Tree325 Retrofit Started

[mmurray70]

I would personally set your max following error a lot higher then the error you have seen while tuning. You might have much higher errors with heavy roughing cuts, or pushing a big drill through steel for example. So you need to set a maximum high enough to avoid nuisance trips in normal situations. I think the max following error is best used to shut the machine down if it gets overloaded to the point of an axis stalling (like a crash, broken tool etc) or any other problem that causes the servo to get a mile behind, like a broken servo motor cable or something. Even if you set it at 500 thats still just 0.030".

As Tom said, direction is easy to fix. Keep in mind that its the motion of the tool relative to the table that your after. Jogging right or X+ will actually cause the table to move to the left. Sorry if this is obvious, I cant remember if this is your first machine or not.

And about that spike, probably is a little stiction after sitting for a while. My Y axis always makes a little thump first when i start the machine. Once it breaks free its fine after. If you did the move twice in a row do you get the same results? I think you will be ok once you fix following error. Good luck with it.

[PeterTheWolf]

....

As Tom said, direction is easy to fix. Keep in mind that its the motion of the tool relative to the table that your after. Jogging right or X+ will actually cause the table to move to the left. Sorry if this is obvious, I cant remember if this is your first machine or not.

Thanks mmurray70, Yes ... this is why I needed to reverse the gain signs. A cut move was negative instead of positive on all axis.

Hi PeterTheWolf,


It seems you have incorrectly configured your rapid parameters into the KMotoinCNC Trajectory Planner Settings rather than the cutting parameters. If I look at the X Axis Cutting Parameters they are:

V=30000
A=100000

dividing by the X resolution of 20225 gives

V= 1.48 ips
A = 4.94 in/sec2

Please set all the Trajectory Planner Settings to the corresponding Cutting Parameters.

Yes, I calculated what I needed for the cutting parameters correctly; however, for some reason I entered the rapid parameters ... Thanks!
I also reversed the signs on the "Input Gain and Output Gains" on all three channels.
Now I am getting correct movement direction and "Single Stepping" also works correctly on the simple "Box" NC code.

Now I will start testing for accuracy.

Question: If I am not seeing cutting moves exactly as programmed, like 1.0" via NC code, do I start adjusting the "Axis Parameters, Cuts/Inch" on the "Tool Setup" to correct this? I made a quick test on the X-axis after I got the Trajectory Planner/Gain Signs correct and I am seeing about 0.005" deviation on a 1.0" move; however, it does return back to zero on a dial indicator I used.


I am going to need help on getting the rest of the electronics wired correctly to KFLOP/Kanalog.

My first challenge I am experiencing is ... I moved the X-axis via KMotionCNC jog until it hit the limit switch and it disengaged ICON relay, which shut the power to the drive/amp down. I then had to manually move the axis off the limit switch before I could get ICON relay to engage again.
How do I over come this such that KLOP/Kanalog shuts the axis movement down at the limit switch such that I can manually jog the axis in the opposite direction with KMotionCNC to get the axis off the limit switch?

I know this is obvious for those who know how to read schematics; however, I am still challenged with comprehending the "Tree325_Schematic" below and it would seem this is the next steps to retrofit this machine. I was hoping to get all the switches/MPG pendant working before working on the spindle motor.


.....
.....

J325 Quick Links:

Existing Machine Schematics
Electronic Cabinet-Right Side
RickB's J325 Retrofit Wiring
Tree Journeyman 325 Designed Specs.


.....
.....

[rbickle]

Peter,

I just saw your thread on the Journeyman. Not sure I understand what you're asking about the 3 phase conversion, but I left mine as it was. The spindle drive and servo drives power from that and I left those alone. Maybe one day they'll quit working and I'll have to find new ones, but for now they're good.

Now that winter is pretty much over in Texas, I'll try to get my video camera out and shoot another video, explaining my construction and wiring of things. Sometimes a kick start is all you need to understand it better.

Rick

[rbickle]

Also, here is my tuning configuration if that helps you...

