I know this is a late answer, but it is difficult to give you some advice without knowing the instrument you using.
Please inform us the type of instrument you use and explain how it works. For instance, are there 1, 2, or 3 syringe pumps or is the flow by a peristaltic pump? How is the buffer / analyte switching done: is there an injection loop or is it done by injection of the sample in the running buffer (biacore style)? The flow channels, are they in series or in parallel? How is the temperature stabilization done? More information can help us better.
On the other hand, when I look at these figures, it looks like an improvement over the earlier you have posted. Part of the signal increment looks like drift. This means that washing probably will solve this. Make sure you have washed the chip (and flow and injection system) properly before capturing the linker DNA. Then wash the system again thoroughly before injection of the DNA.
I use peristatic pump when I inject sample and buffer, in all of channel. and each buffer and sample stored in tube(buffer:50ml falcon tube, sample : 1.5ml ep-tube), and when buffer changes, I mones tube to tube which tube of peristic pump. sample flows 40ul/min.
all of channel flous like =(parellel), they cannot shuffle when flow on chip.