Post Go back to editing


Hi, please advise why is the Relative Gain is expressed in negative dB in the


data sheet truth table: 

Usually negative dB mean attenuation, however, the device is supposed to be a 45db Amplifier. 

Thank you! 

  • Hi artemen,

    I think that the answer to your question is 1) a matter of the convention we've adopted to describe the part's performance and 2) the nomenclature associated with the individual die within the package.

    The part contains two identical amplifier die plus a single variable attenuator die. (There is also a single die to control the attenuator, but we can ignore it for this discussion). In the typical application circuit (shown beneath my signature) the input signal is fed to one of the amplifiers, then to the attenuator, then to the other amplifier, and then output to the load. The amplifiers provides a combined gain of 45 dB (typ.) while the attenuator provides from 0 dB through 31.5 dB of attenuation. You said "Usually negative dB mean attenuation, however, the device is supposed to be a 45db Amplifier"...and that's exactly how the gain control works. The negative dB values in the Relative Gain Setting column indicates the amount of attenuation that has been switched into the signal path, thus reducing the gain of the amplifier from its maximum gain of...

    45 dB (amp gain) - 0 dB (attenuation) = 45 dB gain its minimum gain of...

    45 dB (amp gain) - 31.5 dB (attenuation) = 13.5 dB gain

    In a sense we've described the amp's operation in terms of how it works. Alternatively, we could have defined the Truth Table to show the individual gain states relative to the amp's minimum gain state, which would have resulted in all of the non-zero Relative Gain Settings having positive values.

    In summary, it's all a matter of convention.

    I hope that I've adequately answered your question.



  • Thank you for your prompt and detailed reply! 

    However, it leads to another question:

    my application uses a transceiver module and I need to be able to receive  non-attenuated signal from the same antenna, will HMC681ALP5E introduce any attenuation when in non-max amplifier gain mode?

    Thank you! 

  • Hi artemen,

    I'm not sure that I correctly understand your question. Using the typical application circuit from the data sheet, the minimum gain setting is +13.5 dB (typ.) and the maximum gain setting is +45 dB (typ) there is no state in which the (typ.) net gain would be less than +13.5 dB. The setting of the attenuator (variable from 0 dB through 31.5 dB) is what controls the gain. Because the part provides access to the individual RF inputs and RF outputs for both amplifiers and for the variable attenuator (i.e., six RF ports in total), you could configure your signal path so that it differs from the typical application circuit.

    If I've misinterpreted your question, then please try again.



  • Sorry, you mentioned 6 RFports, I see only RFin and RFout in the datasheet... 

    my question was: since I also need to be able to receive signal from the same antenna (the IC I am using is a transceiver), is there a mode in which HMC681ALP5E  doesn't block received signal? 


    though; I've found the 6 RF ports you've mentioned RFin1,2 RFout1,2, ATTin, ATTout, however it is not clear how can I change the RF flow inside the chip. Would you point me into the right info? Thanks! 

  • Hi artemen,

    For the stock configuration of the evaluation board:

    • The RF signal is input to connector J1 (RFIN), which feeds the signal into pin 2 (RFin1)
    • RF signal from pin 4 (RFout1) is fed into pin 6 (ATTin)
    • RF signal from pin 19 (ATTout) is fed into pin 20 (RFin2)
    • RF signal from pin 22 (RFout2) is output from connector J2 (RFOUT)

    The HMC681ALP5E does not contain any type of RF switching. The RF flow within the chip itself cannot be changed:

    • The RF signal entering pin 2 (RFin1) can only exit pin 4 (RFout1)
    • The RF signal entering pin 6 (ATTin) can only exit pin 19 (ATTout)
    • The RF signal entering pin 20 (RFin2) can only exit pin 22 (RFout2)

    Because I don't understand how you intend to use the part, it might be helpful if you could show me a block diagram of your system.

    Would that be possible?



  • That's what I thought based on the DT, though you wrote the following in your previous reply: "Because the part provides access to the individual RF inputs and RF outputs for both amplifiers and for the variable attenuator (i.e., six RF ports in total), you could configure your signal path so that it differs from the typical application circuit."

    I assume I will put this amplifier befor the txrxswitch. My original thought was to make a stand alone version. 

    And the last quesrion, how many dB can be supplied into the RFin? 

