Question about THC parameters

Your torch speed will be up to your maximum velocity. The whole intent of a PID is it is a self adjusting algorithm that converges on the set point. If you google it, there are plenty of explanations.

I have no intention of digging too much into the details of how a PID controlled THC works but I just want to control the thc exactly like what I want. The reason is that when I cut a rectangle, the torch will cut one edge of the rectangle at the target voltage(69v) but when the torch changes the direction, the voltage will change to 71v or 72v even in the middle of a long edge. I was wondering why the thc wouldn't cut all the same edges of the rectangle at the same voltage. Is that because my thc speed not fast enough or I missed something in the PID setting?

It might be a good idea to disable VAD to see how it goes. I always enabled it to lock the torch around the corners. Will try it next time when I am around the machine.
Yes, THC ACTIVE is on and everything is working as expected(VAD, Up and Down led), just the cutting voltage is slightly different for each edge. By the way, the rectangle I cut is around 50x100mm, my feedrate is 3900mm, my acceleration for x/y/z axis is 1500mm/s/s. Is the dimension of the rectangle too small for the thc to kick in after the VAD lock?

I guess the real question is; Is the cut quality still OK  on the second and subsequent sides?
If so, don't worry.
Typically torch voltages changes 10 volts per mm so 1 volt is only 0.1mm
Say if you reduced to 80% and had to get back to cut speed, It will take 0.52 seconds so that might mean you travel say 3.5mm..
3900mm/min = 65 mm/sec so you will travel about 34mm in that time

There is a difference between the edges/sides, I can see it with bare eyes but it's still in the tolerance. I even changed the threshold to 0.5v to see if there is any difference but not really. I guess I have to live with this until I figure out something. Btw, how do you get the result 0.52s for the cut speed to get from 80% to 100%?

I dunno either now...
3900 mm/min = 3900/60 mm/sec = 65 mm/sec
65 * 20% = 13mm/sec
to accelerate from 52mm/sec to 65mm/sec @1500 mm/sec/sec
= 8.6 ms

Is that right?

My table at 5000 mm/sec/sec acceleration does it in 2.6 ms. I have noticed an improvement with corners since I upgraded it.