Driver Specs for a ST-4045 driver

I am experiencing unpredictable movement results so I am trying to tune my steppers as per the article  linuxcnc.org/docs/html/motion/tweaking-steppers.html
I can't seem to find out or understand the specs I get when looking for my driver (ST-4045).
200khz pulse rate max 
10us step rising edge

So I understand 10us steplength (I believe) but can't find anything on any of the other specs to do the math in the article from.   I need to run the latency test while abusing the machine as it seems I get missed steps or steps in the wrong direction only when there are many quick small movements. 
Any advice would be helpful.

Parallel port or Mesa?

I am using Parallel port. I have full movement on axis but pockets aren't lining up and I miss random chunks of movement on any of the 3 axis. I don't have access to my current settings but they are most likely default or I dumbed them down to get it working just not working properly. I wanted to do the math and make sure the stepconf is right.

Base_period to 100.000
step time to 5000
step space to 5000
reset time to 5000
-
reset should be in hal, the rest in the ini file.

Thank you very much. I will try those settings out and see how the machine behaves. Unfortunately I need to wait until the weekend to get to my machine but thank you for the quick help.

You are welcomed, always.

I have attached the ini and hal file for my current setup.  The issue I am experiencing is when there are small movements on the x axis.  I get +x drift so I am missing -x steps. Y axis seems fine.  Z axis seems fine.   
The test axis in stepconf runs at 1inch/s which is no where near what I can get during cutting.
I saw the Base_Period in the ini and step_reset in the hal but step space and step time I only saw in the stepconf setup not in the hal or ini.
I am going to rewire the breakout and driver and motor for the x axis and see if that helps.

Just for an update.  I rewired the x axis and checked everything.  There was a loose wire but ran a 3 layer 5 inch wide pocket with a 1/4 inch bit.  The second layer outside edge was out 1/8th of an inch but the 3rd layer seemed to line up.  I used the pocket.py to generate the gcode.  The shift is opposite what I had before I rewired it.  The randomness of the issue is giving me trouble.
On the plus side after I rewired the drivers the motors sound a lot healthier.  So bonus.