[TomKerekes]

Hi PeterTheWolf,

Question: If I am not seeing cutting moves exactly as programmed, like 1.0" via NC code, do I start adjusting the "Axis Parameters, Cuts/Inch" on the "Tool Setup" to correct this? I made a quick test on the X-axis after I got the Trajectory Planner/Gain Signs correct and I am seeing about 0.005" deviation on a 1.0" move; however, it does return back to zero on a dial indicator I used.

Yes. If you have an accurate means of measuring longer distances then that would be better. The amount of scaling error on a 20inch move will be 20X larger than on a 1 inch move.

My first challenge I am experiencing is ... I moved the X-axis via KMotionCNC jog until it hit the limit switch and it disengaged ICON relay, which shut the power to the drive/amp down. I then had to manually move the axis off the limit switch before I could get ICON relay to engage again.
How do I over come this such that KLOP/Kanalog shuts the axis movement down at the limit switch such that I can manually jog the axis in the opposite direction with KMotionCNC to get the axis off the limit switch?

I'm not sure what you mean by ICON relay. I thought 1CRE relay enabled the drives. Right now I don't believe you have any Limits configured in KFLOP. Limits are entirely handled in hardware. Once a Limit switch opens the 1CRE relay turns off and disables the Drives. KFLOP has no way to re-enable the drives. I can't figure out how that was handled on the original control. I'm thinking that the signal 207 was somehow driven to +24V to override the Limits. You may need to add a relay that can be controlled by KFLOP/Kanalog to do this. Or add an override push button.

Anyone know how that was handled originally?

Regards

[mmurray70]

A simple but slightly crude way around this is to put a momentary push button in parallel with the limit switchs. So when the limit switches are hit and the loop breaks, you can bypass it temporarily by holding in a button. You just need to be absolutely certain to jog the right direction to back off switch. My old Mazak lathe actually works this way, and a 90's fanuc mill i used to run had a protected switch for doing the same thing, had a cover over it, kinda like the ones you see in the movies to launch a missile lol.

You probably dont need to worry about this too much. Once you get your machine homing properly, and soft limits setup you will most likely never hit a limit switch again. And in the extremely unlikely case that you somehow do hit one, you can turn the screw by hand if your stuck.

I wouldnt rely on an indicator much over a long range. Do you have a 246 block or even 123 block? you could try clamping something to the side to create two steps exactly 6 or 3 inches apart and touch both with an indicator maybe.

[PeterTheWolf]

...
You probably dont need to worry about this too much. Once you get your machine homing properly, and soft limits setup you will most likely never hit a limit switch again. And in the extremely unlikely case that you somehow do hit one, you can turn the screw by hand if your stuck.
....

How and where do I set the home position and soft-limits in KMotionCNC?

Am I able to define a button for home-position? I also want to wire-up a momentary push button to go to home-position once I get it defined.

....
....

[mmurray70]

How and where do I set the home position and soft-limits in KMotionCNC?

Am I able to define a button for home-position? I also want to wire-up a momentary push button to go to home-position once I get it defined.

....
....

First thing is to get your home switches connected to Kanalog and make sure the input is changing state in Kmotion, then you need to do a c-program to home. Yes you can do a user button to home. Does your machine have a third switch on each axis using for homing?

Soft limits are fairly easy, you basically just move the machine to somewhere close to the ends of the travel, just before it will hit limit switches and then note axis position in kmotion, and then fill these values into your init file. But you need to get it homing properly first as the position numbers you have now are meaningless without having the machine homed properly

[rbickle]

Tom,

The ICON relay enables the axes on the Tree mill. It is enabled by 1CRE on the diagram. The E-stop or a limit switch will trigger the ICON relay off, which shunts all 3 axis servo motors to braking resistors.
The 1CRE relay coil is energized in the convoluted diagram on page 2 of the original machine schematics.

Rick

Hi PeterTheWolf,

Yes. If you have an accurate means of measuring longer distances then that would be better. The amount of scaling error on a 20inch move will be 20X larger than on a 1 inch move.

I'm not sure what you mean by ICON relay. I thought 1CRE relay enabled the drives. Right now I don't believe you have any Limits configured in KFLOP. Limits are entirely handled in hardware. Once a Limit switch opens the 1CRE relay turns off and disables the Drives. KFLOP has no way to re-enable the drives. I can't figure out how that was handled on the original control. I'm thinking that the signal 207 was somehow driven to +24V to override the Limits. You may need to add a relay that can be controlled by KFLOP/Kanalog to do this. Or add an override push button.

