L297 controler and StepConf

Having read instruction how to set up parameters of the CNC mill, I found among other listed L297 controller (this controller I do have running my step motors) at the first pages of StepConfig wizard. When I chose this solution and start running test of each axis it is so slow motion you cannot imagine.

Then I turned to use ''others'' and put figures in the first pages as proposed in instruction manual. When testing, motors speed it up very well and I am pretty happy with setup.

Questions:

1) Would this settings ''others'' make any issue with L297 controllers, since I am not using originally L297 setup in StepConfig?
2) How to change parameters in original L297 configuration hence the motors can run smoothly as they run with 'others' and figures from the instruction manual for StepConfig provided in this Forum.

Thank you.

Hi

Then I turned to use ''others'' and put figures in the first pages as proposed in instruction manual.

As you have not said what figures you used, can't tell how they vary from stepconf ones

The figures used in stepconf are
step time 500ns
step space 4000ns
dir hold 4000ns
dir setup 1000ns

The spec sheet for the chip lists
clock time 0.5 us
set up time 1.0us
hold time 4.0 us

Not the same naming or time base but appears to be the same values expressed another way

How have you tested them?
If you have only run them through the the stepconf test, would be advisable to try them for real before deciding.

Do you have other opto-isolators in the circuit, these can add delays and change the overall timing figures

regards

Hi ArcEye

eventually I am in front of LinuxCNC.

My set up is as the following:

step time 10'000ns
step space 10'000ns
dir hold 200'000ns
dir setup 200'000ns

If I chose L297 recommendation, motors simple do not move.

Hi

I imagine that Chris got the figures that stepconf uses from the spec sheet.

Either your chips are different or there is something in the circuit, like opto-isolators; which is slowing the timings.
The end result is pulses registering infrequently or not registering, hence slow movement or none at all.

Your figures are quite high, so well worth experimenting for the optimum.
I would probably halve them, see what happens and start from there.
As set at present they will have a knock-on effect on base thread timings and max pulses, which may have a detrimental effect on overall performance.

Regards

Hereafter attached are files 3SB as photo and scheme of my controller (in red areas are L297) as well as data sheet of the step motors I installed.

I am not familiar with opto-isolators, therefor if you can analyze a bit attachments and let me know what to do.

I can reduce the figures, but please let me know what would be optimal range for the motors I use. Leadscrews are ball screw 5mm/rev.

Ok, becoming a bit clearer

LS297 is an integrated circuit ( a chip ) not a controller board.

This board made by HobbyCNC could have any number of components which would delay the timings of the driver.
It is unclear if it has any opto-isolators, as the schematic is of components on a PCB not the circuit.

Their website has scant information, no downloadable spec sheets, you need to ask them what the timings are for the board.

I can reduce the figures, but please let me know what would be optimal range for the motors I use.

The timings are to do with the driver side of the board not the steppers
Not enough pause and pulses will be missed, too much and the maximum pulse rate will be restricted

To quote the wiki ( which is the same as I have been telling you)

If your unsure about your drive timing start high like 10000 for each and test. remember that signal conditioning and opto-isolation can increase timing requirements. That's why you need to know the timings for a driver board, not just the step translator chip that is contains.
Also notice that some controllers step on the falling edge some on the rising edge. This matters as it will change the timing and will be hard to trace. make sure LinuxCNC follows what the controller expects.

The wiki lists one HobbyCNC board wiki.linuxcnc.org/cgi-bin/wiki.pl?Stepper_Drive_Timing

I don't know if it is the same as yours, I cannot see any boards on the site that physically look the same as yours.

However you will see that the timings are significantly lower than the ones you are currently using, 2000ns for all timings.
I would keep step_space and dir_hold at the 4000ns from the chip data and reduce step_time and dir_setup to that 2000ns figure ( 4 and 2 times higher respectively) and test it


regards

Hi,

I have investigated and there is no opto-isolator on my board made by HobbyCNC. In addition, here are my latest set up:

step time 2000ns
step space 4000ns
dir hold 4000ns
dir setup 2000ns

Now motors are running . I tried to set 1000ns and 1500ns for step time and dir setup but motors are ''blocked'', though with set of 2000ns in both motors run.

Now I have other issue!!!

When I set in LinuxCNC home position and then zero in G54 (G54 G0 X0Y0Z0), starting the program code goes as it is, but somewhere during the graving, the machine goes elsewhere. In fact, on the screen the ''tool'' following the path, but in reality the machine (tool) followed the path, and then still ''following'' the graving but at another place.

When I go back in G54 G0 X0Y0Z0, the position in Y is OK, in Z is OK, but in X moved ~-10mm, which was consequence of misleading the code or something else.

Again, on the screen everything looks OK, red line covering white one (tool path), but in reality the tool from the original path slept to the other location.

Any advice?

I suspect all that is happening is that you are losing steps

One or more of the timing must be too short still. Try step time and step space at double the current value and run again.

Somewhere there will be a happy balance but in the absence of specific timings for your board, you will have to experiment to find the shortest period which will ensure no missed steps

regards

I tried doubling the figures for steps, run the machine again and slippering is reduced by half but still is there. I went even up to 10'000 and no further reduction occured .

SOmehow I did not had this slipperi problem when I tried to cut "LinuxCNC" and other small text transfred in the code.

Yesterdy I formated the HDD, installed XP and Mach3 in order to checkout the machine performance in this environment. I set the machine and gave it to John for conversion.

linuxcnc.org/index.php/english/forum/16-...onf-wizard?start=20#

There in Mach, for all axis I put for step 5 and for dir 0, and the motors are speeding up like a hell . Unfortunately, via Vectric Aspire I try to save a toolpath for Mach3, using *.txt file, which when loaded in Mach3 appears nothing comparing what I see in Aspire and what I saw in LinuxCNC. Therefore I cannot justify if settings in Mach would have or not slippering .

However, since I am Linuxocholic, all computers I have are under Ubuntu 12.04LTS, I will come back in LinuxCNC, but I am afraid that I would have again the same issue I was experianced, and I would not be able to get my first proffesional cut .

I will use converted file and see what would happen with new installation. Would LinuxCNC works in LTS 12.04 and when the new version of the software would come under teh Sun.

Attached is file that I was trying to cut, prepared in Aspire (works in 12.04LTS under wine) and very realisticly shown in LinuxCNC, if someone wants to try it.

How to invesrt step pulses? Do you mean to rewire the motors to the drive or something else?

Once again thank you so much for your continuos support.