Non-Square Step

More
11 Jan 2016 16:32 #68182 by danielemc2
Non-Square Step was created by danielemc2
I have a Sherline mill with driven with a Delegrane CNC interface to Nema style steppers.
It has run for about 9 years with the older EMC program. I used 22mSec steps.
Installed a blank hard drive and loaded the new 2.7....It crashed because the video card was embedded.
I pulled out another older machine with a parallel port and loaded 2.7. It runs without crashing.

I ran Stepconf and chose "Other" since it is a custom controller.
The pins are set correctly but the steppers were weak when driving strong when stopped.
They do drive in the correct directions when jogging different directions.
My first thought was the step time and step space were incorrect.
No matter what the setting I used the problem still existed.

I pulled out the oscilloscope and looked at the step pulse.
As expected the direction line toggles when direction changes.
BUT THE STEP IS VERY SHORT COMPARED TO THE SPACE. (About 10 %)
Inverting the step signal makes the step very long and the space very short. (as expected)
No matter what I set the StepTime and StepSpace the signal widths do not change.

Thanks for any thoughts that I have not covered....Daniel

PS. Tried to change the parallel port bios settings.....No Joy

Please Log in or Create an account to join the conversation.

More
11 Jan 2016 16:46 #68183 by andypugh
Replied by andypugh on topic Non-Square Step

The pins are set correctly but the steppers were weak when driving strong when stopped.

If the axes move the correct distance in the correct direction then this is extremely unlikely to be anything to do with step timing.
The stepper drives will either step, or not step, the commands have no bearing on the motor current.

BUT THE STEP IS VERY SHORT COMPARED TO THE SPACE. (About 10 %)

This is to be expected, step pulses are expected to be pulses, one pulse commands one step.

I am discussing conventional stepper drives here, it is just possible that your machine has direct phase control. How many pins are used to control each axis?

Please Log in or Create an account to join the conversation.

More
11 Jan 2016 18:16 #68188 by danielemc2
Replied by danielemc2 on topic Non-Square Step
You are very correct. I checked the original 2.6 linux box and it has pulses similar to the new one.
Mostly it was the Z axis that was misbehaving.
Since torque is directly proportional to current I checked the lubrication on the z lead screw and it was in need of attention.
Now that axis moves also. However the torque on all three axis is still weak with both p/c's and both programs.

However I think my original calculations were incorrect.
I wired this controller up in 2005.
I am a firmware guy so my EE skills were not the best, but I am improving.
I am driving 6 volt steppers with a 24 volt supply.
Here is the my math.

Stepper: KP58LM2-097 Unipolar 6 wire.
1.2 amps per phase ... motor rated ohms 6v/1.2a = 5 ohms (This seems high)
Total resistance = 24v / 1.2a = 20ohms
Coil series resistor = 20 - 5 = 15 ohms.

These steppers have V+ center tapped on the stator coils.
I put a 25watt 14.5 ohm series resistor on each side of the coil. IE. 12 total resistors.

This all seems a bit linear when I would expect the power curve to be quadratic.
Please enlighten me.
Thanks....Daniel

Please Log in or Create an account to join the conversation.

More
11 Jan 2016 19:23 #68193 by andypugh
Replied by andypugh on topic Non-Square Step
What are you using as drivers?

Using resistors to limit current is a rather poor way to do it.

Your calculations do not consider the motor inductance, this becomes more and more important as the motor speed increases.

You ought to see a significant increase in performance running the steppers bipolar at the full 24V with commercial stepper drivers doing the current limiting.

Please Log in or Create an account to join the conversation.

Time to create page: 0.085 seconds
Powered by Kunena Forum