Hi,

I would appreciate some help regarding a sync. problem I've been having 
for few weeeks now. It's a Siemens SMA4/1 ADM, which is used to groom 
traffic from couple of mobile switches (GSM operator) and send it to 
another location via STM-1 lines leased from incumbent carrier. There 
are 4 STM-1 ports connected to a device, and some 35 E1 links. The sync 
problem arises on one STM-1 port only (which is connected to a Siemens 
mobile swithing center in the same location), I get a lot of PJE+ events 
every day, around 1400-1500. The ADM is being synchronized through an 
STM-1 port coming from incumbent carrier. The Siemens swith concerned is 
also being synchronize from the same carrier, only with 2 MHz 
connection. I tried to sync the ADM to a Siemens switch, but that was no 
good; there were no PJEs on that port, but there were a lot of PJEs on 
all the other ports. Also, device is configured to have 3 possible clock 
sources, and is free to select the best one itself, that's why the one 
from incumbent carrier is selected. Anyone has an idea how to adress 
this problem? 
Thanks in advance,

-- 
Robi

Hello Robi,

You asked:

> I would appreciate some help regarding a sync. problem I've been having 
> for few weeeks now. It's a Siemens SMA4/1 ADM, 

Hm, Siemens SMA is a popular system in this group  ;-)

> which is used to groom 
> traffic from couple of mobile switches (GSM operator) and send it to 
> another location via STM-1 lines leased from incumbent carrier. There 
> are 4 STM-1 ports connected to a device, and some 35 E1 links. The sync 
> problem arises on one STM-1 port only (which is connected to a Siemens 
> mobile swithing center in the same location), I get a lot of PJE+ events 
> every day, around 1400-1500. 

If I read this well, the four STM-1 are connected to four different
systems (probably for protection), right?

> The ADM is being synchronized through an 
> STM-1 port coming from incumbent carrier.

Yes, line-timing as I would expect.

> The Siemens swith concerned is 
> also being synchronize from the same carrier, only with 2 MHz 
> connection. 

Probably a BITS station clock.

> I tried to sync the ADM to a Siemens switch, but that was no 
> good; there were no PJEs on that port, but there were a lot of PJEs on 
> all the other ports. Also, device is configured to have 3 possible clock 
> sources, and is free to select the best one itself, that's why the one 
> from incumbent carrier is selected. 

What is the selection criterium? SSM or priority/availability?

> Anyone has an idea how to adress this problem? 

 From what you describe I would suspect the Siemens mob. sw. center
using the 2 Mbit reference clock.
If there are PJE+ only (and no PJE-) hen the system is not locked.
if there are PJE+ and PJE- and about the same, this is due to
wander.

> Thanks in advance,

Please let me know how you solved the issue.

Cheers, Huub.

-- 
                reply to hhelvooort with 2 'o's
================================================================
              http://members.chello.nl/hhelvoort/
================================================================
Always remember that you are unique...just like everyone else...

In article <4W84e.158$184.49@amstwist00>, hhelvooort@chello.nl says...
> Hello Robi,
> 
> You asked:
> 
> > I would appreciate some help regarding a sync. problem I've been having 
> > for few weeeks now. It's a Siemens SMA4/1 ADM, 
> 
> Hm, Siemens SMA is a popular system in this group  ;-)
> 
> > which is used to groom 
> > traffic from couple of mobile switches (GSM operator) and send it to 
> > another location via STM-1 lines leased from incumbent carrier. There 
> > are 4 STM-1 ports connected to a device, and some 35 E1 links. The sync 
> > problem arises on one STM-1 port only (which is connected to a Siemens 
> > mobile swithing center in the same location), I get a lot of PJE+ events 
> > every day, around 1400-1500. 
> 
> If I read this well, the four STM-1 are connected to four different
> systems (probably for protection), right?
Actualy, no. Two of them are connected as tributary ports and carry 
local traffic (like mob. sw. center and some signalling links which, in 
the end, connect to a gateway in another city), and the other two 
connect to inc. carrier's fibres and go to that other city. Notice that 
they do not work in protected mode, each of them carries a different 
traffic.
> 
> > The ADM is being synchronized through an 
> > STM-1 port coming from incumbent carrier.
> 
> Yes, line-timing as I would expect.
That is correct
> > The Siemens swith concerned is 
> > also being synchronize from the same carrier, only with 2 MHz 
> > connection. 
> 
> Probably a BITS station clock.
Just to make sure we talk obaout the same thing: it's a T3 signal used 
just for sync, it carries no traffic.
> 
> > I tried to sync the ADM to a Siemens switch, but that was no 
> > good; there were no PJEs on that port, but there were a lot of PJEs on 
> > all the other ports. Also, device is configured to have 3 possible clock 
> > sources, and is free to select the best one itself, that's why the one 
> > from incumbent carrier is selected. 
> 
> What is the selection criterium? SSM or priority/availability?
The first criterium is SSM, then, if there are more than one clock 
sources of the same SSM status (quality), the one with the highest 
priority is selected. Currently, the selected clock source is 
prioritised as 3rd, and it's SSM byte is read as PRC).
> 
> > Anyone has an idea how to adress this problem? 
> 
>  From what you describe I would suspect the Siemens mob. sw. center
> using the 2 Mbit reference clock.
Actualy it's a 2 MHz clock (no SSM byte in it)
> If there are PJE+ only (and no PJE-) hen the system is not locked.
That is the situation, there are only PJE+. What are you sugesting, 
holdover? Or even free running? As far as I know, free running is not 
allowed in SDH networks, and I doubt the mob. sw. center could handle 
all the traffic that it receives with no problems if that was the case 
(connection from the SMA ADM is not the only traffic that it handles). 
How can I check whether it is a holdover state? Diagnostic capabilities 
of a mob. sw. center are not very high, stating the reasons here would 
take too long..... 
> if there are PJE+ and PJE- and about the same, this is due to
> wander.
> 
> > Thanks in advance,
> 
> Please let me know how you solved the issue.

