• State Resolved
  • Locked Locked
  • Replies 8 replies
  • Subscribers 50 subscribers
  • Views 1436 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.

Using TL3016ID to convert 2MHz sine to square(0-3.3V level) clock

Anoop Joshi
Expert 1115 points
Other Parts Discussed in Thread: TL3016, TLV3501

Hi,


In an application I'm required to generate a 2MHz 3.3V CMOS square wave clock from a sine wave of the same frequency. I was considering using TL3016ID ultra fast precision comparator for this application, by setting VCC+, VCC- as +3.3V and -3.3V respectively. Can you please say if it is fine, can I expect to obtain 3.3V CMOS levels at the output with these supplies?

Also, I understand that all the specifications provided in the datasheet are for +/-5V supplies, is the performance expected to suffer significantly if I use +/-3.3V supplies, especially in terms of jitter, rise/fall time etc?

Plus, can you comment on the internal hysteresis specification for this part, how much input noise can it tolerate without resulting in faulty switching?

Thanks,

Anoop

  • 0
    Expert 1115 points
    Hi,

    In addition to the previous queries, if the usage of +/-3.3V supplies as mentioned in the previous post is fine, can you please say if I can expect to obtain the same level of performance with a 0-3.3V CMOS level output even with a single supply of 3.3V, ie, +VCC=3.3V, -VCC=GND.

    Thanks,
    Anoop
  • 0
    TI__Guru* 93105 points

    Hello Anoop,

    We don't information about operating the TL3016 with +/-3.3 V supplies, but some safe assumptions can be made based on the +/-5 V supply operating information. The TL3016 will function well with +/-3.3 V supplies, but do expect the ac performance to slow down some because the operating current drops with lower supply voltages. It is difficult to say how much things will slow because that hasn't been characterized, but I wouldn't expect the slow down to be large.  

    The output voltage swing will be limited to the same constraints that are had with the +/-5 V supplies. The low level output will still be about 500 to 600 mV above ground, and the upper level will be limited to about 1.1 to 1.6 V below V+. It will depend on the current the output stage provides as to how close it swings towards ground and the positive supply level. If we assume the worst, then it would swing from 600 mV to 1700 mV which aren't CMOS logic levels.

    The TL3016 does not have built-in hysteresis. It could be added externally if desired. Using external hysteresis allows one to set the level of noise tolerance to what they need. We have a very good TI Precision Design on the subject of comparator hysteresis:

    http://www.ti.com/tool/TIPD144

    Regards, Thomas

    PA - Linear Applications Engineering

     

  • 0 in reply to Anoop Joshi
    TI__Guru* 93105 points

    Hello Anoop,

    The TL3016 is not specified for single +3.3 V/GND supply operation.

    Regards, Thomas

    PA - Linear Applications Engineering

  • 0 in reply to Thomas Kuehl
    Expert 1115 points
    Hi Thomas,

    Thanks a lot for all the details!

    Thanks,
    Anoop
  • 0 in reply to Thomas Kuehl
    Expert 1115 points
    Hi Thomas,

    How about using "TLV3501" for the same application, with a 3.3V single supply( ie +Vs=3.3V, -Vs=GND). The datasheet does mention that it is suitable for single supply operation between 2.7V to 5.5V, and also that the output voltage swings within a max of 50mV from either rail. So this part looks ideal for my application of converting a 2MHz sine wave into a 3.3V CMOS level square wave?

    Thanks,
    Anoop
  • 0 in reply to Anoop Joshi
    TI__Guru* 93105 points

    Hello Anoop,

    The TLV3501 comparator is an excellent fit for your application.

    Regards, Thomas

    PA - Linear Applications Engineering

  • 0 in reply to Thomas Kuehl
    Expert 1115 points
    Hi Thomas,

    It seems I missed out on the jitter aspect here. It seems from the jitter perspective, this may not be a good idea because the Input slew-rate limited jitter J_in = V_hyst / SR, will be pretty significant here?(For a 10-MHz, 0-dBm sine wave, assuming 7mV hysteresis, the jitter comes out to be ~400ps?, referring to e2e.ti.com/.../163080). Can you please say if my understanding is correct on this? I was looking for a jitter <15ps or so for my application, so this may not be a suitable choice from a jitter point of view?

    Thanks,
    Anoop
  • 0 in reply to Anoop Joshi
    TI__Expert 6275 points
    I might also recommend you post this on our Clock and Timing forum with details on the input waveform amplitudes as they may already have a clever solution.