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.

TINA/Spice/LMH6554: 1GHz amplifier response

Part Number: LMH6554
Other Parts Discussed in Thread: TINA-TI,

Tool/software: TINA-TI or Spice Models

Hi,

I am trying to simulate in Tina-TI an amplifier with about 12dB gain upto 1GHz using the LMH6554.

The input impedance is 50ohm and the output is 100ohm.

I gain and phase response is rolling off at around 300MHz and cannot see what is wrong.

Attached is our file for the simulation.

Can someone please help to identify where we are going wrong.

Thanks,

BPLMH6554_1GHz.TSC

  • Bharat,

    An AC sim of your circuit shows bandwidth out to 1GHz, rolling off after that.
    The plot appears below. How does this compare with your results and your requirements?

    Regards,
    John


  • Thanks John.

    That is what we get too but we need it to be less than <0.5dB down at 1GHz.. We could not find any component that would increase the response like plot Av=4 in Fig 2 of datasheet. Any extra peaking beyond 1GHz is OK for our requirement.

    Regards,

    BP

  • Hi,

    Is there any TINA simulation file that can give a response like Figure 1 (Rf=250) or Figure 2 (Av=8 V/V)

    Kind regards,

    BP

  • BP,

    At this point it would be best to engage with the product experts at this point.
    I say this because it can be tricky to correlate model and device behavior at frequencies at and above 1GHz.
    I will transfer this thread to the High Speed Amplifier E2E forum. The engineers that monitor that forum will be able to help.
    Many thanks for your patience.

    Regards,
    John

  • Thank you John.
    We are hoping to resolve this quickly.
    Kind regards,
    BP
  • Hello BP,

    I talked to John on the phone this morning about your post.  I am reading over this thread and starting to run some simulations so that I can help answer your questions.

    I was the applications engineer who wrote the datasheet for the LMH6554 and also validated the LMH6554 TINA model. At the time this product was released we were National Semiconductor, so the model was not validated in TINA-TI, but rather in PSpice.

    Also, the LMH6554 model was not designed to duplicate the datasheet response, but rather it was designed as a physics based model to closely replicate the performance of the LMH6554 device when NOT mounted on a PCB. This means that the interaction of the circuit board and the amplifier is not reflected in the device model. This also means that the model response was not expected to exactly duplicate the datasheet curves.

    In reading your first post it looks like you want to duplicate the datasheet response at a gain of 4V/V and to see the model simulate a bandwidth of 2.1GHz at that gain condition?

    I will see if I can get a TINA-TI response that duplicates that shown in the datasheet. I think that this will require duplicating the parasitic inductance and capacitance of the PCB in the TINA-TI model.



    Regards,
    Loren

  • LMH6554_1GHz_lds.TSCHello BP,

    In your attached file you are using a differential input signal source and the load resistance is 10k Ohms.  Can you confirm if you want a single ended or differential input design? 

    I am attaching a schematic that is single ended input and more closely duplicates the datasheet curve.  My attached circuit includes termination resistors on the load.   This means that the load that is not on the PCB is only 100 Ohms and there are load impedance matching resistors on the PCB.  This matches the laboratory setup.  

    My attached schematic also includes a single ended signal source.   This also matches the datasheet conditions.  I am getting similar results to the datasheet by using 4 capacitors to bypass the gain and termination resistors.  This simulates some of the parasitic capacitance of the circuit board we used. 

    Keep in mind that if you use a different PCB layout in your final design your results will be different from our datasheet curves as well. 

    Regards,

    Loren

  • Hello Loren,

    Thank you for the quick reply. We are looking for a fully differential design with input (50ohm) and a differential output (100ohm). Ideally a flat response to 1GHz.

    Kind regards,
    BP
  • 8524.LMH6554_1GHz.TSCHi Loren,

    I have updated the design (attached) with your suggestion, it looks better. I would like 9dB gain so is it best to drop the input resistors?

  • Hello BP,

    Please use the attached excel spreadsheet to select resistor values for your 9dB 50 Ohm input, 100 Ohm output design.  Note that when using a terminated load you will have 6dB of loss so that you will need to target 15dB of total gain (5.6V/V).

    The capacitors I added in the TINA model are used to trim the model performance and should not be needed on the actual design.  These components represent PCB effects, not real components that you would add to the design.  You can use our evaluation board as a sample board layout.  If you use our evaluation board and the resistor values calculated in the attached spreadsheet you will not need any capacitors to get the response that is shown in the datasheet. 

    Can you give me more details on your application?  Will you need a filter between the amplifier and the load?  Do you need termination resistors, or is the load very close to the amplifier (under 1cm)? 

    Your attached schematic is using a differential input.  Using single ended input signals requires a different resistor configuration as detailed in the spreadsheet and shown in the attached TINA schematic. 

    Regards,

    Loren

    1374.FDA single to diff resistor calculations with BW and Noise calculations to share.xlsx

    LMH6554_1GHz_9db.TSC