calibration drift/instability

I would like to hear if anyone has some suggestions for the following problem. I am
currently trying to set up a conc. measurement method using a CM5 chip. Analyte is
55 kDa, the ligand 50 kDa. The chip is immobilized with 2000 RU. The problem is that
ex. 4 calibration curves in series suffer from drift/instability (the data points do
not lay close to a average curve!). Regeneration is fine and there is not baseline
drift. I tend to think that I am having some conformational change problems with the
immobilized ligand. How do I stabilize the ligand - another reg. solution, washing
after regeneration?

Let me here some suggestions?
Michael

First step is to read the pages on

<Getting the best results>

In <Experimental artifacts> pay special attention to

- Analyte concentration
- Steric hindrance
- Independent vs linked reactions

I hope you get some clues to solve ypur problem.

Arnoud