Hi,
I looked at the figures. You use the Quickinject in stead of the Kinject. The Kinject will probably give you less spikes before and after injection of the sample.
The quick jump at the end of the sample injection can be due to sample / flow buffer differences. When you have a rapid binding you wil not see a junmp at the start. As long as the jump is low, there is no concern. Otherwise dialyse your sample.
If the problem persist, try injecting flowbuffer to detremine if the injection system is working properly. If not, clean! and re-evaluate again.
I hope this will help you
Arnoud