Selectivity improvement solutions proposal for UmTRX

Jean-Samuel Najnudel - BJT PARTNERS SARL jsn at bjtpartners.com
Fri Mar 16 00:17:58 UTC 2012


Hi Andrey,

Thank you very much for your reply.

Thanks a lot for all the details you send.
Yes, you are all right, Hittite mixers looks much more suitable than the
ADL5801. These Hittite mixers have also a very high IIP3 and a resonable
Noise Figure. Moreover, I agree with you, the ADL5801 integrated IF amp is
not really suitable in our design.

Again, I agree with you. Receiving 2 carriers simultaneously on each Rx
path will not be easy (LMS phase noise, ADC dynamic range, extra digital
filtering...). We should keep just a single carrier per Rx path. This will
allow to do switched diversity. This is already great.

Regarding the choice between HMC483MS8GE/HMC485MS8GE and
HMC686LP4E/HMC687LP4E, I cannot figure out which one would be the most
suitable for our design. For the filter, it is also quite difficult for me
to help you to decide between the TB0130A and the TB0448A.
By the way, do you think this would be fine to use a different mixer for
900 and 1800 MHz bands ? As chip packages remain the same, I think it would
be fine but I would like to double check with you.

Regarding the GSM spec, I believe these blocker tests are hard to pass and
not that useful in most practical situations. However, I do not care that
much about passing this spec for my network deployment in Mayotte but I
really believe passing the spec will be very important for you if you wish
to sell your hardware solution to some major operators. Moreover, as we
would anyway need a superheterodyne selective filtering to get a reasonably
narrow Rx sampling band (< 1.5 MHz LMS band), it does not cost that much
more to try to pass the GSM spec.

By the way, I really believe this does not worth designing 2 different
preselectors. I think it would be more reasonable to design only the
Hittite based high performance preselector. First of all, as you mentioned
and as I calculate on my side also, the ADRF6601 will not even pass the
micro BTS spec. Moreover, the time you spend designing and debugging 2
different preselector would probably cost more than the 40 USD per board
you save on the boards you will use for low cost applications. Moreover, 2
different designs would mean lower quantities components order. This would
mean higher buying prices, higher shipping costs per chip... In practice,
the cost difference would be probably 25 to 30 USD per board, not 40 USD.

I really think we should only make one design with the Hittite high
performance preselector. If we really want a low cost femtocell, a
universal SDR board or a lab experiment system, we would not even populate
the preselector components. If we need GSM deployment applications (except
femtocells), nobody (even me ;-)) would actually care about the extra 40
USD BOM.

I looked at the picture of your Rx path calculations. As discussed together
in Barcelona, I am not sure we should add an LNA on the UmTRX. TMB LNA
would already have a very good Noise Figure (about 0.8 dB or even a bit
less) and a quite high gain (between 20 to 30 dB depending of the cable
loss). OIP3 of the TMB would be typacally around 35 dBm. Adding an extra on
board LNA would be quite bad for the total Rx path IIP3.
Moreover, as there is some PCB leakage from Tx to Rx in the UmTRX, we might
avoid onboard amplifiers. An on board amplifier might actuallly amplify the
leakage.
I actually suggested this LNA a few months ago but, in Barcelona, you
convinced me this LNA was not really a good idea. What do you think ?

Regarding the PA parameters, for both models we can get at a good price,
P1dB would be between 8 (Infineon) to 15 (Freescale) Watts at 27 VDC (a bit
lower at 24 VDC). Gain would be around 40 dB (excluding cable loss). Please
let me know if you need others LNA and/or PA specs of the TMB I can get
from the manufacturers I am discussing with.

Thank you very much for your great work on the design of the UmTRX. I
really believe this will be a great board.

Best regards.

