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.

LMH5401: Cannot reproduce the specifications in the datasheet of LMH5401

Part Number: LMH5401
Other Parts Discussed in Thread: TINA-TI, LMH6401

Hi Team,

We have received an inquiry from our customer as follows.

We are trying to use an LMH5401 in one of our designs and neither our design, nor the EVM produces what the data sheet says it should. Need to figure out why that's so.

We are trying to use the LMH5401 in our design. Our frequency range is 30 MHz to 6 GHz. What we find is that the system seems to have a rapid roll off at frequencies of more than about 4 GHz. We've tried the designs in the data sheet and also the 5401EVM and none of them produce the performance as specified in the data sheet.

As a starter the data sheet says that Figure 1 was derived from measurements made using the EVM. We cannot reproduce that graph for the 2V/V and 4V/V curves using the schematics provided in the data sheet Figure 56 and the EVM. That would be a good starting point. Also we actually need to cascade two or more amps to get the gain required and we find that cascading more than one amp drops the 3 dB frequency very fast. Even the TI-spice seems to show this behavior so there seems to be something that's not mentioned in the data sheet. For instance does this part have a very high input capacitance?

Thank you for your support!

Regards,

Danilo

  • Hello Danilo,

       Is this the same customer and issue as in this thread? Would you be able to share the simulations files of the reproduction Figure 56 and the EVM? Also any simulation files of the cascading amplifier design. 

    Thank you,

    Sima 

  • Morning Danilo, yes this discussion has been going on for a bit - might be nice to have the design requirements - gain, BW, Swing, Load 

  • Mike,

    If this is me, and it seems to be, my desired frequency region (instantaneous bandwidth) is from 30 MHz to 6 GHz. I'll take a gain of 2 V/V or 4 V/V or whatever I need to be able to get the 6 GHz 3 dB. 

    The eventual overall gain I need is about 20 dB so I suspect I'm going to have to cascade amps, and that's fine if I can get a cascade to operate fine.

    The output is to a 100 Ohm load on each of the two output ports

  • Sima,

    Ask and you shall receive

     6366.LMH5401 Eval.TSC

    This one is the example TI provides for the TI-spice for this part. You'll notice this doesn't even produce the specified 3dB bandwidth of 6 GHz.

    Here's an output and the simulation file for a system cascading 3 amps. It's interesting to note how the slope of the gain curve at the high frequency end changes as each amp is added to the chain.

     Tiger exact vals 3 stages.TSC

    We get results reasonably close to the simulation when we measure the EVM and a board of our own design.

  • Please see the original thread. This appears to be a load impedance issue. Please ensure that you are presenting a 200 Ohm differential load to the device to achieve the desired bandwidth.

    Best,
    Sam

  • We have the right load. In fact if you look at the sim file I attached (which is TI's) you'll see that. And the simulation doesn't give the results TI specifies. So that isn't it. It's something else.

  • Hello Ash,

       Thank you for the simulation files, I will look into it and get back to you with an update by end of day today.

    Thank you,

    Sima

  • Hello Ash,

       Sorry, it took me a while to find the original EVM/datasheet measurements. If you are able to accept my friend request, I can share the original excel file of Figure 1 privately. It looks like before the gain was normalized, the gain using resistance values in Table 2/Figure56 for 4V/V gave a total output gain of 3.14dB. The -3dB point for the 4V/V is around 6.2GHz. I am awaiting response from my coworkers who characterized this device on how they setup the device for it to output 3.14dB rather than 6 or 12dB. Tina-TI matches this frequency cutoff at this gain which corresponds to the scope measurements.

       In the meantime, I am looking at adjusting your three stage design to achieve a higher bandwidth. This reference design uses LMH5401 and the LMH6401 cascaded to achieve a high variable gain at a high bandwidth. From earlier threads, it looks like you need a bandwidth of at least 6GHz? What is the necessary output voltage? 

    Thank you,

    Sima 

  • Sima,

    I would like to get a total gain of 25 V/V for the differential gain.

    The output voltage we are looking for ranges from 50 - 100 nV/√Hz upwards depending on the frequency (Lower frequency has a higher input voltage ~ 3-4 nV/√Hz) .

    The frequency coverage is from 30 MHz - 6 GHz

    Thanks

    Ash

  • Hello Ash,

       Unfortunately, a  frequency at 6GHz will be difficult to achieve with this setup. Even if we are able to get 6GHz for each 2-4V/V gain stage, at later stages in the design this will not follow a small signal frequency response plot, but the large signal frequency plot shown in Figure 2 where 2-4V/V gives a bandwidth of around 5GHz. 

    Thank you,

    Sima