Anyone know how that was handled originally?

Regards

[PeterTheWolf]

.....
.....

A simple but slightly crude way around this is to put a momentary push button in parallel with the limit switches. So when the limit switches are hit and the loop breaks, you can bypass it temporarily by holding in a button..

OK ... I tried a momentary switch across wires #255 and #207 which indeed over-rides the limit-switch once it disengages the relay for the power to servo/amps.

However, this to me, seems a little too scary. Because if I am traversing (rapid) and it hits the X or Y limit switch and if I do not release the traverse button right away the counts continue to increase.
So, then when I use the momentary over-ride switch (to engage the amp relay) to bring it off the limit switch the axis goes to the new position which is beyond the limit switch (in the wrong direction).

What I did find is, that if I hit the limit switch and while the amp relay is disengaged, move the KMotionCNC position in the opposite direction about 1/2" or so, then when I hit the momentary switch KMotionCNC instantly moves to the new position in rapid speed. Once the new position is reached the axis go into a drift until I re-initialize the KMotionCNC via INIT button.

Isn't there a better way to handle this? Certainly the original Tree design did not work like this.



.....
.....

J325 Quick Links:

Machine
Existing Machine Schematics
Electronic Cabinet-Right Side
Electronic Cabinet-Back SIde
Existing Drive Board SD1525-10
J325 Servo Drive-SD1525 Manual
3-Phase Rotary Convert Used
RickB's J325 Retrofit Wiring
KFLOP 5VDC/15Watt/3A Power Supply
KANALOG Mounted & Connected
Kmotion - Axis Encoder Manual Test of Position via Manual Movement
Kanalog-Encoder Voltage High/Low Checks & 1KOhm Resister
Kmotion Configuration Screens "RUN-AWAY"
Tree Journeyman 325 Designed Specs.
Tree Journeyman 325 Axis-Tension Frequency Settings
Final Axis Tuned Error Parameters


.....
.....

[mmurray70]

I think whats happening is you are overriding the limit switches and turning on the amplifiers without having the axes enabled in kmotion. They would have been disabled by following error which is normal. The axes drift until the axes are enabled again by your init file.

How are you enabling your drives now exactly? Are you still using a switch or something from the testing stage to power on the drives with that relay? You should probably be getting Kflop to do this now. So when run your init file, kflop would switch on the relay to provide power and then enable axis at the same time. So you get no drifting. If you hit a limit switch then you would have to push momentary switch and nothing would happen since everything is disabled. You would need to hold override switch, init again, and then move off.

[PeterTheWolf]

I think whats happening is you are overriding the limit switches and turning on the amplifiers without having the axes enabled in kmotion. They would have been disabled by following error which is normal. The axes drift until the axes are enabled again by your init file.

This certainly sounds correct and what seems to be happening.

How are you enabling your drives now exactly? Are you still using a switch or something from the testing stage to power on the drives with that relay?

Yes ... I still have a jumper wire connected from #200 to #206 as advised back on #41 of this thread.

You should probably be getting Kflop to do this now. So when run your init file, kflop would switch on the relay to provide power and then enable axis at the same time. So you get no drifting. If you hit a limit switch then you would have to push momentary switch and nothing would happen since everything is disabled. You would need to hold override switch, init again, and then move off.

How do I get Kflop to do this? Do I actually connect wire #200 and #206 to the KFLOP and then write "C" code to do something?

.....
.....

J325 Quick Links:

Machine
Existing Machine Schematics
Electronic Cabinet-Right Side
Electronic Cabinet-Back SIde
Existing Drive Board SD1525-10
J325 Servo Drive-SD1525 Manual
3-Phase Rotary Convert Used
RickB's J325 Retrofit Wiring
KFLOP 5VDC/15Watt/3A Power Supply
KANALOG Mounted & Connected
Kmotion - Axis Encoder Manual Test of Position via Manual Movement
Kanalog-Encoder Voltage High/Low Checks & 1KOhm Resister
Kmotion Configuration Screens "RUN-AWAY"
Tree Journeyman 325 Designed Specs.
Tree Journeyman 325 Axis-Tension Frequency Settings
Final Axis Tuned Error Parameters


.....
.....