Jean-Samuel.
:-)


2012/3/15 Andrey Sviyazov <andreysviyaz at gmail.com>

> Hi Jean-Samuel!
>
> About  ADL5802, you can't use it as Down and Up converter simultaneously.
> Also I'm afraid that we can't yet receiving 2 GSM carriers simultaneously.
> So lets done single channel first but 2 or more keep in mind for
> experiments.
>
> I try to find better components too.
> ADI mixers have internal IF AMP and it is not good in this case.
> Components of Hittite are known for me as a very good through my main job.
> I'd like to recommend to use any of the next:
> for GSM900
> 1/ HMC830LP6GE + HMC483MS8GE + TB0130A.  (20.6+4.69+3=$28.29)
> 2/ HMC830LP6GE + HMC686LP4E + TB0448A. (20.6+9.67+3=$32.97)
> for DCS1800
> 1/ HMC830LP6GE + HMC485MS8GE + TB0130A .  (20.6+4.69+3=$28.29)
> 2/ HMC830LP6GE + HMC687LP4E + TB0448A. (20.6+9.67+3=$32.97)
> By the way, LMS lower limit 0.3GHz as per datasheet.
>
> About GSM 05.05 specs I still can't understand blocker requirements:
> MS spectrum with RBW=200kHz have -65dBc level at 600-1200kHz offset,
> therefore blocker MS with -26dBm will be jammer for wanted MS with level
> less then -91dBm and noise level of receiver's heterodyne isn't matter in
> this case.
> I really can't understand why CW levels -26dBm and -16dBm blocking tests
> required.
> May be it just universal test of heterodyne quality?
>
> I think we should be reasonable people, and therefore we should use
> parameters which really necessary for us.
> So, ADRF6601 parameters seems to be quite enough even if it pass only mBS
> requirements.
> On the other hand, BOM difference between ADRF and HMC's around $20-25
> only and it isn't much for normal BS.
> In short, it seems that we should do three options front-end mezzanines:
> 1/ without channel preselector for picoBS or nanoBS upto 0.5W/ch.
> 2/ preselector ADRF based for microBS upto 2W/ch.
> 3/ preselector HMC based for normal BS with 10-20W/ch TMB.
>
> Please find attached pictures that my simple calculations for RX chain.
> Also, please let me know real PA parameters which you decide to use
> for TMB.
>
> Best regards,
> Andrey Sviyazov.
>
>
>
> 15 марта 2012 г. 4:42 пользователь Jean-Samuel Najnudel - BJT PARTNERS
> SARL <jsn at bjtpartners.com> написал:
>
> Hi Andrey,
>>
>> Thank you for your e-mail.
>>
>> Yes, you are all right.
>>
>> It think your calculations are good. I just did not knew how hard was the
>> GSM macro BTS spec.
>>
>> I browsed a bit the web and found the GSM 05.05 specs which confirm what
>> you say. I also found some similar information and calculations in an
>> academic paper which confirm your figures.
>> http://www.uta.edu/rfmems/Conferences/2001_SPIE_MicroMEMS/4592-20.pdf(page 5 and 6)
>>
>> Anyway, I tried to look at other components than the ADRF6601.
>>
>> I found a quite low phase noise VCO/PLL from Hittite which seems to be
>> able to let us probably pass the macro BTS spec or at least the micro BTS
>> spec.
>> http://www.hittite.com/products/view.html/view/HMC830LP6GE
>>
>> For the mixer, we may use a separate component like the ADL5801.
>>
>> Please let me know what you think about these chips. Please do not
>> hesitate to let us know some other suggestions if you know or if you can
>> find some other components that would have better performances.
>>
>> Actually, even if the specs are not easy to pass, I still feel quite
>> optimistic as it was possible to pass these specs 15 years old components.
>> Anyway, if the macro BTS specs are really too hard to pass, we may focus on
>> the micro BTS spec. This would already be great to convince the market you
>> may be interested in and the performances would be good enough for most
