Thread Links Date Links
Thread Prev Thread Next Thread Index Date Prev Date Next Date Index

Re: [802.3ae] RE: [802.3ae_Serial] 802.3ae PRBSs are upside down

For every bit-based or pattern-based BERT, there is always a choice to use
inverted data independent of the type of pattern being sought.  Some BERTs
do this automatically, others must be explicitly told to use "inverted"
data.  The bit-based BERTS invert the incoming data, then, before using it
as a seed to the internal reference generator.  They also invert the
incoming data while then doing comparisons to the internal reference
generator's output to make bit error comparisons.  The pattern-based BERTS
do a similar thing.  They load their pattern RAMs from the pattern off the
disk (in it's true form, for example) and then invert the incoming data (on
command) to compare against it.  It does not invert the reference pattern in
the pattern RAM.

note, we mention bit-based BERTS--in fact, BERTs don't work bit-based
either.  The PRBS reference generators are all done in parallel where the
polynomials have been pre-raised by some power of 2 (e.g. 32, 64 or 128
depending on BERT architecture).

Tom Waschura
SyntheSys Research, Inc.

----- Original Message -----
From: "Tim Warland" <twarland@xxxxxxxxxxxxxxxxxx>
To: <pat_thaler@xxxxxxxxxxx>
Cc: <stds-802-3-hssg@xxxxxxxxxxxxxxxxxx>
Sent: Wednesday, March 06, 2002 7:21 AM
Subject: Re: [802.3ae] RE: [802.3ae_Serial] 802.3ae PRBSs are upside down

> pat_thaler@xxxxxxxxxxx wrote:
> > The only question is whether to use the current pattern or to invert it.
> > reading of O.150 is that it implies the inverted pattern is used, but it
> > doesn't state it and some parts of the spec appear to say that either
> > inverted or normal might be used for a given pattern. Inverted and
> > non-inverted should perform an equally effective test. The concern with
> > inverting the signal is that people are trying to get these chips done
> > implementations under way will end up with the wrong polarity. It would
> > helpful if someone could verify whether BERT testers support both
> > or just one.
> Thanks Pat.  This is really the crux of the issue - should we mandate use
> of the inverter to make the 10GE standard more compatible with existing
> test equipment.
> I don't represent a test equipment provider, but I'll start the dialogue
> order to gauge whether to invert the PRBS31 generator.
> For bit based BERTs (i.e. those testers for which you program every bit)
> it is irrelevant.
> For Pattern based BERTs (i.e. testers pre-program with varying length
> PRBS streams) most specify the O.150 as the standard pattern and
> allow the "inverted" option (which from the discussion actually removes
> the inverter we're considering installing). There may be some testers
> in this group which only look for the O.150 pattern.
> For Network based BERTs (i.e. those testers which are designed to
> test the optical stream in a network application) most specify O.150
> compatibility. For more details on this, you can try a YAHOO search
> using the key words "O.150 PRBS MEASUREMENT" (although
> there are other equally applicable searches).
> So I'm looking for feedback. Who has used test equipment which only
> accepts O.150 format PRBS. What are other test equipment providers
> recommending.
> N.B. We still have to discuss the WIS test pattern, I'd prefer to do this
> in a separate thread.
> --
> Tim Warland  P. Eng.
> Applications Engineer
> Quake Technologies   (613)270-8113 ext 2311
> Tough Times don't last, tough people do