I am trying to calibrate a 30 cm/7.5 mm i.d. column with commercial MW standards from Bio-Rad. Five components of different MW are included in the mix: thryoglobulin, gammaglobulin, ovalalbumin, myoglobin and vitamin B-12 (elutes in that order). The mobile phase is aqueous-based containg salt (MgCl2) and buffer and some EDTA. My problem is that the peak separation visually looks good but the plate count is very low. It looks as though I have good separation between peaks - no overlap, no shoulder/rabbit ears and minimal tailing. Any ideas why the plate count is low? It should be >10,000 and I am coming up around 6000.
Thanks.

The only way I have been able to duplicate plate counts is to run the exact same method the column vendor used. If that method shows the column to be "good", I set up my analysis, record the plate count and use that as my benchmark for determining whether the column is going bad.

Are you making equivalent comparisons, apples with apples rather than apples with oranges, so to speak? Are you comparing your current plate counts with data from previous analyses you or your co-workers have performed, or with that supplied by the column manufacturer accompanying the column? Has all plate count data you are comparing been run under identical conditions? How old is the column?

I ordered the standard the supplier used and ran that. The plate counts are still low. The spec sheet states that the plate count should be around 12,000 and I get around 7,000.
H.J.- I compared the chromatogram from the MW standards with a co-worker and the results are similar, but he never calculated the efficiency. The most recent data (i.e. the new standard) was run under identical conditions and the column is 6 months old.
Thanks.

There are still two possibilities for the low plate count. Either the column has deteriorated, or your instrument has too much bandspreading to count the plates.
Here is what you need to do to figure this out. Inject the plate count marker on the column and measure the plates and the peak width in volume units. Then disconnect the column from your instrument, and replace it with a low-dead-volume coupling. Now inject the same sample at the same flow rate. Compare the peak widths with column and without column to each other. If the peak width without column is a substantial fraction of the peak width with column, the problem is not the column but the instrument.
Specifically, the squares of the extra-column peak width and of the true column peak width add up to generate the square of the total peak width. Equations can be found in any decent textbook. I can send you the stuff too.