>> practical situations in my deployment in Mayotte.
>>
>> Best regards.
>>
>> Jean-Samuel.
>> :-)
>>
>>
>>
>> 2012/3/14 Andrey Sviyazov <andreysviyaz at gmail.com>
>>
>>> Hi all!
>>>
>>> I am again about far-near problem.
>>>
>>> If we have heterodyne noise -135dBc/Hz at 600kHz offset (ADRF6601), then
>>> for blocking signal at the same offset and at 200kHz RBW we get additional
>>> noise level 135-53=72dBc relative to blocking signal level.
>>> To keep "normal GSM900 BS" sensitivity -104dBm we must keep additional
>>> noise as low as -107dBm, therefore blocking signal maximum level must less
>>> then -107+73=-39dBm.
>>> But in GSM-05.05 (sec 5.1) I saw blocking characteristics requirements
>>> for normal BTS must be -26 dBm at 0.6-0.8 MHz offset and -16 dBm at 0.8-1.6
>>> MHz offset.
>>> So, I do not know how and who can meet those requirements and I am
>>> really hope that there are fundamental mistakes in my calculations.
>>> Correct me please.
>>>
>>> Best regards,
>>> Andrey Sviyazov.
>>>
>>>
>>>
>>> 16 января 2012 г. 19:22 пользователь Jean-Samuel Najnudel - BJT PARTNERS
>>> SARL <jsn at bjtpartners.com> написал:
>>>
>>> Hi Alexander,
>>>>
>>>> These last days, I tried to find a solution for the selectivity
>>>> improvement.
>>>>
>>>> I have 5 solutions to propose. 1st and 2nd are inboard solutions. 3rd,
>>>> 4th and 5th uses an external board. Some seems to be much better than
>>>> others.
>>>>
>>>> 1/ We could use an IF frequency above 375 MHz to be able to connect the
>>>> IF signal dirtectly to the LMS, without any upconvertion back to RF
>>>> frequency. This would save some components.
>>>> We could use the ADRF6601 (PLL/VCO + mixer) and the TB0448A IF SAW
>>>> filter.
>>>> The ADRF6601 is single chip PLL/VCO and mixer. This would be quite
>>>> convenient.
>>>> The TB0448A is cheap (< 3 USD), narrow band (good selectivity) and 400
>>>> MHz center frequency (> 375 MHz LMS lower limit).
>>>>
>>>> Cost of this solution would be about 60 USD and selectivity would be
>>>> really good.
>>>> The main disadvantage of this solution is the filter would restrict the
>>>> signal to a single GSM carrier. This would avoid us to get both GSM
>>>> carriers on each LMS. We would not be able to get true diversity. We would
>>>> only be able to get switched diversity.
>>>>
>>>> After the LNA, RF SAW filter and the RF switches, we can split the
>>>> signal between the current RX path to LMS RX LNA 3 and a new alternate RX
>>>> path (ADRF6601 => TB0448A => LMS RX LNA 1).
>>>> Depending of our need for selectivity, we would be able to select 1 of
>>>> these 2 RX path (direct RX path to LMS RX LNA 3 or IF filter RX path to LMS
>>>> RX LNA1).
>>>>
>>>> This would allow to use the board either as a normal wideband SDR board
>>>> or with a very selective filter.
>>>>
>>>>
>>>> 2/ A very nice option would be to use a variant of the 1st solution
>>>> with a wider bandwidth SAW IF filter. For example, if we use a 400 to 600
>>>> KHz bandwidth IF filter, we would also get a very good selectivity and we
>>>> would also be able to sample both GSM carriers on each LMS. This would
>>>> allow a good selectivity and full diversity.
>>>>
>>>> The problem is we would need a 400 to 600 KHz SAW IF filter, with good
>>>> selectivity, reasonable price and an IF center frequency above 375 MHz. I
>>>> was not able to find such a filter.
>>>>
>>>>
>>>> 3/ As suggested a few days ago, we may use the external selectivity
>>>> improvement board design I sent you. Instead of the Triquint 856378 IF SAW
>>>> filter, we could use the TAISAW TB0448A narrow band filter. This TAISAW
>>>> filter is really much cheaper than the Triquint. This would save a lot of