[m_c]

The KFlop should be programmed to detect the limit, disable the drives (I assume there is some form of enable signal from the KFlop to the drives?), and then disable the axis (if the limit switches aren't connected to the KFlop and set as limits for the corresponding axes, which this type of setup won't be, then the axis won't be automatically disabled on a limit hit)

When you then press the limit switch bypass, you should then need to re-init the KFlop, at which point it should set the destination to the current position, and then you can safely jog away from the limit.

On my machines, the limits are connected as part of the E-stop loop (it's how they came from the factory), so what I have is an E-stop monitoring function, that basically monitors various things (i.e. e-stop circuit, toolchanger status, spindle status - the exact things vary between machines). If any of the conditions aren't met, then it triggers an e-stop cycle, that disables all movement, before displaying a message saying what triggered the e-stop.
The initialisation file is then required to be run to re-enable the machine.

I've attached a couple ZIP files with the code for my machines. They should give you a general idea of how I've got my code structured.
The Cyclone is a lathe and was the first machine I retrofitted. It uses servos, and Modbus to communicate with a PLC that handles the turret.
The Triac is a small mill, and is my current project. The code is not complete, but contains a few refinements to the general structure of the Cyclone code (uses MsgBoxNoWait, so things can happen in a slightly different order). It also only runs steppers, so there is no dealing with axis positions.

If you're going to use any as a base, use the Triac code, as it's the most refined. I've only included the Cyclone to give you an idea how I deal with servo disabling/re-initialisation (I also think this isn't the latest version, as I've modified the code a couple times on the machine, this is just the version I happen to have on my laptop - I do have newer backups, just not handy)

[mmurray70]

How do I get Kflop to do this? Do I actually connect wire #200 and #206 to the KFLOP and then write "C" code to do something?

.....

Basically yes. Im not sure if its best to simply wire in these two points, or if its best to try and turn on the CRE1 relay. You might want to wait for Toms advice about exactly whats best to do and what outputs to use.

[TomKerekes]

Hi PeterTheWolf,

OK ... I tried a momentary switch across wires #255 and #207 which indeed over-rides the limit-switch once it disengages the relay for the power to servo/amps.

However, this to me, seems a little too scary. Because if I am traversing (rapid) and it hits the X or Y limit switch and if I do not release the traverse button right away the counts continue to increase.

I wouldn't expect this to happen if the Max Following Error limits are set properly. I'd expect an immediate following error and the KFLOP Axis should disable and the Jogging motion of the Axis Destination should stop. Furthermore when the Axis is re-initialized and re-enabled the Destination should be set to the current Encoder Position and how far the Axis was Jogged past the Limit shouldn't matter.

Were your Max Following Errors set properly?

The recovery process should require you to press the Limit Switch Override Button, re-Initialize to Enable the Axes, then Jog out of the limit. Initializing first with the drives still disabled may cause a jerk when the drives are eventually enabled.

It would be nice to know how the original system worked. I can't find anything in the schematics that would override the limits disabling the contactor.

I think the 3 signals 208, 207, 255 could be connected to Kanalog Opto Inputs so KFLOP could be aware of EStop status, Limits status, the other disables status as it seems was done with the original control.

Regards

[PeterTheWolf]

Were your Max Following Errors set properly?

I believe I have these set correctly.
X-axis: Rapid Error=18 counts, Max. Following Error=40
Y-axis: Rapid Error=16 counts; Max Following Error=30
Z-axis: Rapid Error=7 counts; Max Following Error=20

It would be nice to know how the original system worked. I can't find anything in the schematics that would override the limits disabling the contactor.

I am sorry; however, I cannot say how it originally worked as this machine was purchased with the originally controller nonfunctional.

I think the 3 signals 208, 207, 255 could be connected to Kanalog Opto Inputs so KFLOP could be aware of EStop status, Limits status, the other disables status as it seems was done with the original control.

Can I please get so help on what exactly I need to do to get these signals connected to Kanalog Opto inputs so KFLOP will be aware and respond to this signal?

I need to understand how this works since I have many, many more signals I need to connect as I move forward with this retrofit.

