• State TI Thinks Resolved
  • Locked Locked
  • Replies 6 replies
  • Answers 2 answers
  • Subscribers 48 subscribers
  • Views 1186 views
  • Users 0 members are here
Support feedback
Options
Options
Related

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.

TLV2333: Output waveform distortion

Ruixin Feng
Intellectual 2535 points

Part Number: TLV2333
Other Parts Discussed in Thread: LMV358, OPA2365, , OPA2333

Hi team,

customer use tlv2333 in below circuit. when using tlv2333, output distorted as third picture. when using opa2365/lmv358, output is not distorted as second picture. 

The input signal is 1kHz.

Power supply is 3.3V single supply.

I asked customer to lower the input frequency in case the distortion is caused by lower bandwidth, when input frequency down to 50Hz, output is not distorted. 

But for 1kHz, 6.8 gain, 350kHz bandwidth seems ok. Could you pls advise on it? Thanks!

TLV2333 circuit

yellow input and blue output using opa2365

blue input and yellow output using tlv2333

  • 0
    TI__Guru* 93105 points

    Hello Ruixin, 

    A correctly applied and configured TLV2333 op amp circuit should be useable for a 1 kHz and a gain of 6.8 V/V.

    You mention you are powering the TLC2333 with a single +3.3 V supply. What is the amplitude of the 1 kHz or 50 Hz waveform? Is the input waveform centered around 0 V, or is there a positive DC level added such that the common-mode voltage (Vcm) is raised above ground?

    I see a noisy sine waveform on the top traces images, and then what appears to be a clipped sine on the lower traces. If the input sine wave is centered about 0 V and not raised to be within the TLV2333 linear Vcm range [(V–) – 0.1 to (V+) + 0.1 V] the output waveform will be clipped. And if that happens the op amp's normal operation will be badly disrupted and its electrical parameters will be severely degraded. The output signal in that case will not only show the clipping, but may now show the normally attenuated internal switching transients.

    The TLV2333 is an auto-zero op amp and uses relies on internal switching techniques to minimize its voltage offset. The LMV358 and OPA2365 are conventional op amps and do not rely on internal switching for voltage offset reduction. If their Vcm range is exceeded they too should exhibit adverse input/output effects.

    Regards, Thomas

    Precision Amplifiers Applications Engineering

  • 0
    TI__Guru 71711 points

    Adding to Tom's answer, I believe the issue you allude to is caused by three factors:

    1. You reference the output voltage to ground (see below bottom of R4), which would require the 250mV input signal to go to +/-1.7V (+/-250mV * 6.8), which is not possible on 3.3V single supply - see below using ideal op amp.

    Since the output cannot get all the way to ground, let alone below it, Vout saturates 29mV above ground (see red circles below).  Additionally, it shows overload recovery distortion as it is pulled up above its negative rail - see distortion in green circles (TLV2333 is based on OPA2333).

    2.  In order to accommodate +/-250mV input signal, you must reference the output to mid-supply - see Vref.  As you may see, under such conditions, the output attempts to go from -50mV to 3.35V; Vout = Vref + Vin*Gain = 1.65V+/-250mV*6.8 

    Since TLV2333 linear output range (see AOL condition below) requires it to be at least 100mV from either rail, you must lower the gain or the input amplitude to assure output to be within 100mV to 3.2V range.

    3.  Lowering the input signal to +/-200mV will result in output of  +/-290mV to 3.01V, which is within the output linear range - see below linear operation of TLV2333.

  • 0
    TI__Intellectual 2535 points

    Hi Thomas, Marek,

    Really appreciated your patient answer! I totally understand it.

    I'd like to require you about the recovery time of LMV358. I noticed that in datasheet of tlv2333, there is a overvoltage recovery figure. But there isn't any description of overvoltage/overload recovery in LMV358 datasheet. Is there any other parameter could be used for implying recovery time?

    Besides, I'd like to learn that what affect the overvoltage recovery speed of amplifier. Why does tlv2333 has over 50us recovery time but for example, opa2365 has only less than 1us recovery time?  Does bandwidth or RRO affect it? 

    Looking forward to your reply!

    Rayna 

  • 0 in reply to Ruixin Feng
    TI__Guru* 93105 points

    Hello Ruixin,

    The TLV2333 and OPA2365 are very different operational amplifiers in terms of how they are intended to be applied in applications. The OPA2365 is a high-speed op amp having a 50 MHz gain-bandwidth, while the TLV2333 has a 350 kHz gain-bandwidth. Note that the TLV2333 quiescent current is typically 17 uA per amplifier, while the OPA2365 quiescent current is typically 4.6 mA per amplifier, or about 270x that of the TLV2333.

    Operating current almost always is a good indicator of speed and bandwidth and that is the case for these two op amps. Higher current, often equates to higher speed and bandwidth capacbilities. It is no surprise that the OPA2365 overload recovery is less than 1 us, while for the TLV2333 it is on the order of 50 us based on their huge operating current differences. Basically, speed is a matter of being able to quickly charge and discharge the internal junction capacitances of the op amp's internal transistors. The TLV2333 is an auto-zero design that employs thousands of transistors, while the OPA2365 is a more conventional design employing far, far less transistors.

    The LMV358 is a legacy, general purpose bipolar op amp. It is supported by TI's General Purpose Op Amps group. Likely overload recovery was not an important parameter relative to the applications where it would be used so it wasn't included in the datasheet. You might want to contact the General Purpose Op amps e2e forum with your questions about it if that is important to you.

    Regards, Thomas

    Precision Amplifiers Applications Engineering

  • 0 in reply to Ruixin Feng
    TI__Guru 71711 points

    Adding to Tom's reply, since the settling and the overload recovery times are inversely-proportional to op amp quiescent current, IQ, one could estimate them. Thus, based on IQ difference between TLV2333 and LMV358 the overload recovery time for LMV358 to first order should be around 4us; t=~(17uA/210uA) * 50us.

  • 0
    TI__Intellectual 2535 points

    Thanks Thomas, Marek!

    It's really helpful!