Why would my attenuation go up and down?

Date_Time,Sync_dn,SNR_dn,PWR_dn,Atten_dn,CRC_dn,HEC_dn,Sync_up,SNR_up,Pwr_up,Atten_up,CRC_up,HEC_up

2006/10/08,08:43:28,8128,6,19.0,24.0,587,321,448,24.0,11.0,10.5,0,0
2006/10/08,08:43:58,8128,5.5,19.0,25.0,587,321,448,24.0,11.0,10.5,0,0
2006/10/08,08:44:28,8128,4.5,19.0,25.0,592,325,448,24.0,11.0,10.5,0,1
2006/10/08,08:44:56,,,,,,,,,,,,
2006/10/08,08:45:28,6528,8.5,19.0,25.0,1,0,448,24.0,11.0,11.5,0,0
2006/10/08,08:45:55,,,,,,,,,,,,
2006/10/08,08:46:28,4160,7,19.0,27.0,0,0,448,14.0,10.5,13.5,0,0
2006/10/08,08:46:58,4160,7,19.0,28.0,146,93,448,14.0,11.0,13.5,3,5
2006/10/08,08:47:26,,,,,,,,,,,,
2006/10/08,08:47:58,7232,0.5,19.0,24.0,173,141,448,21.0,11.0,11.0,1,3
2006/10/08,08:48:26,,,,,,,,,,,,
2006/10/08,08:48:56,,,,,,,,,,,,
2006/10/08,08:49:28,2944,10,19.0,27.0,2,2,448,11.0,11.0,13.0,3,2
2006/10/08,08:49:56,,,,,,,,,,,,
2006/10/08,08:50:28,6688,7,19.0,24.0,0,0,448,22.0,11.0,11.0,0,0
2006/10/08,08:50:58,6688,7,19.0,24.0,0,0,448,22.0,11.0,11.0,0,0
2006/10/08,08:51:28,6688,7,19.0,24.0,0,0,448,22.0,11.0,11.0,0,0

Ok looking at the above data my attenuation down seems to vary from 24db to 28db and loses sync quite a few times. Any ideas what might cause this apparent variation in attenuation. Could noise affect the attenuation or are we looking at a line fault ?

Oh oh oh and why would it vary between 9am and 5:30pm on a Sunday Any theories ?

Edited by Banger (Mon 09-Oct-06 01:02:20)

I may have answered my own question here but I can now hear a crackle on my telephone line would this affect attenuation ?

Would I contact BT about the voice side ? Is the same procedure to check for the crackle from the test socket ?

Yes, always check the line at the test socket. Dial 17070 and select "Quiet Line Test". That won't time-out, and gives you a totally quiet line to check for noises. Report it to BT (or whoever provides the line). I wouldn't mention ADSL too much. By the way, do the test with the ADSL modem off - ADSL can sometimes put clicks and crackles on the line itself. A variation of a couple of dB's on a line isn't uncommon - it's actually a variation in testing accuracy.

Edited by deleted (Mon 09-Oct-06 09:14:09)