Showing results for 
Search instead for 
Did you mean: 

SOGEA Broadband speed highly variable - is there a wiring issue


We have a building which has a 100 pair copper cable from the street cabinet  which is 300m from the DP point in a riser cabinet.  This is jumpered across krone strips to a 100 pair which runs up the riser to the comms room on the 3rd floor.  From here, there are 3 x 50 pair copper cables down to each floor (ground, 1st & 2nd) 

We have 20 DEL lines working fine.

We then have 7 x SOGEA Broadband lines which follow the same cabling route to the 3rd floor comms room and then get distributed to the floors below.

When installed, and subsequently, the Openreach testing shows that these lines should all deliver 80/20.  However, when we add a BT router to the lines we start to get vary variable results.   The best we have got is 77/19 which ties in with the line tests that Openreach tests indicate.  However we have seen significant drop off with some lines dropping to 5/2.  We have also seen 5/19 which is first for all involved as we have not seen higher up speed than down before.

To try and resolve the issue we jumpered in master sockets for all 7 Broadband lines in the 3rd floor comms room and while it is not the BT DP, it is the closest workable are.  Most tests provide 74/19 on the routers here.  When we move down a floor, we drop to 50/19.  This seems quite a drop off for perhaps a max of an additional 10m.  Also, the down speed drops off but the up speed has stayed the same. Go to the ground floor and we get lower speed, down to 7/11.  Again odd to see the up speed being higher than the down.  Leave it a while and the speed changes to 5/2.

One engineer suggested that one of the wires used for jumpering across the krone strips looked a little thin - but we have replaced it with no change.

This has led us to question whether we have an issue with cabling or if we are doing something wrong with having a master socked jumpered in parallel  with the other connections.

I have posted a diagram below which shows the setup and the point where we switch from the twisted pairs  to the structured cabling.  Have we done something wrong and is there something about the SOGEA broadband that is different that is causing this.  

We have tested all sorts of scenarios but can't seem to avoid this significant drop off.  We have also on one occasion punched in a master socket to the DP point where the wiring enters the building thus removing all other remaining wiring and got 18/9.  

Is it possible to have too thick a cable that might cause this (the link from the floor coms room to the patch panel) or is having a RJ45 to BT socket adaptor on the same line in parallel to a BT master socket causing an issue?  I have an Openreach engineer and another telephone engineer scratching their heads on this one.  The only consistency is that the line test from the street cabinet to the building always tests 80/20.

BT to router connectivity diagram.jpg