• State TI Thinks Resolved
  • Locked Locked
  • Replies 5 replies
  • Answers 1 answer
  • Subscribers 48 subscribers
  • Views 698 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.

Original question:

Switching a 0-28V pulse in less than 10ns

THS3095: Switching a 0-28V pulse in less than 10ns with a Current Feedback Amplifier

Kerem Oktem
Prodigy 240 points
Part Number: THS3095
Other Parts Discussed in Thread: TINA-TI, LMH6881, THS3217

Hello,

I am working on a circuit that has an pulse input from an FPGA with LVCMOS3V3 level. The pulse can range from 50ns to 1ms.

The circuit should convert this LVCMOS3V3 signal to a variable 5V to 28V signal with a rise and fall time less than 10ns. (Level Conversion)

I had 2 solutions in mind

1. Supply -2V negative supply and a variable position supply to the THS3095. Use it in non-inverting configuration and saturate the amplifier with a constant gain and by changing the positive supply I thought I could change the level of the output pulse. In the lab, when not saturated the amp functions perfectly with very low rise and fall times, but when the amp is saturated I experienced longer pulse lengths and unexpected rise and fall times and weird pulse shapes. I get the impression that saturating a current feedback amplifier for a pulse amplification implementation is wrong? Is this true? And why do I have this kind of a result

2. Supply -2V negative and 30V positive supply to the THS3095. Use it in non-inverting configuration and change the gain of the amplifier with a digitally controllable potentiometer. But  it is really hard to find a potentiometers in sub kOhm range to use the suggested resistor value that is stated in the datahseet of THS3095. Also there is a spec in the datahseets in the digital potentiometers that states the bandwidth of the potentiometer. Does the bandwidth of the digital potentiometer affect the response an amplifier in terms of rise and fall times of the output.

I would appreciate and help.

Kind Regards

Kerem

  • 0
    Guru 48395 points

    I sounds like you you really want to use that logic signal to update a DAC output where the DAC is used to generate a variable level that then feeds into a fixed gain amplifier stage. do not try to run the amplifier into saturation as its recovery is not well controlled or specified. 

  • 0 in reply to Michael Steffes
    Prodigy 240 points

    Dear Michael Steffes (or anyone who could help me),

    Thank you very much for your reply.

    I thought for a new solution to my application. I simulated the circuit that I designed on Tina-TI, the results seams good, however I wanted to make sure if I missed anything.

    The block diagram is as follows:

    I am also attaching the Tina-TI simulation to this message.PulseGen.TSC

    I could not really figure out some points.

    1. The datasheet of LMH6881 says in Figure-38 with a single ended input there has to be decoupling capacitor. I believe when there is cap at the input, DC does not passes to the amplifier. In my application I should be able to supply to the input a DC 3.3V.

    2. I need to convert the differential voltage of LMH6881 to single ended for the THS3095. THS3217 is supplied with +-6V in the simulation. When I lower the supply voltages (I believe I should be able to) the simulation says "Convergence Problem". So the question is, I could not find the input voltage requirement like common voltage input voltage and the relation to the differential amplifier.

    3. Oddly enough when I change the rise and fall of the pulse or I use a step signal I get the same error "convergence problem"

    Do you think that this solution would work for my application? Or could there be a better solution.

    Looking forward hearing from you.

    Kind Regards

    Kerem OKTEM

  • 0 in reply to Kerem Oktem
    Guru 48395 points

    Hey Karem, 

    1. Not sure why you are trying to apply the LMH6881, wouldn't a DAC be what you really want? That is what the THS3217 is intended to be used with

    2. If you have a blocking cap at the LMH6881 inputs, you are not DC coupled and cannot get good step responses, 

    3. Convergence? under Analysis at the bottom go to options and change the integration to trapezoidal and the matrix solver to Davis KLU - that works for me a lot of the time,  

  • 0 in reply to Michael Steffes
    Prodigy 240 points

    Hello Michael,

    Thank you very much for your valuable response.

    I agree with you completely at this point after all my attempts. However there is a huge problem why I can not do this solution.

    The board has 16 channels of this 10ns pulses and many other channels. In the end the FPGA board that I am using has no more free inputs and output.

    This is the reason why I am trying this solution.

    In this respect, even if this might not be the best solution to my problem. Do you think that the amplifier combination that I stated earlier, using a variable gain amplifier, would solve my problem?

    With regards to the blocking capacitor, I agree with you completely, however the simulations shows me a great step response with the capacitor. I'm starting think if there is something wrong with the simulator.

    Kind Regards

    Kerem

  • 0 in reply to Kerem Oktem
    TI__Mastermind 23226 points

    Do you have any further questions?