Do not remove the residual freon as that is enough to effectively allow the P6015 to work up to 40KV. It is the freon vapor that increases the dielectric strength, probe does not need to be completely filled with freon.
That bump at the pulse edge is likely due to grounding. Before trying to adjust the P6015, remove the probe body from the housing by un-screwing the probe handle, disconnect the BNC connector & cable, then slide out the probe body. Once this done connect the generator ground as close to the probe tip as possible with a 50 ohm termination to the probe tip.
Think 100+ Mhz required system response or 3.5nS.
The edge and top of the step response should improve, then adjust the P6015 termination box carefully trying one adjustment at a time to note it's effect. If the dip goes away with the improved grounding and generator termination, leave the adjustments alone. The pulse edge and top response of the P6015 should be almost aberration free at the edge with an absolutely flat top at the step.
If a probe had that amount of pulse response aberration (both probes as displayed), I would consider it totally un-acceptable, the try to figure out why and what is happening.
Bernice
OK, the Tektronix P6015 probe has arrived! The fluorocarbon bottle was empty, no surprise there. There is a very small amount of liquid in the probe, but I don't need it anyway as it can handle 13kV without it.
I also purchased a Testec TT-HVP 15HF as I will need to connect to a couple of places to compare timings.
To check compensation I used a siglent SDG1020 in pulse mode at 20Vpp, 100Hz.
The yellow trace is the Testec TT-HVP 15HF and the blue trace is the Tek P6015. Timebase is 50us. I haven't touched the P6015 but did twiddle the Testec. Do these waveforms look ok?