancel this common mode noise riding on your required signal if captured differentially.

    I looked over the schematic and Gerber files, and the circuit looks okay. However, one thing that we normally do is sandwich the Power layer in-between two GND layers, which provides good supply bypassing. As a result, I would recommend to include an additional GND layer in-between the Power layer and the Bottom layer for the next board revision.

    Best Regards,

    Rohit

  • 0 in reply to Rohit Bhat
    Prodigy 90 points
    Dear Rohit,

    1. We are running with Li-Ion battery which is clean . We have checked power supply and its regulated
    2. Single ended o/p : We are taking o/p w.r.t ground
    3. We did your 50ohm one, the magnitude of spurious frequencies is lower than before but still a considerable magnitude is there

    Do you think
    1. We placed fer-rite beads at quite distance from Op-Apms and also instead of Tantalum de-coupling capacitors we are using ceramic . Do you think it might give rise to any issue?
    2. Our excitation frequency(carrier wave to sensor; 1.7Mz ) has -80dB SFDR. Do you think the harmonics of it is generating spurious frequencies?
    3. Any cross talk in layout is taking place ?

    Thanks!
  • 0 in reply to Prakhar Jain
    TI__Genius 15975 points

    Hi Prakhar,

    To answer your questions:

    1. We placed fer-rite beads at quite distance from Op-Apms and also instead of Tantalum de-coupling capacitors we are using ceramic . Do you think it might give rise to any issue

    Ans: It is a possibility that not using Tantalum capacitors on the supply lines and placing ferrite beads at quite distance from Op-Amps could be causing power supply ripple. But, based on what you are describing that the power supply is regulated and you don't see any ripple on the power supply, I don't think it could be a potential major source for the noise coupling.

    2. Our excitation frequency(carrier wave to sensor; 1.7Mz ) has -80dB SFDR. Do you think the harmonics of it is generating spurious frequencies?

    Ans: I am sort of having a hard time understanding your waveform. Do you know the time duration of the spurious frequency and also is the vertical axis 1mVpp? You mentioned that placing the 50-ohm close to the THS4130 output reduced the magnitude of the spur. Have you looked to see whether bypassing the output filter flattens out the output waveform? It is possible the THS4130 output needs higher isolation value resistor than the 50-ohms while driving the filter cap. 

    3. Any cross talk in layout is taking place ?

    Ans: If you are showing waveforms in the mV range, then crosstalk could certainly be a factor for single-ended signals.

    Have you probed individual stages in the signal path with an oscilloscope to isolate from where this spur could be coming from? Also, are you making sure the GND of the measurement unit and your board are similar? Sometimes you see a hum noise coupling on the single ended signals if the GNDs are not the same.

    Best Regards,

    Rohit