LNA and Noise Figure improvement

Jean-Samuel Najnudel - BJT PARTNERS SARL jsn at bjtpartners.com
Fri Nov 2 10:21:22 UTC 2012


Hi Andrey,

Thank you for your reply.

I agree we do not care very much of 0.2 dB lower NF.
I also understand we cannot really increase the gain.

By the way, do you think this is ok to have the LNA on the UmSEL board ? Do
you think we do not need some specific shielding or separated enclosure, at
least for the 1st LNA stage ? I think this should probably be fine but I
would like to double check and know your point of view about this.

Again, I agree we do not really care about a 0.2 dB sensitivity difference.
However, I am still very interested in your calculations to have an
apporximate idea of what kind of NF we should get in the GSM 1800 band.
For the GSM 900 band, the NF is fine. However, what kind of figures do you
think we could expect in the GSM 1800 band ?

Best regards.

Jean-Samuel.
:-)


On Fri, Nov 2, 2012 at 10:42 AM, Andrey Sviyazov <andreysviyaz at gmail.com>wrote:

> Hi Jean-Samuel.
>
> Regarding extra LNA, I should apologise, but still no time to calculate it.
> On my opinion (first look) this is not good idea.
> We already have quite high gain and pretty good NF.
> And I really can't understand why do you think that it isn't enough.
> Please explain me what do you wish to obtain if NF will be 0.2dB lower.
>
> Best regards,
> Andrey Sviyazov.
>
>
>
> 2012/11/2 Jean-Samuel Najnudel - BJT PARTNERS SARL <jsn at bjtpartners.com>
>
>> Hi Andrey,
>>
>> Would you have some updates about your NF calculations in both 900
>> (MGA-13116) and 1800 (MGA-13216) bands ?
>>
>> This information would be very interesting to figure out if we need an
>> external 1st stage LNA in front of the UmSEL Rx path.
>> Thanks a lot for your kind help regarding this front-end related question.
>>
>> Best regards.
>>
>> Jean-Samuel.
>> :-)
>>
>>
>>
>> On Tue, Oct 30, 2012 at 5:02 PM, Andrey Sviyazov <andreysviyaz at gmail.com>wrote:
>>
>>> HI Jean-Samuel.
>>>
>>> Not really sure to do UmTRXv3 right now.
>>> On my opinion now we should found as much issues as possible with v2.
>>> For example, now I am working with power consumptions.
>>>
>>> I have to think a bit about NF improvement you suggest.
>>> I'll let you know my calculations by tommorow.
>>>
>>> Best regards,
>>> Andrey Sviyazov.
>>>
>>>
>>>
>>> 2012/10/30 Alexander Chemeris <alexander.chemeris at gmail.com>
>>>
>>>> Forwarding to the mailing list.
>>>>
>>>>
>>>> ---------- Forwarded message ----------
>>>> From: Jean-Samuel Najnudel - BJT PARTNERS SARL <jsn at bjtpartners.com>
>>>> Date: Tue, Oct 30, 2012 at 6:25 PM
>>>> Subject: LNA and Noise Figure improvement
>>>>
>>>>
>>>> Hi Andrey,
>>>>
>>>> I understand you plan to prepare the UmTRXv3.
>>>>
>>>> I would need to discuss about the Rx Noise Figure improvement.
>>>> With UmSEL, according to your calculations, NF would be around 0.7 dB
>>>> (excluding duplexer IL) with the MGA-13116.
>>>> It would be interesting to try to improve this figure with a very
>>>> first stage LNA based on the MGA-633P8. This would let us reach a NF
>>>> as low as about 0.4 dB.
>>>>
>>>> Adding this very first stage LNA in the Rx path brings a few questions.
>>>>
>>>> 1/ To improve isolation with the other RF parts, would we need to have
>>>> this LNA in a separate shielded enclosure ?
>>>> If yes, we would need a 5V power supply for this LNA. To have a better
>>>> shielding between the voltage regulator circuitery and the LNA, it
>>>> should be nice to have the 5V voltage regulator outside of the LNA
>>>> enclosure. To make this possible, it would be very useful to add a 5V
>>>> voltage regulator and connector on the UmTRXv3.
>>>> If not, we do not need specific modifications on the UmTRXv3 but we
>>>> would need to add an MGA-633P8/634P8 as a very first stage in the
>>>> UmSEL.
>>>>
>>>> 2/ Whatever external or on the UmSEL, this extra LNA will increase the
>>>> total gain. Considering your calculation, the UmSEL IIP3 is already as
>>>> low as around 1 dBm. If we add a very first stage LNA, this IIP3 will
>>>> go down to about -15 dBm. This could be a problem for inband blockers.
>>>> Even if, in real life situation in the field, I am not sure it will be
>>>> a very big problem, it will be a problem to pass the spec.
>>>> Would you have another idea to improve the NF ?
>>>> If not, could you please let me know what NF we could expect with the
>>>> current design (UmSEL with MGA-13116/13216) in both 900 and 1800 bands
>>>> ?
>>>>
>>>> Best regards.
>>>>
>>>> Jean-Samuel.
>>>> :-)
>>>>
>>>>
>>>>
>>>>
>>>> --
>>>> Regards,
>>>> Alexander Chemeris.
>>>> CEO, Fairwaves LLC / ООО УмРадио
>>>> http://fairwaves.ru
>>>>
>>>>
>>>
>>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.osmocom.org/pipermail/umtrx/attachments/20121102/53570c84/attachment.html>


More information about the UmTRX mailing list