E1/T1 frequent alarms/Clock sync issues in WCDMA <<Tx guys>>

[vaampeere]

Dear ppl,

We are facing a problem of clock sync in out network.Usually a clock is purchased for a TX link for 2G and 3G, but here the clock is being provided by 2G site, which in turn is being provided by an external link from a company. (such things in a new network !)

Now, I have noticed that when the clock sync alarms come on site, the O & M people go on to the site , and re punch the E1/T1's, which immediately clears the alarms.
My question is:

1) what are the immediate practical outcomes of an E1/T1 active alarm Or clock sync alarm?

2) What options are there for removing it, apart from purchasing separate GPS clocks?

3) Why on re-punching the E1/T1 alarms disappear?

Thx.

[bawbag11]

1) what are the immediate practical outcomes of an E1/T1 active alarm Or clock sync alarm? Audio distortion and lost signalling possibly leading to dropped calls / SHO failures and call setup failures especially if you use UM mode.

2) What options are there for removing it, apart from purchasing separate GPS clocks? Fix the underlying transmission issue that is causing the problem.

3) Why on re-punching the E1/T1 alarms disappear? Are field removing the wires then re-punching them? in this case the LOS would re-start the clock DAC.

You don't say if you are using IMA groups and the physical interface between TX and NodeB is 75ohm coax or 120 twisted pair.

In short, do loopbacks to check for transmission disturbances. If the E1 goes over IP/ATM transmission then ensure there is no contention happening in the TX network. There is also some talk about grounding in the Hedex help regarding these alarms.

You can also monitor the clock DAC and see the drift, if there are any sharp deviations then I would definitely suspect bad transmission.

Regards

BB

[vaampeere]

Bawbag,

Thank you for you detailed reply.

I'm not sure about the IMA groups, but the connectivity between NodeB and Tx is twisted pair. Network overall is IP based.

Voice distortion is there with these alarms and yet there is another instance of call with bad levels (indoors especially), call distortion and *****ing , while still not being handed over to GSM, even when the levels are beyond the thresholds defined for CS/PS IRAT handovers.If it goes to 2G, then the call is clear...but generally it gets stuck with worst levels on 3G indoors.

From the Tx dept., the db levels are within good Rx levels On U2000 on the Tx interfaces, and the issue exists only with the Clock sync.These alarms are also a bit confusing as they appear whenever a site power goes down and NodeB comes back up again.

Thanks for clearing up the issue and I hope with clock license purchasing (for 2G and 3G) separately the issue can be disappear.

BR.

1) what are the immediate practical outcomes of an E1/T1 active alarm Or clock sync alarm? Audio distortion and lost signalling possibly leading to dropped calls / SHO failures and call setup failures especially if you use UM mode.

2) What options are there for removing it, apart from purchasing separate GPS clocks? Fix the underlying transmission issue that is causing the problem.

3) Why on re-punching the E1/T1 alarms disappear? Are field removing the wires then re-punching them? in this case the LOS would re-start the clock DAC.

You don't say if you are using IMA groups and the physical interface between TX and NodeB is 75ohm coax or 120 twisted pair.

In short, do loopbacks to check for transmission disturbances. If the E1 goes over IP/ATM transmission then ensure there is no contention happening in the TX network. There is also some talk about grounding in the Hedex help regarding these alarms.

You can also monitor the clock DAC and see the drift, if there are any sharp deviations then I would definitely suspect bad transmission.

Regards

BB

[bawbag11]

Hi,

I would definitely do a loopback test to double check the TX circuit. If TX Network is IP you are using pseudowire and too many times the TX people avoid setting the correct CIR rate end to end which causes the exact problem you describe.

Good luck

BB

[vaampeere]

Hello Bawbag,

We just got a word from Huawei TAC that our clock settings for BSC-->2G (DTMU)-->WMPT (source of clock feed) are based as CLKMODE (Line) while the definition should be 'PEER' for this.However when done with the config, after adding the PEER MODE, the PEER line is still showing as unavailable?.Any experience on this scenario?.

Thx.

[Lav Anders]

Clock Working mode=Manual
Selected Clock Source=Peer Clock
Clock Source No.=0,1,2,3 or 4...etc

Make sure that the E1 in 2G have transmit and receive. 2G E1 transmit<-->3G E1 receive and 2G E1 receive<-->3G E1 transmit.

Good Luck

[vaampeere]

Ok got another word from TAC..and based on our BTS for 2G as BTS3012 and BTS3900 for only UMTS, and also the line coming from BTS of 2G, they are retracting the statement of PEER saying that it should be LINE as before.However, they are also of the opinion that they have not seen such configurations before, as usually UMTS should not be serve by 2G BTS, but its own LINE source from 2G.They are saying further, that if such a configuration is present, then the topology should also have a very good ring network in case no cascade effect takes place.