What about the amount of PJE+ (1400-1500), they don't interrupt the 
traffic (so far), but it's been going on for weeks now, when do I need 
to get worried?
Thanks a lot for your trouble,

-- 
Robi

In article <4W84e.158$184.49@amstwist00>, hhelvooort@chello.nl says...
> Hello Robi,
> 
> You asked:
> 
> > I would appreciate some help regarding a sync. problem I've been having 
> > for few weeeks now. It's a Siemens SMA4/1 ADM, 
> 
> Hm, Siemens SMA is a popular system in this group  ;-)
> 
> > which is used to groom 
> > traffic from couple of mobile switches (GSM operator) and send it to 
> > another location via STM-1 lines leased from incumbent carrier. There 
> > are 4 STM-1 ports connected to a device, and some 35 E1 links. The sync 
> > problem arises on one STM-1 port only (which is connected to a Siemens 
> > mobile swithing center in the same location), I get a lot of PJE+ events 
> > every day, around 1400-1500. 
> 
> If I read this well, the four STM-1 are connected to four different
> systems (probably for protection), right?
Actualy, no. Two of them are connected as tributary ports and carry 
local traffic (like mob. sw. center and some signalling links which, in 
the end, connect to a gateway in another city), and the other two 
connect to inc. carrier's fibres and go to that other city. Notice that 
they do not work in protected mode, each of them carries a different 
traffic.
> 
> > The ADM is being synchronized through an 
> > STM-1 port coming from incumbent carrier.
> 
> Yes, line-timing as I would expect.
That is correct
> > The Siemens swith concerned is 
> > also being synchronize from the same carrier, only with 2 MHz 
> > connection. 
> 
> Probably a BITS station clock.
Just to make sure we talk obaout the same thing: it's a T3 signal used 
just for sync, it carries no traffic.
> 
> > I tried to sync the ADM to a Siemens switch, but that was no 
> > good; there were no PJEs on that port, but there were a lot of PJEs on 
> > all the other ports. Also, device is configured to have 3 possible clock 
> > sources, and is free to select the best one itself, that's why the one 
> > from incumbent carrier is selected. 
> 
> What is the selection criterium? SSM or priority/availability?
The first criterium is SSM, then, if there are more than one clock 
sources of the same SSM status (quality), the one with the highest 
priority is selected. Currently, the selected clock source is 
prioritised as 3rd, and it's SSM byte is read as PRC).
> 
> > Anyone has an idea how to adress this problem? 
> 
>  From what you describe I would suspect the Siemens mob. sw. center
> using the 2 Mbit reference clock.
Actualy it's a 2 MHz clock (no SSM byte in it)
> If there are PJE+ only (and no PJE-) hen the system is not locked.
That is the situation, there are only PJE+. What are you sugesting, 
holdover? Or even free running? As far as I know, free running is not 
allowed in SDH networks, and I doubt the mob. sw. center could handle 
all the traffic that it receives with no problems if that was the case 
(connection from the SMA ADM is not the only traffic that it handles). 
How can I check whether it is a holdover state? Diagnostic capabilities 
of a mob. sw. center are not very high, stating the reasons here would 
take too long..... 
> if there are PJE+ and PJE- and about the same, this is due to
> wander.
> 
> > Thanks in advance,
> 
> Please let me know how you solved the issue.

What about the amount of PJE+ (1400-1500), they don't interrupt the 
traffic (so far), but it's been going on for weeks now, when do I need 
to get worried?
Thanks a lot for your trouble,

-- 
Robi

Hello Robi,

You wrote:

> Actualy, no. Two of them are connected as tributary ports and carry 
> local traffic (like mob. sw. center and some signalling links which, in 
> the end, connect to a gateway in another city), and the other two 
> connect to inc. carrier's fibres and go to that other city. Notice that 
> they do not work in protected mode, each of them carries a different 
> traffic.

OK.

>>>The Siemens swith concerned is 
>>>also being synchronize from the same carrier, only with 2 MHz 
>>>connection. 
>>
>>Probably a BITS station clock.
> 
> Just to make sure we talk obaout the same thing: it's a T3 signal used 
> just for sync, it carries no traffic.