>>>> budget. However, we would still need 4 mixer and 2 PLL/VCO for each LMS RX
>>>> path. This external board would cost approximately 100 USD (excluding PCB
>>>> and assembly). We would need 2 of these boards for each UmTRX board. This
>>>> would make 200 USD per UmTRX. Including PCB and assembly, toatl cost would
>>>> be around 300 USD. This is not compeltely unrealistic but it seems still
>>>> quite expensive.
>>>>
>>>>
>>>> 4/ Another solution would be to build a single carrier version of the
>>>> 3rd solution design. We would need only 1 RF path (PLL/VCO + mixer) with
>>>> only 1 narrow band filter per LMS RX path. This would not need any splitter
>>>> or combiner. Design would be quite simple and cost would be about 2 times
>>>> lower. However, as we will have only 1 carrier on each antenna, we would
>>>> not be able to get diversity at all.
>>>>
>>>> This solution would finally not have many advantages compared to 1st
>>>> solution. It would cost more and would not allow any kind of diversity.
>>>>
>>>>
>>>> 5/ Last solution would be to build an external diversity improvement
>>>> board, as 4th solution, but with a wider band IF SAW filter.
>>>>
>>>> We could use the following RF path:
>>>> LNA => RF SAW filter => mixer => IF SAW filter => mixer => RF SAW
>>>> filter.
>>>> Dual mixer could be the ADL5802 connected to the ADF4350 PLL/VCO.
>>>>
>>>> We could use the TB0218A IF SAW filter. This filter is quite affordable
>>>> (< 10 USD). Selectivity is good and bandwidth is wide enough to select 2
>>>> GSM carriers (separated by 400 KHz).
>>>>
>>>> Cost of such external diversity improvement board would be quite
>>>> reasonable.
>>>> This would be a very nice solution to select 2 GSM carriers. Connected
>>>> to the UmTRX, this selectivity improvement board would allow to get both
>>>> switched or true diversity.
>>>>
>>>> As TB0218A center frequency is 140 MHz, we would not be able to connect
>>>> directly the IF signal to the LMS. We would need to up convert the signal
>>>> back to the RF frequency.
>>>> As IF down converted signal is upconverted back to the original RF
>>>> frequency, it would be possible to use this selectivity improvement board
>>>> with any kind of existing OpenBTS (UmTRX, USRP, SSRP...) or OpenBSC
>>>> (Sysmocom BTS, IP.access nanoBTS...) hardware to improve the Rx
>>>> selectivity. This would offer a wider potential market than an inboard
>>>> solution.
>>>>
>>>>
>>>> Considering all these solution, I believe 1st and 5th solutions seems
>>>> to be the best choices. 2nd solution would also be really nice but I was
>>>> not able to find the appropriate IF SAW filter.
>>>> Please let me know your opinion regarding each of these two solutions.
>>>>
>>>> By the way, the TB0448A and TB0218A SAW filters looks really good but I
>>>> am not 100% sure the GSM carrier spectrum distortion due to the pass band
>>>> ripple of the SAW filter is acceptable.
>>>> Center part of the GSM carrier (f +/- 100 KHz) is fine but side parts
>>>> of the GSM carrier (bellow f - 100 KHz and above f + 100 KHz) may be cut a
>>>> bit by the SAW filter.
>>>>
>>>> Could you also please check the TB0448A and TB0218A datasheets to
>>>> double check if the usable bandwidth is wide enough ? Especially, do you
>>>> think cutting a bit the side parts of the GSM carrier may cause problem ?
>>>>
>>>> Anyway, please let me know your point of view regarding these
>>>> selectivity improvement solutions.
>>>>
>>>> Best regards.
>>>>
>>>> Jean-Samuel.
>>>> :-)
>>>>
>>>>
>>>
>>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.osmocom.org/pipermail/umtrx/attachments/20120316/ddbce852/attachment.html>


More information about the UmTRX mailing list