Thanks Alex. Do you know if this GPSDO would work: https://www.ebay.com/itm/GPS-Receiver-GPSDO-10MHz-1PPS-GPS-Disciplined-Clock-SINE-WAVE-GPS-RECEIVER/262981677836?hash=item3d3aedf30c:g:8HUAAOSwxu5ZD~MH
Feature: POWER:DC12V 1.5A GPS ANT POWER:DC3.3V 1PPS OUTPUT:Square WAVE,3.3Vpp 10M OUTPUT :SINE WAVE,1Vrms (13dBm+-2dB) RS232 OUTPUT:GPS NMEA SIGNAL SIZE:W*H*D=107*55*120mm ACCESSORY:AC110-220-DC12V ADAPTER,GPS ANT. Thanks! Rafael Diniz On 01/15/2018 11:46 AM, Alexander Huemer wrote: > Hi! > > On Mon, Jan 15, 2018 at 10:07:50AM -0200, Rafael Diniz wrote: > >> I'm doing some tests with an old USRP2 + RF900 daughterboard, and I get >> a very unstable system (with osmo-trx, osmo-bts-trx and osmo-nitb) - >> some phones can connect, some dont. I'd like to identify the problem - >> is it a bad GBit ethernet or the lack of a GPSDO connected to the 10MHz >> port? > > According to [1] the clock accuracy of the USRP2 is 20ppm. > GSM 05.10[2] states that accuracy has to be 0.05ppm or better. > That's quite a bit off, so I don't think it's surprising that you don't > see much stability. > I believe it makes a difference whether the MSs see other BTSs or not. > If the MSs see only the unstable BTS it is more likely to work out. > Not sure what is the case here. > >> And after some time, osmo-trx stops reading samples from UHD and >> osmo-bts-trx shuts down. > > Not sure if that is a possible effect of an unstable clock. I could > imagine so, but again, I am not sure. > > I guess the obvious advice is to attach a more stable clock and try > again. > > Kind regards, > -Alex > > [1] http://openbts.org/w/index.php?title=Ettus_Research_USRP > [2] > http://www.etsi.org/deliver/etsi_gts/05/0510/05.00.00_60/gsmts_0510v050000p.pdf >
signature.asc
Description: OpenPGP digital signature