  • Hi artemen,

    There is an important distinction to be made between the HMC681A part and the HMC681A EVB. The part itself allows access to the individual RF inputs and outputs for both amplifiers (let's call them Amp1 and Amp2) and for the variable attenuator (let's call it ATTEN). For example, if you designed your PCB to do so, you could use just Amp1 followed by ATTEN. Also, if you designed your PCB to do so, you could use just ATTEN followed by Amp2. Or string them up differently. The four die within the package effectively constitute three independent blocks: Block1=Amp1, Block2=(Serial controller + variable attenuator), Block3=Amp2. On your own PCB you could configure them in any way you desire, but on our EVB the only available configuration is RFIN>Amp1>ATTEN>Amp2>RFOUT.

    Regarding the maximum signal amplitude applied to the input of either of the two amplifiers, see the bright yellow highlighted sections of the data sheet's AMR table below. You will also want to be aware of the linearity (e.g. P1dB & IP3) associated with each gain setting. For that information see the data sheet plots, which contain data for operation using the data sheet's Application Circuit.



  • that is negative dB again, correct me if I am wrong this means that I'll have to attenuate my signal to at least -10db? 

    so this is not a TX amplifier, but a Receiver amplifier? 

    so, according to [1] if I input at 10db, I will have only 20db amplification? 

  • We need to be clear regarding dB and dBm. The RF Pin and Pout values are in dBm, while the gain and relative gain are in dB. Due to the part's OP1dB (which is in dBm), you will never get Gain = +20 dB when using RF Pin = +10 dBm.

    In the max gain state, you need to limit the RF Pin to -10 dBm. This is the RF Pin AMR for the max gain state. As the gain setting decreases by X dB, the RF Pin AMR increases by X dB, to a maximum RF Pin of +10 dBm. 

    I have created and attached a spreadsheet showing the Control Word, the state of each attenuator stage, relative gain, reference gain, small signal gain (absolute), and RF Pin AMR for that gain state. Maybe that spreadsheet will help you to better understand the part's operation.



  • Hi Artemen,

    Couple of things - From what you wrote it looks like you are trying to connect HMC681 to a received signal. Usually you will be lucky if you have -20 or -30 dBm signal coming in. Also to avoid unwanted signals getting in to HMC681 you will have to use some sort of Filter to let through only the wanted signal. Now  a little bit about signal levels in dBm.

    Let us say signal is like water level in a dam. When there is no water, it is at -174 dBm. (if you were letting water through a 1 Hz wide window, so it is called -174 dBm/Hz. If you increase window to 10 Hz the total power inside the window will increase to -164 dBm - ie -174 dBm + 10*LOG10(Window width). By this logic if you were to look at the bottom of the dam with absolutely no useful signal (noise only) through a band pass filter that is 1 MHz wide the noise power observed at the output of Band Pass Filter will be -174 +60 = -114 dBm. Now imagine you have a single CW signal of - 40 dBm coming inside this Band pass filter. Then the signal to Noise would be {-40 dBm - (-114 dBm)  = 74 dBC meaning 74 dB between Carrier signal and noise power). Now let us consider that you did not have a Band Pass Filter at the Input of HMC681, then all signalsor noises from near DC to 1 GHz (the upper operating frequency of HMC681) will get in to input of HMC 681. Now noise power going in is -174 dBm + 10*LOG10(1 GHz) = -174 +90 = -84 dBm. For the same -40 dBm signal as before with out an input filter, now you have only 44 dBC signal to noise ratio.

    OK. Now let us consider how HMC681 works. It receives the -40 dBm signal and Amplifies by a gain variable from 13.5 dB to 45 dB. thus the signal output can vary from (-40Bm+13.5 dB =) -26.5 dBm to (-40 dBm +45 dB = ) +5 dBm. How it achieves this variable gain is what Analog Devices Engineer explained above. For every amplifier there is maximum input signal it can receive with out saturating or damaging the amp. In this case what he is saying is the signal level at the input should not exceed -10 dBm. Now you have to be careful. If the signal was at -10 dBm and if we set the gain at 45 dB, we should get -10 dBm +45 dB =+35 dBm. But how ever HMC 681 can only deliver +19 dBm even when it is slightly compressed. So you can see that we can not set gain greater than 29 dB for a -10 dBm input signal.

    Analog devices could have made our life a little bit better by stating what is the gain and P1dB & NF for the First Amp and same thing for the second Amp. But I assume the NF of First stage must be around 2.5 dB and gain of 20 dB or so. Let us hope SMcBride may be able to give us an answer.