Hello I am using a HP1100 system and I am trying a IEC separation in TSK gel SP-2SW.
My buffers are A:NaH2PO4 10mM pH 3 and B:NaH2PO4 0.5M pH 3. The gradient is run in 100min from A to B. The diference in absorbance between A and B is very small at 220nm but when I run the gradient there appear two "hills" between 20 and 50%B, not so big but I am interested in high sensitivity detection and this could interfere in analyses . Every gradient looks like this .The column and the buffers are new.
I will appreciate any help. Thank you all.

While running your system using Eluent A some impurities accumulate on the column. These elute when the ion strength is increased by Eluent B, resulting in spurious peaks.

Try to insert a high capacity strong ion exchanges -before- the injector. That may reduce the problem.

I would also look into trying to identify the source of the impurities responsible for these baseline disturbances. The disturbance could be caused by impurities in your phosphate, impurities in your phosphoric acid used to adjust the pH (I assume that's what you're using to adjust the pH) and possible impurities in your deionized water. Given the concentration of phosphate you are using, there's a good chance it's coming from your phosphate. You should be using an "HPLC grade" phosphate for this application. Perhaps just switching brands will solve the problem.

Thank you Einar and Chris. I appreciate your comments.