Troubleshooting BER- The basics.
BER Troubleshooting? This comes up A LOT. Troubleshooting BER has a bit of mystic magic about it. Its a lot of numbers in an odd scientific notation that baffles a lot of techs. Especially the new guys. First off lets start with breaking down the measurement a BER test spits out. 1.00-E-09
What does that mean? Ask any tech and his answer will be the same.
“If its below 1.00-E09 its bad and I have to find whatever is causing it to be bad…”
“What do the numbers mean though?”
Well, here they are-
X amount or errors out of X amount of data sent . That is all.
|1.00E-01||1/10||One in Ten|
|1.00E-02||1/100||One in One Hundred|
|1.00E-03||1/1,000||One in One Thousand|
|1.00E-04||1/10,000||One in Ten Thousand|
|1.00E-05||1/100,000||One in One Hundred Thousand|
|1.00E-06||1/1,000,000||One in One Million|
|1.00E-07||1/10,000,000||One in Ten Million|
|1.00E-08||1/100,000,000||One in One Hundred Million|
|1.00E-09||1/1,000,000,000||One in One Billion|
|1.00E-010||1/10,000,000,000||One in Ten Billion|
|1.00E-011||1/100,000,000,000||One in One Hundred Billion|
|1.00E-012||1/1,000,000,000,000||One in One Trillion|
|1.00E-00||0X1||Zero (no errors)|
When you are on a service call for tiling or any issue really and the QAM channel you are testing shows a BER value of 5.00E-05 it means simply this- There are 5 errors for every one hundred thousand bits or packets sent. That doesnt sound like much at all. But if you take into account how many packets are sent per second( 5mil) you would see this is a HUGE amount of errors.
The key to running a good solid BER test is patience and immobility. Trilithic, the makers of the 860DSPi that most techs have these days, suggest to keep your meter completely still when running a BER test. Also, since you are measuring over time its suggested to run the test or “sit on the channel” with your meter for at least a minute. You will not get an accurate test or measurement if you just hit your QAM button and channel up as soon as the carrier locks on. Remember, BER is a measurement of errors over time. So give it some.
Repairing BER however is much easier then extrapolating the scientific notation posted above. Its still cable right? So what causes poor BER? A bunch of stuff.
There are more involved issues that can cause poor BER. One possibility is a poorly balanced run. It has been witnessed that a run just off by 2dB or so can throw errors into your plant. More towards the end of the cascade as well. If your a service tech and have BER issues at the tap then you should be referring it to maintenance. Most MSOs agree that a BER tolerance of 1.00E-08 is acceptable. The differnence however might be if we are reading Pre or Post errors. ( more on that in a second.)
Going back up to the top of the article, when taking a measurement take care for accuracy. It isnt entirely uncommon for a tech to not find an issue at the home, BER or otherwise, while the customer swears it tiles all the time. The tech heads to the tap, climbs up the pole
and starts a BER test. As he tests he bangs on the tap a few times assuming there is something loose or damaged that he can coax out of hiding and low and behold! BER errors!! ( of course there are you rookie!) Please refrain from doing this. Any disturbance can cause a few packets to be thrown asunder. This is normal. A little pop of a correctables or uncorrectables generally is not going to be your issue. About the most I can advocate at the tap would be reading your QAM and maybe swaying the feeder a bit. Not shaking or rattling, but sway it a bit as it would sway in the wind. This will turn up radial cracks and loose connectors and will show as drops of MER or loss of Low End signal. That can be indicative of an intermittent issue on the plant. For the love of god do not shake or bang on the taps themselves.
In the end tracking BER issues is the same as tracking any issue. Find where its bad, track back to where it is good and troubleshoot in-between. High levels, Low levels, noise, loose connectors. You name it, they cause it. But at least now hopefully you understand what BER is, how to read it and how it is measured.
BER is a measurement of errors over time. So you need a little time to build up some packets so your meter can take an accurate sample. You shouldnt need more then a minute of testing to get an accurate QAM reading. Per spec 33 secs should get you a pretty accurate BER reading.
This takes us into Home Certification. It is not uncommon for a QAM to fail a BER test during a home cert depending on how many channels are set up to be checked for BER. The Cert test does not sit on every channel for 30 secs to 1 min. Therefor its testing threshold is a bit lower. Less bits tested divided by any number of errors will show a BER issue. Then you sit on the channel after the test and see no issue, so you test again. Frustrating I know. This is most noticeable on Docsis channels being tested. The meter is testing for Upstream, Downstream, MER, BER, etc. These are a lot of calculations in a short time. ( although it feels like forever when you are at the tap) Be aware of this when tracking your issue. Do not be thrown down the wrong path.
Do NOT use your Home Cert to troubleshoot. The Home Cert test is simply a verification AFTER your issue is resolved. Running around performing home certs to troubleshoot means we are skipping channels that need to be tested, not spending time studying a channel and possibly not listening to the customer in regards to what their issue really is. How many times have we set up a line call for high upstream when we were there for tiling but found no issue? We hope the line tech will come across something else. And we often STOP troubleshooting when any line issue is identified.
This can take us into many other avenues that Ill try and jot down at a later date. I hope this clears up BER reading and tracking.
If you would like to discuss this further, ask some questions or offer some insights please comment on this article or, even better, join the forum and continue the current discussions regarding BER and troubleshooting.