[USRP-users] Output power of E310

Marcus Müller marcus.mueller at ettus.com
Tue Mar 15 12:50:51 EDT 2016

You're asking the right questions!
So, no, as far as I can tell, they're not generally "wrong"; they still
might be wrong for your use case, though!
Multiple aspects play together, here:

1. Frequency dependency of the device as is:
Generally, output power does change with frequency; it's practically
impossible to avoid that with a budget. Amplifiers have their own
response, so do transformers, the quality of matching of devices, PCB
traces, SMA connectors, package pads etc changes with frequency too, and
so do the characteristics of semiconductor mixers.
2. The AD9361, as fantastically as it spans 70-6000 MHz, already has
three different ports for three bands for the above reasons. Behind
these three ports, there's different hardware, so they behave
3. The filter responses of course aren't always perfectly flat, and
aren't calibrated to match their passband attenuation.

We're currently only addressing point 3., because frankly, it's the only
thing that UHD can influence.

That being said, the upper frequencies used for the different filter
paths (450, 700, 1200, 1800, 2350, 2600, and 6000 MHz) aren't
necessarily "good" for you. If you're observing e.g. spectrum a bit
above and a bit below one of these frequencies, then switching between
the different filter paths will leave you wondering why your signal
power suddenly changes.

Best regards,

On 03/15/2016 05:37 PM, Claudio Cicconetti wrote:
> Dear Marcus,
> I suppose modifying the UHD code would not be a blocking issue (though
> having an API would definitely make one's life easier): by following a
> mix of manual + Philip instructions I was able to build an image from
> source code.
> However, I am still not getting if the current implementation of the
> filter boundaries are i) wrong, or ii) wrong under some circumstances
> such as mine, or iii) correct for all cases but I am missing something else.
> Best regards,
> Claudio
> On 03/15/2016 05:30 PM, Marcus Müller via USRP-users wrote:
>> The point still is that to change the boundaries at which a different
>> filter is used, you'd have to modify UHD's e300_impl.cpp around lines
>> 1160ff. UHD currently doesn't expose an API to do so.
>> On 03/15/2016 04:30 PM, Marcus D. Leech via USRP-users wrote:
>>> Ah, found it.
>>> There's an update_bandsel() method in e300_impl.cpp
>>> On 2016-03-15 11:23, Marcus D. Leech via USRP-users wrote:
>>>> Claudio:
>>>> I'm hunting through the code to see where it selects the filters
>>>> based on frequency.  All of this code was recently re-factored, and I
>>>> have a query in to R&D to figure this out.
>>>> There doesn't appear to be a top-level "set_filter()" although such
>>>> is used internally in the drivers, which select the appropriate
>>>> filters automatically.
>>>> On 2016-03-15 07:15, Claudio Cicconetti wrote:
>>>>     Dear Marcus, all,
>>>>     I have upgraded to latest (Rev4) image, the results are basically the
>>>>     same: output power varies significantly and non-monotonically with
>>>>     center frequency.
>>>>     Is there any (simple) way to change the configuration of filters?
>>>>     Finding a "good" configuration for a given frequency with trial&error
>>>>     would be acceptable for us, since the center frequency is fixed and
>>>>     known in advance.
>>>>     Btw, I tried to generate the image from source following the
>>>>     instructions in the manual, but the compilation fails ('unknown type
>>>>     name' errors when compiling volk).
>>>>     Best regards,
>>>>     Claudio
>>>>     On 03/14/2016 02:25 PM, Marcus D. Leech via USRP-users wrote:
>>>>         On 03/14/2016 04:04 AM, Claudio Cicconetti via USRP-users wrote:
>>>>             2. getting a pre-built image from
>>>>             http://files.ettus.com/e3xx_images/ If the option 2. is
>>>>             the correct one, can you please provide a pointer to the
>>>>             correct image to copy to the SD?
>>>>         I'd use the latest Rev4 image:
>>>>         http://files.ettus.com/e3xx_images/e3xx-release-4/ettus-e3xx-sg1/
>>>>         Use the "dev" image
>>>>             Best regards, Claudio
>>>> _______________________________________________
>>>> USRP-users mailing list
>>>> USRP-users at lists.ettus.com <mailto:USRP-users at lists.ettus.com>
>>>> http://lists.ettus.com/mailman/listinfo/usrp-users_lists.ettus.com
>>> _______________________________________________
>>> USRP-users mailing list
>>> USRP-users at lists.ettus.com
>>> http://lists.ettus.com/mailman/listinfo/usrp-users_lists.ettus.com
>> _______________________________________________
>> USRP-users mailing list
>> USRP-users at lists.ettus.com
>> http://lists.ettus.com/mailman/listinfo/usrp-users_lists.ettus.com

More information about the USRP-users mailing list