sciphone dream g2 mtk 6235 status? ;)

This is merely a historical archive of years 2008-2021, before the migration to mailman3.

A maintained and still updated list archive can be found at https://lists.osmocom.org/hyperkitty/list/baseband-devel@lists.osmocom.org/.

Mychaela Falconia mychaela.falconia at gmail.com
Tue Feb 6 00:28:25 UTC 2018


On the subject of RF test equipment for the testing and development of
GSM MS devices, chiming in here as a "professional hobbyist" active
manufacturer of such devices, my choice of equipment is Rohde&Schwarz
CMU200.  This model has been the "gold standard" in the world of GSM
chipset vendors and mainstream phone/modem manufacturers: TI used the
CMU200 as their primary choice of instrument when validating their RF
chips (Clara, Rita, LoCosto and so forth) and their complete GSM MS
development boards with those chips, and they provided turnkey
software to their downstream customers (phone/modem manufs) so the
latter could do production line testing and RF calibration of their
end product devices using the very same R&S CMU200.

CMU200 units are readily available on ebay at very reasonable prices -
that's where I got mine.  Most of them (including the one I got) are
equipped with hardware options not only for GSM, but also for WCDMA
(UMTS/3G) - as I see it, the _*only*_ sensible reason to invest the
absolutely enormous amount of effort to run libre MS firmware on some
newer-than-Calypso chipset would be to get UMTS/3G support (expending
such Herculean effort to move from the Calypso to some newer chipset
that is still 2G-only just to please some gerontophones is not
justifiable in my book), hence I similarly see little point in
investing in RF test equipment that is not 3G-capable.

The CMU200 can be operated both standalone (using the front panel
display and controls) and under external host control, and for the
latter mode it supports both GPIB and RS-232.  I use the RS-232 option
for my automated RF test and calibration processes, as it conveniently
avoids the need for exotic interface adapter hardware and equally
exotic drivers and libraries.  The documentation is very good; I got
it in the form of an original R&S documentation CD, but if someone
else needs it, I can upload it to my FTP site.

If you go the ebay-sourced CMU200 route, one issue to watch out for is
units with subtle defects which are not immediately obvious.  My unit
came with a defective main RF generator: the RF analyzer functions
worked fine (analyzing the signal put out by the DUT), but when I
tried to make the CMU200 generate its own signal, the output was so
severely attenuated that it was almost undetectable.  The Aux Tx
generator worked, and I used it as a hacky workaround before I fixed
the main one.  To fix the main signal generator, I first isolated the
fault to the Rx/Tx board inside: I checked the signal at the internal
interface from the Rx/Tx board to the front end module, and found it
to be just as badly attenuated as the external output, pointing the
blame to the Tx half of the Rx/Tx board.  Then I found a replacement
Rx/Tx board on ebay, bought it, swapped it in, and the main signal
generator started working.

Right now my CMU200 is at R&S's service center in Maryland where I
sent it for recalibration - one of my requirements is to measure the
Tx power levels put out by my FreeCalypso GSM MS devices as precisely
as possible (for production calibration), and to feed signals of a
precisely known power level to these devices for Calypso Rx path
calibration.  Given that I have swapped an internal module in the RF
path of my CMU200 in the process of repairing it, I now need to have
the instrument recalibrated in order to get that absolute highest
power level accuracy I desire, which is why I sent it to R&S.  The
cost of this calibration service is more than the cost of the used
unit itself, but that's life.

Professional hobbyist regards,
Mychaela



More information about the baseband-devel mailing list