pls help me find stepconf time for my DIY board

I build my Break Out board and Driver board.
- I use 4n25-Opto for protect my LPT-port
- I use L297 + L298 for driver board.
I've done some research/measure reading and by oscilloscope this is my result

1st pic is , I measure td (delay from input) 8.5us , tr 7.2us (rising edge max) , tf 3.4us (falling edge max) by oscilloscope
2nd pic is timing of L297 for driver board. tclk 0.5us , ts 1us , th 4us
3rd pic is how my calculate for STEP SPACE in StepConf that I use.




This is my question for StepConf setup
(1.) for STEP TIME ,
I do: tclk (of L297) + tr (rising of 4n25) + tf (falling of 4n25) = 0.5us + 7.2us + 3.4us = 11.1us or 11,100ns maybe I put 12,000ns for play safe. Am I correct?


(2.) for STEP SPACE ,
min space between each pulse.
So I do: th (of L297) + ts (of L297) - tclk (of L297)

but I need to add rising&falling of 4n25
So I do: [th (of L297) + tr+tf (rise&fall of 4n25)] + [ts (of L297) + tr+tf (rise&fall of 4n25)] - [tclk (of L297) + tr+tf (rise&fall of 4n25)]
= (4us + 10.6us) + (1us + 10.6us) - (0.5us + 10.6us)
=15.1us or 15,100ns maybe I put 16,000ns for play safe
am I correct or not?


(3.) for DIRECTION HOLD ,
I do: th (of L297) + tr (rising of 4n25) + tf (falling of 4n25) = 4us + 7.2us + 3.4us = 14,6us or 14,600ns maybe I put 15,000 for play safe. Am I correct?


(4.) for DIRECTION SETUP
I do: ts (of L297) + tr (rising of 4n25) + tf (falling of 4n25) = 1us + 7.2us + 3.4us = 11,6us or 11,600ns maybe I put 12,000 for play safe. Am I correct?


(5.) for delay of 4n25 (td) is 8.5us.
So I plus it with Base period Maximum jitter. Which my pc is : 7000ns
So I do 7000ns + 8.5us = 15.5us or 15,500ns maybe I put 16,000 for play safe.

But base on this: wiki.linuxcnc.org/cgi-bin/wiki.pl?Tweaki...ftwareStepGeneration "Making a step pulse requires at least two periods"
So I may plus 2times delay of 4n25: 7000ns + 8.5us + 8.5us = 24us or 24,000us maybe I put 25,000us for play safe

Which one is correct? 16,000ns or 25,000ns?


(6.) When I enter 25,000ns in Base period Maximum jitter. I get Max step rate. 12345Hz (number so cool!)
This is my 1st cnc , I plan to use
screw pit 5mm , step motor 1.8 degree per step , direst drive for micro step = 1.
So that I'll get: 360/1.8 = 200step/1 circle step motor turn ,
And: 200step/5mm = 40step/1mm
My max speed: 12345/40 = 308mm/s

Actually 10mm/s - 30mm/s is all I need for my cnc , so 308mm/s is acceptable for this machine.
Am I correct?


(7.) In theory , the maximum operating frequency for 4n25 of break out board should be : inverse number of the sum of tr (rising edge) and tf (falling edge)
So: 1/(tr + tf) = 1/(7.2us + 3.4us) = 94.33kHz
But that is only theory. Maybe in reality it going to be lower.
Am I correct?


I have lean that for add opto into circuit effect more on timing/speed decrease.
which is add delay to 8.5us , rising to 7.2us , falling to 3.4us. that is much.
But I think it more safe to my pc incase of short circuit of driver board and noise reduction.
So my next design I'll choose more faster opto or change to use rd-isolator

It long question please help me to correct these. And comment me anything. Or if my info is not enough please let me know.

Thank in advance!

I'll start experiment from bigger number 1st. Thank.
But I also want to know about my calculation also whether it right or wrong. Since it'll increse my understaning in using Pulse generation in Linuxcnc also help me to choose better electronice part in my next design.

Thank. Please do review my calculation.

Coz I'd like to see where is limit of my design board. Also to see where am I in compare with commercial board.
Thank