[Gunmachinist]

All right I've brrn working on this for 2-3 days and no real progress. what do I need to do.

[Gunmachinist]

Didn't realize the top pic has all the information covered, I'll get a better picture as soon as I can.

[TomKerekes]

Hi Gunmachinist,

I'd suggest working on one axis at a time. Its less confusing.

You should understand that statements like "All right I've brrn working on this for 2-3 days and no real progress" provides no information on what you did, what happened, and what you understand happened. So we can't really help. The screen shot of axis 2 hides all the critical parameters (motion Profile, PID parameters, and max limits). Its better to take individual screen shots of separate windows with Alt Print Screen of Config/Flash, Step Response, and IIR Filters. We also don't know what amplifier you are using, what mode, and what power supply. I guess its hard to understand we are not there looking over your shoulder to see what is going on. So its necessary for you to explain.

[Gunmachinist]

Hi Gunmachinist,

I'd suggest working on one axis at a time. Its less confusing

Ok, I'll attempt to do a better job explaining.
I ended up making a power supply from old parts of the animal system. I connected a transformer to a square (that I can't remember what it's called) and a capacitor. It supplies 50v unknown ampsbut it is run to my amc servo drives through a 15 amp fuse. I haven't had the power fault out yet with this supply. I've been changing the A and V parameters mostly trying to get the red and blue lines to match. The picture which shows all the parameters is the axis I'm having most trouble with it will move about 1 inch and the return 2.

[TomKerekes]

OK now we know you are using 50V.

Look at the plot with the motion profile parameters visible and explain what you understand is happening.

[Gunmachinist]

OK now we know you are using 50V.

Look at the plot with the motion profile parameters visible and explain what you understand is happening.

On that motion profile I'm not sure exactly what the PID parameters actually do. The V= velocity A=acceleration and J=jerk. Time is how long the motions are recorded and size is the distance of movement which I have set to about 2" of movement and record for 3.5 sec. From reading the information on your website and wiki I understand that you try to determine what the highest V is and then reduce by about 20% then find the highest A value then find J. I don't understand what I'm looking for to determine when V is max. I can up the V value until its extremely high without faulting or making irrational movements. When I increase A I can't tell if anything changes for the better but if I decrease A I get a little better movement. I don't understand that from looking at the graph, what I can do to get better adjustment and stop over shooting.

[TomKerekes]

Attachment 448498

Can you see that the Position (red) plot is changing much slower than the Command (blue) plot?

Can you see that the Output (green) quickly goes to the maximum 2047?

This is like trying to follow a rocket in your car. All all you can do is floor it and fall hopelessly behind. When the Rocket reverses and passes you then you just turn around and again fall hopelessly behind going the other direction.

Look at the slope of the Position (red) while the output (green) is floored. Notice it is a fairly straight line that changes about 15000 counts over 1 second. So the maximum speed of your system is about 15000 counts/sec. Which is about 2 inches/second. To go faster you would need to have higher motor voltage or make some other physical change.

So you need to slow down the motion profile (blue) to around V=15000 so it is something your system can actually do.

The setting of V=1million is crazy. But with Acceleration 50000 it would take 20 seconds to reach that speed. Because your test move only takes about 1 second it never gets anywhere near that speed. More like 50000 which is still much faster than your system is capable of. That's why setting V higher and higher has no effect. Your Acceleration and Move size is limiting the Velocity not the Velocity Limit. You might also note that the blue plot is 'S' shaped and has no constant slope/constant velocity. That is another clue that the max V is never being reached.

Reduce V until the Position is able to follow the Command and the Output isn't floored.

[Gunmachinist]

So, should I reduce the output? When adjusting my values for V I should start at 15000 and work up? Should I start A at 100000 and work down after I find V max?

[Gunmachinist]

It's Getting better, but I'm not shure which direction to go with the output(green line).

[TomKerekes]

Looks much better. You didn't respond to my explanation, answer questions, or tell us what you did, or what happened, so I'm not sure how to help.

Output (green line) is going to about 1400 of 2047. So you potentially might be able to move or accelerate a bit more.

Output of 1400 is needed but Integrator is only allowed to go to 1000 so is probably only helping to a limited degree.

[Gunmachinist]

Looks much better. You didn't respond to my explanation, answer questions, or tell us what you did, or what happened, so I'm not sure how to help.

Output (green line) is going to about 1400 of 2047. So you potentially might be able to move or accelerate a bit more.

Output of 1400 is needed but Integrator is only allowed to go to 1000 so is probably only helping to a limited degree.

So, should I reduce the output? When adjusting my values for V I should start at 15000 and work up? Should I start A at 100000 and work down after I find V max?<<<<that was my answer
to your last post.

The last I did was to adjust the V to what you suggested and played with A more. I then got pissed when I cant see what I'm doing is doing anything so I made D=30 and the Blue and Red lines came closer so I made P=30 and that did something. I then worked up and down with both until I got what you see in my last post with the exception the green line is now AFU. On the graph I believe the numbers on the left are size and bottom is time but what do the numbers on the right side indicate? What does Integrator do?

