Louis Hart asked:

>How does one make measurements of the dull region in the Hull cell?  Our
>chemist tells me that, after the copper electroplating bath is
>"adjusted", his Hull cell should read or measure 0 - 3 millimeters.  I
>asked how precisely that length could be measured.  He didn't know for
>sure, but felt the measurement was relatively coarse and might not be
>good to more than one significant figure.

The measurement resolution of more than about 1 mm would be impossible.
Even 1 mm would be tough IMO.

>We are trying to put our processes under SPC, where possible.  We have a
>way to do it for the plating bath, but need a good way to measure Hull
>cell dull region.  Thanks for any suggestions or ideas.  Anyone who wants
>to discuss further is welcome to contact me.

In a Hull cell, the sample coupon is inserted in a sample of a plating bath
at an angle, such that across the coupon, plating current density ranges
from way below normal electroplating bath current to well above. The coupon
can be examined after plating. When the bath is properly balanced with
electrolyte, additives, etc. the result should be a relatively large width
where the plating "looks" good, i.e. no burning, fast enough rate of
deposition to cover completely.

If the test is done properly, both ends of the coupon will look "bad" and
there will be a gradation near the center with acceptable plating results.
This indicates what the current density range is giving good deposition.
Since there is no clear demarcation line, it will be subjective as to the
width of the "good" plating, based on color, thickness, and roughness of
the deposited metal.

Usually, bath additions or treatments will be made to the sample, and the
test would be repeated until results return to normal range. The same
treatment can then be made to the operating bath on a volume corrected
basis.

Attempting to measure the width of the "good" plating and apply statistical
techniques might be akin to measuring how much good skin there is on a
person's face to gage their age. IMO, such statistics would be of little
merit as a measure for control of the process. This is because the addition
of a very small amount of contaminant might turn the process bad (and the
Hull test result) in a flash, the difficulty of discerning the "good"
range, and the fact that the meausrement result is only an indirect process
control measure - not an attribute.


regards,

Jerry Cupples
Interphase Corporation
Dallas, TX USA
http://www.iphase.com