Yes.

>>>I tried to sync the ADM to a Siemens switch, but that was no 
>>>good; there were no PJEs on that port, but there were a lot of PJEs on 
>>>all the other ports. Also, device is configured to have 3 possible clock 
>>>sources, and is free to select the best one itself, that's why the one 
>>>from incumbent carrier is selected. 
>>
>>What is the selection criterium? SSM or priority/availability?
> 
> The first criterium is SSM, then, if there are more than one clock 
> sources of the same SSM status (quality), the one with the highest 
> priority is selected. Currently, the selected clock source is 
> prioritised as 3rd, and it's SSM byte is read as PRC).

OK.

>>>Anyone has an idea how to adress this problem? 
>>
>> From what you describe I would suspect the Siemens mob. sw. center
>>using the 2 Mbit reference clock.
> 
> Actualy it's a 2 MHz clock (no SSM byte in it)

OK. That is what I assumed.

>>If there are PJE+ only (and no PJE-) hen the system is not locked.
> 
> That is the situation, there are only PJE+. What are you sugesting, 
> holdover? 

This seems to be the most likely, however according to the standards
holdover should not last indefinitely, but should be used to bridge
the MTTR (mean time to repair).
Note that a 1st generation SDH system in holdover still sends
SSM: "PRC".

> Or even free running? As far as I know, free running is not 
> allowed in SDH networks, and I doubt the mob. sw. center could handle 
> all the traffic that it receives with no problems if that was the case 
> (connection from the SMA ADM is not the only traffic that it handles). 
> How can I check whether it is a holdover state? Diagnostic capabilities 
> of a mob. sw. center are not very high, stating the reasons here would 
> take too long..... 

I would expect the system to report loss of sync input.
Maybe it has (on the timing unit?) an indicator "holdover"
or on the EMS sync status "holdover".

> What about the amount of PJE+ (1400-1500), they don't interrupt the 
> traffic (so far), but it's been going on for weeks now, when do I need 
> to get worried?

This is about 90 ppm offset which is large but if the buffer capacity
of your system is able to survive this, you're OK. One PJE is 24 bits,
if they happen too often you buffer may fill and overflow, causing
a re-sync. This may affect the quality of the transported speech.

Cheers, Huub.

-- 
                reply to hhelvooort with 2 'o's
================================================================
              http://members.chello.nl/hhelvoort/
================================================================
Always remember that you are unique...just like everyone else...

Robi wrote:
> In article <4W84e.158$184.49@amstwist00>, hhelvooort@chello.nl
says...
> > Hello Robi,
> >
> > You asked:
> >
> > > I would appreciate some help regarding a sync. problem I've been
having
> > > for few weeeks now. It's a Siemens SMA4/1 ADM, which is used to
groom
> > > traffic from couple of mobile switches (GSM operator) and send it
to
> > > another location via STM-1 lines leased from incumbent carrier.
There
> > > are 4 STM-1 ports connected to a device, and some 35 E1 links.
The sync
> > > problem arises on one STM-1 port only (which is connected to a
Siemens
> > > mobile swithing center in the same location), I get a lot of PJE+
events
> > > every day, around 1400-1500.
> > [...]
> > > The Siemens swith concerned is also being synchronize from the
> > > same carrier, only with 2 MHz connection.
> >
> > Probably a BITS station clock.
>
> Just to make sure we talk obaout the same thing: it's a T3 signal
used
> just for sync, it carries no traffic.

Howdy Robi,

Regardless of what rate the Siemens box is supposed to be using for
it's reference (you seem to mention three different ones: T3, STM-1,
and E1), it looks to me like the mob. sw. center is in freerun/holdover
and the rest of the equipment (including the ADM) is timed correctly.
At least, if I correctly understand what you're saying about the
configuration.  If the mob. sw. center is not timing properly, the ADM
would see PJE's - exactly as you do.

You could prove this to be true by timing the mob. sw. center to the
ADM, if it can do that.  The PJE's on the ADM should settle down.

[...]
> > If there are PJE+ only (and no PJE-) hen the system is not locked.
> That is the situation, there are only PJE+. What are you sugesting,
> holdover? Or even free running? As far as I know, free running is not

> allowed in SDH networks, and I doubt the mob. sw. center could handle

> all the traffic that it receives with no problems if that was the
case

See below...

> What about the amount of PJE+ (1400-1500), they don't interrupt the
> traffic (so far), but it's been going on for weeks now, when do I
need
> to get worried?

As Hubb alluded to, it depends on the type of traffic you're running
and how tolerant it is of jitter (which is what pointer adjustments
cause).

PJE's themselves do not hit traffic - it's how other equipment handles
those pointer adjustments that *may* cause hits.  But let's pretend for
a moment that you are taking hits, and that it is on voice channels.
Most people probably wouldn't bother complaining about it anymore, in
this age of horrible voice quality on cell phones.  I'm certainly not
saying that you shouldn't fix it - I'm just asking if it may be
interrupting traffic and you just don't know it?

Good luck,

   Marc