So you potentially might be able to move or accelerate a bit more.

How? If output max is 2047 should integrator upped until output is maxed on the graph?

A video of this explaining all the different settings would be beneficial.

[TomKerekes]

So, should I reduce the output?

I don't understand what you mean by this. You don't have direct control over the Output. The Servo will Output whatever is needed to do the job.

Its like driving a car and maintaining 30mph going over a hill. The output is like the gas pedal position while driving over the hill. Imagine you monitor the pedal position and see it never goes down more than 70% so is never "floored". So now you know you can take that hill at 30mph, and you might be able to do 40mph. Or if we reduce to 20mph we may only need 50% of pedal.

What does Integrator do?

The servo Integrator works much like a cruise control in a car. Say the cruise control is set for 30mph and the current speed is 29mph. Then the cruise control will keep pushing the pedal more and more and more until the speed is exactly 30mph (or pedal is floored). Now if you restrict the cruise control to pushing the pedal half way down and going 30mph takes more than half pedal then the speed may stay at less than 30mph.

So my suggestion was to increase max integrator to 2047 to allow the Integrator to use full power if needed.

When adjusting my values for V I should start at 15000 and work up?

Yes. Basically in general walk before trying to run. The basic tuning process is to start with something that works but might be slower or less accurate than what is possible. Then increase things to go faster or more accurately until something goes wrong. Then you know what the limits are.

Should I start A at 100000 and work down after I find V max?<<<<that was my answer to your last post.

That's a question not an answer. And is unrelated to my questions. Please go back and answer the two questions.

The last I did was to adjust the V to what you suggested and played with A more.

I would have found what V was possible first.

I then worked up and down with both until I got what you see in my last post with the exception the green line is now AFU.

Its not really bad at all. Its like a plot of your gas pedal position while driving. It varies up and down where there is accelerations and deceleration and based on road conditions.

I believe the numbers on the left are size and bottom is time but what do the numbers on the right side indicate?

Good question. The blue and red are in encoder counts and use the scale on the left. The green is the DAC output and uses the scale on the right.

[Gunmachinist]

I've been working on trying to get my servos all tuned over the past couple weeks and had to re run my wiring because the servos weren't acting correct. I found 2 wires that were accidentally swapped which corrected the servo situation. I replaced my AC to DC power inverted to supply the max 85v to my AMC servo controllers.

I almost had all 3 servos tuned and decided to upgrade to the current software Version 4.35d. I can command movement in the console screen, but can't get my motors to run in step response screen. WTF gives?

[TomKerekes]

The Step Response Screen uses the settings in the Screens (Step Response, Config/Flash, and IIR Filters). Could you have parameters in the screens set incorrectly?

You might read Upgrading from previous KMotion Versions.

[Gunmachinist]

Ok, somehow I have everything working again and all axis are tuned (see screen shots). I have loaded all axis into a c program, now I need to connect the spindle and get it working with Kflop/Kanalog to control spindle speed, and coolant on/off. I kept a board from the Anilam system that I believe will allow me to control spindle speed and coolant on/off. I have included the schematic for the board, what do you think about using it with JP8, SW0-7? I am feeling better about this the further I get.

[TomKerekes]

Hi,

You might want to re-tune axes 0 and 1 as they are saturating the output and not following well.

Yes you should be able to drive those relays by driving the pins on P3 with the Kanalog Relay Drivers on JP8. You will need to connect the 24V power supply common to Kanalog GND. Its not clear to me where that is. Possibly P5-23 bit it isn't clear.

[Gunmachinist]

What do you mean by saturating the output? I thought if the blue and red matched, that is what I needed to look for.

I was planning on attaching the 24v comm at TB2-15 and 24v+ at TB2-18 (its marked on the board). Do I need to connect 24v through Kanalog?

I also plan to connect an encoder to the spindle so I can control the speed in my G code program.

[TomKerekes]

What do you mean by saturating the output? I thought if the blue and red matched, that is what I needed to look for.

Saturation is like in your car the pedal is to the floor and you still can't keep up:

Attachment 451042

I was planning on attaching the 24v comm at TB2-15 and 24v+ at TB2-18 (its marked on the board). Do I need to connect 24v through Kanalog?

Kanalog is like a switch to Kanalog GND. Its not clear how your machine is wired or what TB2 Pin15 (Coincidence) is. But the basic idea is:

Attachment 451054

[Gunmachinist]

QUOTE=TomKerekes;2409882]
Kanalog is like a switch to Kanalog GND. Its not clear how your machine is wired or what TB2 Pin15 (Coincidence) is. But the basic idea is:

Attachment 451054[/QUOTE]

I assume Kanalog ground would be a ground terminal on JP6 or JP 11? It also looks like there is a 24v connection on one if the P3 connections.

[TomKerekes]

I assume Kanalog ground would be a ground terminal on JP6 or JP 11?

yes

It also looks like there is a 24v connection on one if the P3 connections.

I would expect that to be somehow already wired to a +24V supply. It does not need to go to Kanalog as shown in my diagram.