• State Resolved
  • Locked Locked
  • Replies 8 replies
  • Subscribers 47 subscribers
  • Views 3794 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.

Designing a simple integrator with THS4211

Jim Huang77508
Prodigy 55 points

Other Parts Discussed in Thread: THS4211, OPA656, TLV3501, TINA-TI

Hi All,

   I am putting together a simple triangle wave generator for a bigger project. I already have a clean 5V, 20MHz sqaure wave with 7nS risetime/falltime. I am tyring to use the THS4211 as the opamp for the integrator portion to generate the triangle wave. I am using the model in TINA to verify the resistive/capacitive values.

   As of now, I have 390 Ohms (R1) at the inverting input, 2.5VDC at the noninverting input node, and 32 pF feedback (C1), with 1MOhm (R2) around as well to set the DC operating point. The pole formed by R2, C1 sets my near DC pole, and unity gain bandwidth is set by C1, R1. I do not have a load. I am simply scoping the output node of the THS4211. I am powering the opamp to: 0V on GND, and 5V on VCC.

   My output looks nothing like a triangle wave, and I have tried to debug it for a while now but to no avail. Any tips on any glaring error in my setup? I checked the operating Iq of the part and it is within normal spec.

Jim

  • 0
    Genius 5430 points

    Jim;

    The input bias current of the fast bipolar op amp THS 4211 is driving it into the rail. Using a fast FET op amp, OPA656, in a somewhat similar circuit, it works well. I used a bipolar power supply but I'll leave it to you to modify the circuit for your application.

    Fast Integrator.TSC
  • 0 in reply to Neil Albaugh
    TI__Genius 17905 points

    Jim,

    Neil is correct that the input bias currents of the THS4211 can create an offset that drives the output to the rail. Is that what you are seeing, that your output rails?

    One way to fix this would be to add the same value resistor to the inverting input in series with your 2.5V reference at the non-inverting input. This will result in bias current cancellation and recenter your output at 2.5V. See the attached TINA where I have tweaked the R and C values.

    I am also concerned about your input signal - you have a 20MHz square wave with 7ns tr/tf? A 20MHz square wave would imply a 50ns period or 25ns high, 25ns low. 7ns of rise and fall time is a pretty significant portion of this high and low time and the output will be rounded during these 'slow' edges, causing your output to look more sinusoidal. You may consider 'buffering' your square wave input with a fast comparator such as the TLV3501 to get those edge rates up to ~2ns. I have shown the simulation results for 2ns input rise/fall times and for 7ns input rise/fall times in my TINA simulation.

    THS4211 Integrator 6-23-2011.TSC
  • 0 in reply to Kris F
    Genius 5430 points

    Jim;

    Kristoffer is correct about the rise & fall time of your input 20MHz square wave-- in fact, I reduced my signal rise & fall time to 1ns so that the integrator output looked like a true triangular wave; 7ns gives a rather goofy- looking output waveform.

    I blindly assumed that your 20MHz square wave was a bipolar +/-2.5V waveform but Kristoffer used a 0/+5V waveform.... one of us is right but I don't know which  :)

  • 0 in reply to Neil Albaugh
    Prodigy 55 points

    Thanks for the tips from both of you. I'll condition the 0-5V square wave signal with what Kristoffer mentioned, and also select one o the OPA JFET input opamp parts.

    Going to head over to the lab now to try out the resistive to ground to see if it alleviates the base current of 15uA on the THS4211 part. Thanks again!

     

    Jim

  • 0 in reply to Jim Huang77508
    Prodigy 55 points

    Having trouble opening the attached circuit schematics form both of you. Is there a compatibility issue with TINA files? I am using TINA7.

    Also, I attached the TLV3501 part to sharpen up the rise/falltimes of the 20MHz square wave. But when I hooked it up to the input resistor of the integrator portion, my triangle wave looks strange. It looks like the slope (I/C) is too high, and it doens't change when I change the capacitor value. This is not an issue if I use an ideal Vsource with programmed 2nS rise/falltimes and 2.5V VDC with 2.5V amplitude square wave.

    I tried uploading the schematic but firefox is giving me fits and I am uploading the waveform/schematics here instead. Sorry for the inconvenience... thanks as always for your help!

    Jim

     

  • 0 in reply to Jim Huang77508
    Genius 5430 points

    Jim;

    If you download the latest revision of TINA-TI, version 9, you can open them. Files created with v7 can be read on v9 but not the other way around.

    Did you notice that VF1 is trying to swing below ground? This is a no-no for a single-supply op amp.

  • 0 in reply to Neil Albaugh
    Prodigy 55 points

    Thanks for the response. I was also looking into that "below the rail" weirdness. Here's the final schematics/waveform. And I updated my TINA to version 9 as well so I was able to view both your schematics. Thanks again for the tips. Have a great weekend!

    Jim

  • 0 in reply to Jim Huang77508
    Genius 5430 points

    You're welcome, Jim. It looks like you're circuit is working OK now.

    Have a good weekend yourself-- tomorrow I can get back to work on my race car.