BSR64K corrected/uncorrectable codewords | docsis.org

You are here

BSR64K corrected/uncorrectable codewords

4 posts / 0 new
Last post
bama2545
BSR64K corrected/uncorrectable codewords
AttachmentSize
Image icon test modem.jpg63.34 KB

We got a BSR64k running in docsis 1.1. We have hooked a modem up directly to the back of it and it is running corrected codewords when streaming data over the internet with an occasional uncorrectable. Is this normal? We are thinking that this should be clean and not running any corrected data. We have tried different modems and slots as well (sb5101, sb5120, and slots 11,12,13). Corrected codewords are about the same no matter what.

Software - 3.1.1.16P05.KRCU
CMTS 1x4 - RDN PPC 3.1.0.26
SBG900 - sbg900-2.1.9.7-scm00-nosh
16qam - 32mhz
256qam - 117mhz

Modem Levels
Rcv Power @ modem - 9
Transmit Power - 47
Downstream SNR - 42
Upstream SNR - 28

Attached is a graph showing this happening

kwesibrunee
Though it by cisco, the best

Though it by cisco, the best documentation on FEC (Forward Error Correction) and, signal quality in general, I have found is on Cisco's site Here
From the Article

The presence of uncorrectable FEC errors is a good measure when noise gets to an intolerable level or when packets are colliding with each other from bad timing or poor headend splitter or combiner isolation. With regard to the latter, a packet intended for one CMTS upstream port is “leaking” onto another upstream because of the poor isolation.

A large increase in uncorrectable FEC errors results in voice quality issues.

Correctable FEC errors are seen as the level of noise is increased. Correctable FEC errors do not result in packet drops or poor voice quality, as long as there are no accompanying uncorrectable FEC errors.

Increasing the FEC T-bytes in the US modulation profile may help up to a certain point, but it depends on the noise source. Seven to ten percent FEC coverage seems optimal.

bama2545
The only thing involved here

The only thing involved here is about 5 feet of coax and a splitter. We have a 1 in 2 out splitter but we are running through it backwards. Downstream 0 to output 1, Upstream 0 to output 2 and the cable modem hooked up to the input of the splitter. Could hooking it up this way be causing the corrected codewords? All I know is that it works this way and still just curious if it is common to get corrected codewords when hooked up directly to the back of the cmts with one modem on the entire card.

kwesibrunee
Corrected codewords are not

Corrected codewords are not necessarily bad and Uncorrectable codewords are only a problem if there are too many of them i.e. > 1-2% of total codewords. The cisco article I referenced earlier explains better than I ever could. (FEC is part of the Docsis standard, but the standard is a bit dry and hard to understand whereas the cisco Doc is relatively straightforward.).

Are you using a Diplex Filter or a regular splitter? Use a diplex filter if possible, makes all the difference in the world for Lab settings.

Another thing to look at is because your only going 5 ft, look at the timing offset for the modem, my guess is that it is set to something like 2200 may also be causing your issues. For a lab setting you could override it with a more appropriate setting i.e. the lowest allowed by CMTS since your going 5 ft. Or conversely you could connect a spool of coax i.e. 500ft to better mimic a real world scenario.

Log in or register to post comments