So, if I start with the signal on wire #255, which is the E-Stop/Limit switch on the front user panel. This switch is a Normally Closed (NC) switch.
When KmotionCNC is initialized all axis will servo; however, when the E-Stop in applied the switch goes open and dis-engages the servo/amp contact ICON via 1CRE relay.

Not knowing how to exactly read this Tree schematic this is what I interpret what is happening:

Wire #200 provides +24VDC to wire #255 which goes thru X-axis limit switch then to wire #256 into Y-axis limit switch then into wire #207 then into I think the Z-axis proximity limit switch, which I think is wire #201coming off the 2CRE relay ..................... So, I am thinking wire #201 wire then should be connected to Kanalog JP15 OpTo Input pin 15 with Wire #4 (GND) on JP15 OpTo Input pin 16.

Would this be correct and how the E-stop/OT limit switches should be connected to Kanalog?

And if this is correct, what should I start out with for C-Code?

I've attached a couple ZIP files with the code for my machines. They should give you a general idea of how I've got my code structured.

Thanks M_C .... however, when I look at your code ... I see how much I don't know and how much I need to learn.

I think if I can get the help for at least one switch wired-up correctly to Kanalog/KLFOP and the C-code working it should give me a better understanding for all the other switches I need to connect.

I think my biggest challenge will be the spindle-motor via VFD ... saving that for the last step in this retrofit.

.....
.....

J325 Quick Links:

Machine
Existing Machine Schematics
Electronic Cabinet-Right Side
Electronic Cabinet-Back SIde
Existing Drive Board SD1525-10
J325 Servo Drive-SD1525 Manual
3-Phase Rotary Convert Used
RickB's J325 Retrofit Wiring
KFLOP 5VDC/15Watt/3A Power Supply
KANALOG Mounted & Connected
Kmotion - Axis Encoder Manual Test of Position via Manual Movement
Kanalog-Encoder Voltage High/Low Checks & 1KOhm Resister
Kmotion Configuration Screens "RUN-AWAY"
Tree Journeyman 325 Designed Specs.
Tree Journeyman 325 Axis-Tension Frequency Settings
Final Axis Tuned Error Parameters


.....
.....

[TomKerekes]

Hi PeterTheWolf,

I believe I have these set correctly.
X-axis: Rapid Error=18 counts, Max. Following Error=40
Y-axis: Rapid Error=16 counts; Max Following Error=30
Z-axis: Rapid Error=7 counts; Max Following Error=20

Those look like good values. But if KFLOP was really set to those values then you should get a following error when the axis hits the limit switch and stops moving. Something isn't right. Maybe you didn't merge those settings into your Initialization C Program? Or didn't execute the C Program? Or executed the wrong Program? Please post your final Initialization C Program. Also verify the Init button is configured to the that program. You can also "Upload" an axis to KMotion.exe to see what the Following Errors are currently set at.

Not knowing how to exactly read this Tree schematic this is what I interpret what is happening:

Wire #200 provides +24VDC to wire #255 which goes thru X-axis limit switch then to wire #256 into Y-axis limit switch then into wire #207 then into I think the Z-axis proximity limit switch, which I think is wire #201coming off the 2CRE relay .....................

I think that is roughly correct. Basically a big string of switches and relay contacts that all have to be closed to get power to 1CRE.

But I don't see any connection from #200 to #205. Rather I think the actual circuit path is:

#200 (24V) -> 203 -> 204 -> 206 -> 255 -> 256 -> 207 -> 208 -> 201 -> 209 (1CRE) -> 4 (GND)

Voltage propagates through the switches from left to right. Any open switch kills power to all nodes to the right. So you might imagine connecting 9 24V light bulbs to all those points to GND. With all the switches closed all 9 lights would be on. With any single open switch the lights to the left of the switch would be on and those to the right would be off. Note with more than one open switch the leftmost switch would determine the state of the lights. The diodes just act like always closed switches but prevent any possibility of current flowing backwards.

Kanalog opto inputs are basically like 24V light bulbs (opto LED + series resistor). So we can connect them up like light bulbs to the points of interest we care about. However unlike simple bulbs, to turn on current must flow in a particular direction. Connect the +to the point that will go to +24V and the - to the GND (which is what you described). I'd start with one such as your #201 which could be interpreted by KFLOP as "All OK".

