Post Go back to editing

Need help on AD8036 chip application

Hello:

I am using an AD8036 clamping amplifier in this scheme

It is used to clamp a square wave with boundaries +3.8V and +1.2V at 9KHz. Output is connected to a UMJ-123-D14-G VCO.  This is the output I get:


Now I got very strange rising edge and falling edge (please click on the figures to get a bigger one).

Could someone please explain the reason and give me some advise to solve this problem?  I need high speed, high accuracy on timing (low jitters of rising and falling) and very stable voltage step.  Change another chip is also appreciable.

Sincerely

  • Hi, abforce.

    Can you modify your circuit by doing the following:

    1. Add a parallel combination of 10uF and 0.1uF between +Vs (pin7) and ground. Do the same to -Vs (pin4).

    2. Add 0.1uF between VH (pin8) and ground. Do the same to VL (pin5).

    3. Add 0.1uF bypass capacitor between each input pin (pin 2 and pin 3) and ground to have a stable response.  

    Let me know if this works.

    Regards,

    Anna

  • Hello Anna:

    Thanks to your reply! I should post the full schematic.

    C4,C5,C13,C14 are use to bypass the VCC and -VCC in parallel and are close to power pins while C15 and C16 are used for pin8 and pin5. I do followed the datasheet, All bypass capacitors are within 3mm range to respective pins.

    But I haven't try method 3. And I'll try it now. Just give you the result as soon as I have it.

  • Hi Anna:

    3.Add 0.1uF bypass capacitor between each input pin (pin 2 and pin 3) and ground to have a stable response.

    My signal is attenuated significantly and turn into a triangle wave. But anyway, the strange wave form is gone. However I paid a huge price in timing performance.

  • Hi, abforce.

    I think there's nothing wrong with your circuit. I guess the cause of the ringings are the circuit connections.

    Have you fabricated your circuit in a PCB? If yes, please check your PCB traces. The shorter and the nearer to the DUT pins, the better. Add a ground plane, place feedback resistors as near as possible to the inverting pin and follow other layout considerations on page 21 of the AD8036 datasheet. You may also refer to this good article for PCB layout - http://www.analog.com/library/analogdialogue/archives/39-09/layout.pdf

    If you are evaluating your circuit in a breadboard, build it in a vector board instead because breadboards are very noisy and very unstable. If possible, fabricate it because high speed measurements are critical.

    If you are using probe for your measurements, replace the alligator clip (ground lead) with the probe tip ground (like a coil of wire) and insert it in the probe. This probe setup will give you a stable and precise waveform.

    Also if you are using connectors for your supply, inputs, etc., use short connectors or if possible use SMA for SMAs are fully terminated (50ohms) and are very precise.

    Hope this helps! Apology for the late reply.

    Regards,

    Anna

  • Thank you for your reply!

    Now I'm checking the PCB layout. If I want to connect the output to another voltage feedback op-amp, is there any technique document indicates in this condition? I want to know what kind of connection is need  to get a better performance, i.e., an isolation resister next to the output pin, a output resister connected to ground or just connecting the output pin to the input pin of the other Op-Amp? Thank you!

    Ran

  • Thank you for you reply!

    Your advise solved my problem, with a little price of getting a slower transition. In fact I notice that there is a little 'peak' remained in the rising edge, just like the small sharp peak appeared in the third picture, I mean, the one before the later rising edge.