AFG (as gate)+2602B SMU (as source and drain) measure the device I-t curve

Hi all, recently, I use the Keithley 2602B SMU and AFG arbiartry function generator to test the I-t curve of organic electrochemical transistor (OECT). In this measurement, the AFG was set as a gate generating voltage pulse (T=1 s, High altitude= 1 V, and Low altitude = 0), and SMU adopted as drain and source (source is the LO connected to ground) with gate is 0V. Detials are listed in FIgure1. In this way, I cannot get the a gradual rising or falling current with time as reported in many literatures. There is a periodic changed current with huge current range. (Figure 1) However, keep all settings the same, the gradualy changed I-t curve can appear (shown in Figure 2) when I use the keysight+AFG. I'm so confused if this is a wiring error on my part or something else with the instrument.

In the KickStart, for SMU1 where you force 0V, the current limit is set to 1uA.
If you expect more Ids current than this, please increase.  

You mention the AFG is driving the gate.  AFG typically uses BNC.  Is the outer shell of this connected to the common LO of the SMU channels?

I expect each SMU connects their HI to the Source or Drain.  The LO of these two SMU are connected together for a common reference point.
You also need the LO of the Vgs from the AFG to share this common reference point.