After that is connected test by looking at the Digital IO Screen to see if the state changes as it should (state=1 when everything is ok and enabled).

Next you can add a C Program loop to constantly read this input and if it goes to 0 disable whatever you wish (XYZ Axes). See the EStopDisablesAll.c example.

On another topic regarding Push Button Override: You might research the model number of the 1CRE relay to see how much current the coil requires. If it is less than 25milliamps then a Kanalog Output can be used like your pushbutton to bypass the limit switches.

HTH
Regards

[m_c]

Personally I would not connect the limit switch override to the controller. To me, I want the limits to kill movement regardless of how I've programmed the controller, as if it was connected to the controller, you have to be fully aware that that output was only active when needed, otherwise you're risking running into the hard stops at full power. With it on a physical switch, should anything go wrong, movement gets killed, and should anything go wrong during recovery, you just need to release the switch.


Peter, the main things to start of with, with my files, is start with the basics.
Each machine has it's main init file. In the case of the Triac, it's TriacInit.c (I think the Cyclone one is a far longer name, to identify from when I moved from the original Mach3 init file). That is the main file, that loads all the critical stuff to get the machine moving.

Then I have a header file that gets included. Triac.h, and Cyclone.h.
That contains constants, that give each commonly used value, a name. So in the case of the Triac, rather than having to use 136 every time I want to check the E-Stop circuit status, I can type ESTOP.
I.e. instead of -

ReadBit(136); // Check Estop

I can use the far more readable

ReadBit(ESTOP); // Check Estop

As you should see by looking at some of the code, not only does it make the code more readable, it also means that should a value need changed, it only needs changed in one place.

The next piece of the puzzle after that, is the EStop file. This contains a function that gets called continually from the main init file loop. The whole point of this function is monitor everything on the machine that could cause a situation I don't want. Things like has the E-stop circuit failed, is the spindle on but the tool carousel still out, is the drawbar active while the spindle is on, and so on.
It uses a couple other programing conventions, but I wouldn't worry about that just now. Just shout when you get to that point, and I can explain how things get handled if needed.

Once I get the Triac finished (just need to find time to wire up the new operator panel, and add connectors for a touch probe), I was going to do a full write up on how all the files interact.

[rbickle]

m_c,

Can you please tell me the correct syntax to do this? I've been searching the channel structure, but I'm not sure which variables are correct.

Thanks,
Rick

When you then press the limit switch bypass, you should then need to re-init the KFlop, at which point it should set the destination to the current position, and then you can safely jog away from the limit.

[m_c]

m_c,

Can you please tell me the correct syntax to do this? I've been searching the channel structure, but I'm not sure which variables are correct.

Thanks,
Rick

When you initialise the KFlop, you should have a process that enables the servos, then enables the axis, so you should have code similar to this -

SetBit(SPDLENA);
			Delay_sec(0.3);		// delay to avoid PID windup due to servo drive taking time to enable
			EnableAxisDest(7, ch7->Position);

That's actually part of my Triac spindle on code, as it uses a servo spindle, but the process is the same for axis servos.
The SetBit turns on an opto that powers the servo drive enable signal.
As that particular drive takes a short time to become fully active after receiving the enable signal, I had to add a short delay to avoid the axis either disabling due to following error, or banging to required speed, as the KFlop/Kanalog expects it to move, but it doesn't for a short period.
Then the relevant channel gets enabled, at which point it comes under closed loop control.
I think using EnableAxis(7) achieves the exact same thing as EnableAxisDest(7, ch7->Position).

With that in place, the only difference when you've hit a limit switch, is you need press and hold the override button until everything is initialised, so you can jog of the limit switch, at which point you can release the override switch.

Although, this method is only for if you have limit switches physically disabling things. If the limit switches aren't disabling anything, but are instead feeding back to KFlop inputs, they should be configured using the axis/channel limit switch options. At which point, if you hit a limit switch, the KFlop is smart enough to simply let you jog back of the hit limit switch, and no re-initialisation should be needed.
Off course, if you have accurate homing, then you should be able to set softlimits, which should prevent you from hitting limit switches anyway.