From: Ernst_Wahl [wahl@dcf105.agere.com]
Sent: Wednesday, April 30, 2003 10:21 AM
To: stds-1450-4@majordomo.ieee.org
Subject: stds-1450.4: binnign snippets


Here is a snippet of code generated by TPG3 according to Teradyne rules:

    seq leak_par(VDDH, 100us, iihInp1Grp, ".IIH")
        $(test_num + 827)
        > (ILEAKN) < (ILEAKP)
        "%6.3f uA" "iihInp1 @ VDDH"
        f(RunAllTests == False ? ( TestEnvNr == SHOP || TestEnvNr == EOL ? FBinIIHEol : FBinIIH ) : NO_BINNING)
        fa(if (RunAllTests == True)  IihFail = True;);

Teradyne associates with the test "seq" a fail bin "f(...)" and a fail
action "fa(...)".  Possible but not used here are a pass bin "p(...)"
and a pass action "pa(...)".

You may rejoice because this fits the currently official proposal almost
exactly: the fail bin is separate from the fail actions.

I'm dismayed because i can't control whether the action or the binning
takes place first, or worse yet what happens if i want to intersperse
them ?

The model where binning is one of the actions can be made to conform to
the Teradyne straight-jacket but not vice versa.



Just for giggles, here is the Agilent version of the same code
(approximately: they didn't start from the same TPG3 database):

    run_and_branch(Enabled_65_67) then
        {
        run_and_branch(iihInp1) then
            {
            }
        else
            {
            if @RunAllTests == 1 then
                {
                @IihFail = 1;
                }
            else
                {
                if @TestEnvNr == @EnvPackage or @TestEnvNr == @EnvEOL then
                    {
                    stop_bin "74","Failed IIH @Post-Burn/EOL test.",,bad,noreprobe,red,2,over_on;
                    }
                else
                    {
                    stop_bin "76","Failed LEAKAGE IIH test.",,bad,noreprobe,red,2,over_on;
                    }
                }
            }
        }

Here the binning/actions are not directly associated with the test
"iihInp1", which is defined elsewhere, and run_and_branch is much more
restrictive (limited to 2 ports) than our flownodes but i have better
control over the binning and variable setting sequence.  Why should STIL
be more restrictive than Agilent ?

-----------------------------------------------------------------------
Ernie Wahl           Agere Systems              Tel:   610.712.6720    
Computer-Aided Test  22F-218B, 555 Union Blvd   Fax:   610.712.4235    
                     Allentown, PA, 18109       Email: ejwahl@agere.com
-----------------------------------------------------------------------