Post Go back to editing

Out of band performance

Thread Summary

The user inquired about using the ADAR1000 for phase shifting and attenuation at 1 GHz, despite its 8-16 GHz range. The final answer suggests using individual vector modulators or a combination of analog/digital phase shifters and attenuators for 1 GHz signals. For unformed signals, an IQ modulator like the ADL5375 or a complete transceiver like the AD9361 or ADRV9009 is recommended. The accompanying answers note that the ADAR1000's performance degrades significantly at 1 GHz, and synchronization of multiple AD9361 devices requires more than just a common clock, as detailed in a specific thread.
AI Generated Content
Category: Hardware
Product Number: ADAR1000

Hi!

Does anyone know what the out of band performance for the ADAR1000 is? Essentially, this is 8-16GHz part. If you were to attempt to use it lower in frequency, let's say 1GHz, could you still use the attenuators and phase shifters? I understand you would be limited by total phase shift, but could you use this to basically fine-tune a lower frequency phase? Attenuation is probably easier, right? That might be less dependent on frequency?

Thanks!

  • There are a number of factors acting against you on this. The phase shifters inside the ADAR100 are IQ modulators. So the RF input signal is split into quadrature components. The accuracy of this quadrature split will affect the accuracy of the phase setting. You can see this in Figure 36. If you are wanting to operate at 1 GHz, that is over two octaves lower than the plotted data. You can see from Figure 36 that even at 6 GHz, things are starting to degrade.

    The other issue is frequency response of the part. You can see in Figure 10 that the response roll-off is quite steep at low and high frequencies. 

    The part will still operate but you are going to see very low output power and the phase shifting accuracy will be poor. I think that your best bet is to get and evaluation board and try this out. 

    Eamon

  • So for lower frequency arrays, does anyone have a suggestion for individual antenna element attenuation/phase control? It would seem to me the only option would be to use discrete phase shifters and VVAs... But this is not my area of expertise. For very large arrays, is it primarily a game of analog splitting/combining and individual antenna gain/phase control or would a better approach be putting a IQ capable source at each element and do the same with IQ impairments with a shared LO?

  • There are some options available to you. 

    1. If the signal you want to transmit is already formed (e.g. you have a QAM carrier at 1 GHz that is coming from somewhere else), you could use
      1. individual vector modulators
      2. analog or digital phase shifter cascaded with an analog or digital attenuator
    2. You can find some options herehere. Beamformers, Phase Shifters, and Vector Modulators | Analog Devices
    3. If your signal is not yet formed but you have the modulation information in I and Q format, you could use
      1. An IQ Modulator such as ADL5375 (www.analog.com/adl5375

        ) at each element. The I and Q inputs of the IQ modulator will be fed with the I and Q data that would have been digitally phase shifted so that the signals at the individual elements are also phase shifted relative to one another.

      2. A complete transceiver such as AD9361 (www.analog.com/ad9361) or ADRV9009 (www.analog.com/adRV9009) (there are multiple options). This is essentially the same as approach a above but you now have IQ Mods, DACS, IQ Demods, ADCs and PLL all integrated into one chip. Also, many of these integrated transceivers have multiple channels in one chip. 

    Eamon

  • I think the AD9361 is a good call, especially if we already have some experience with a Pluto. To get multiple AD9361 devices to output signals in-phase, is it as simple as providing a common clock via the XTAL? Basically, how would one put multiple sources that may be spatially separated to be correlated?

  • The process of synchronization is not trivial. Just because you are running multiple devices form the same external PLL, doesn't mean that the outputs will be phase aligned. The thread below speaks to the process for ADRV9009. My understanding is that this is easier to do on the newer transceivers, that is, ADRV9XXX part numbers.  - Eamon

     

    (32) Multi-ADRV9009 phase synchronization with external LO - Q&A - Design Support ADRV9008-1/ADRV9008-2/ADRV9009 - EngineerZone