Hurco BMC20 Dynomotion Retrofit

[TomKerekes]

Hi Troy,

The Coordinate System definition:

DefineCoordSystem(0,1,-1-1);//defineXY(Z and A not used)

needs to be moved from the definition part of the program to a place that will be executed ie in the main() function just before the forever loop.

Regards

[Need TECH Help!]

.

[Need TECH Help!]

Hi Troy,

The Coordinate System definition:

DefineCoordSystem(0,1,-1-1);//defineXY(Z and A not used)

needs to be moved from the definition part of the program to a place that will be executed ie in the main() function just before the forever loop.

Regards

Hay Tom,
I tried that last night and got an error when compiling, too few arguments to function.(So i went to bed ) Then after i read your post today i noticed when i removed the Z (2) from being defined that i forgot to put a ',' after the -1. Now the step buttons work.

Now that i can execute gcode i have been testing movement of Y axis. I have gage blocks on table as my known standard. (See pic). Block lengths is 4.0" and 2.0".Feed move at 20. IPM. When i do a feed distance move of 4.0" the indicator reading is off by .001" . So did a test with 2" block and a 2" feed move and show a indicator reading off by .0005". Then did a 6" move and indicator reading is off by .0015". That figures out to be about .00025" per 1.0". I know ballscrews have a lead pitch error per foot, but this would come out to .003" per foot!

Tested X axis with 4" gage block and indicator reading is 0.

Got any ideas what i might have off?

Thanks,
Troy

[TomKerekes]

Hi Troy,

It isn't clear to me exactly how you are performing the test. But I'm sure you are doing something valid. When you move back to zero does the position repeat? The idea would be to determine if you have some issue with loosing encoder counts or drift, or if it is truly a scaling error. If it is a scaling error you could correct for it by changing your Axis resolution (from 20320 to either 20325 or 20315 - it wasn't clear which way you were off). I can't think of anything that would cause such a small scale error other than the lead screw.

Regards

[Need TECH Help!]

Hi Tom,
The test is zeoring out indicator on end of gage block then zero axis. Then do feed move to other end of block and read indicator. Yes when i move back to 0 it repeats.The actual movement from a programmed feed move of 4.0" was actually 3.9990. So i changed the KMotionCNC steps/in to 20325 and indicator now reads 0 after move and repeats back to 0 starting point.How would i adjust C code to get the correct resolution for rapid and jogging moves?
Will need to bring home my .0001" indicator from work to do final check. Indicator iam using is graduated to .0005". There might be a possibility that someone has had the ballscrews reground at some point.Because iam getting no backlash when jogging in small increments (.0001-.0002") in both directions of axis.
Anyhow, after i adjusted to 20325 and got good reading with feed move i tried a position move (G00) and it appears to be right on, but drive faulted out at end of move. Which brings up another issue with my C program.

The ALM pin output of drive is connected to Kanalog Opto Input bit 141. My C code is supposed to clear some I/O bits and put KMotionCNC into ESTOP and have a pop up message stating "Y axis Drive Error". But C code is not doing it. When looking at Kmotion.exe Digital I/O screen bit 141 is toggling just like the flashing error light on servo drive. How can a signal like this be used in the C code?

Thanks again,
Troy

[TomKerekes]

Hi Troy,

The test is zeoring out indicator on end of gage block then zero axis. Then do feed move to other end of block and read indicator.

Just curious how do you do that? Wouldn't the block be in the way during the move and break the probe? And wouldn't you need to use the other side of the probe tip to probe the other end of the gauge block adding the probe width to the measurement?

The ALM pin output of drive is connected to Kanalog Opto Input bit 141. My C code is supposed to clear some I/O bits and put KMotionCNC into ESTOP and have a pop up message stating "Y axis Drive Error". But C code is not doing it. When looking at Kmotion.exe Digital I/O screen bit 141 is toggling just like the flashing error light on servo drive. How can a signal like this be used in the C code?

I don't see 141 defined anywhere in your code. I see:

#define YSDERRORLINEBIT 140

should that be 141?

Regards

[bhurts]

Your gauge block should be held against a backstop. You would touch the gauge block with the indicator and preload a bit to be able to have a little positive reading as well. Then you would zero the axis so both it and the indicator read zero. Then move the gauge block out of the way and move the indicator into the backstop. The move size is whatever the length of the gauge block your measuring. If the indicator reads zero when it stops then you have moved exactly the length of the gauge block. If it is reading positive then you moved to far the distance from zero shown on the indicator. If it's negative then you haven't moved far enough.

Ben

Sent from my HTC6525LVW using Tapatalk

[Need TECH Help!]

Hi Tom,
There is a block trammed in straight and clamped to table. Then the gage block is set against the fixed block. After you zero out, you slide gage block out of way or move other axis (in this case X axis) over to a set amount say .25" or just enough to clear block. (The further you move axis away from zero point the more room for error if blocks are not trammed in exactly straight.) Then do test move, check indicator reading, move back to 0 on Y axis(or what ever axis is being tested) and either slide gage block back in to place behind indicator probe or move other axis(in this case X) back to set zero.

Sorry, i had my X and Y ALM wires switched and what should be Y axis ALM on 140 was 141. So, i have switched wires back and now the #define YSDERRORLINEBIT is 140 again.

[bhurts]

It looks like in his method he did not move the block. Instead he moved in the opposite axis a bit to slide off the edge of the block them moved against the stop. This can work fine also but may add some error if the axis are not exactly 90 degrees each other.

Ben

Sent from my HTC6525LVW using Tapatalk

[Need TECH Help!]

Hi Ben,
Looks like we are posting over top one another . I actually tested it both ways to make sure. The picture was for more reference purpose.

Troy

[mmurray70]

The ALM pin output of drive is connected to Kanalog Opto Input bit 141. My C code is supposed to clear some I/O bits and put KMotionCNC into ESTOP and have a pop up message stating "Y axis Drive Error". But C code is not doing it. When looking at Kmotion.exe Digital I/O screen bit 141 is toggling just like the flashing error light on servo drive. How can a signal like this be used in the C code?

Thanks again,
Troy

Its probably a good idea to wire in the alarm signal, but keep in mind if you set max following errors the machine will stop anyway as soon as theres a problem with a drive and the move is not completed. I just set my machine to do an E-stop when following error was exceeded. Saved the hassle and clutter of extra wiring and saves a few inputs if you need them. Its not as critical say as spindle drive alarm, you definitely want to monitor that. Machines dont cut well at all after the spindle stops lol.

[Need TECH Help!]

Its probably a good idea to wire in the alarm signal, but keep in mind if you set max following errors the machine will stop anyway as soon as theres a problem with a drive and the move is not completed. I just set my machine to do an E-stop when following error was exceeded. Saved the hassle and clutter of extra wiring and saves a few inputs if you need them. Its not as critical say as spindle drive alarm, you definitely want to monitor that. Machines dont cut well at all after the spindle stops lol.

I have a similar approach using a watchdog, but if user runs the Main INIT code again the following error is reset and user can jog other axis without a fault until user jogs axis that faulted, then KmotionCNC will go into ESTOP again. With the ALM input from drive KMotionCNC stays in ESTOP until servo drive is reset to clear the fault.

I actually found the issue with my code when i was working on my MPG code. When i added the Servo Drive Error input part of code...

// Handle Y Axis Servo Drives Error Line
		result = Debounce(ReadBit(YSDERRORLINEBIT),&ysdecount,&ysdelast,&ysdelastsolid);
		if  (result == 1)
		{
			DoPC(PC_COMM_ESTOP);
			ClearBit(145);
			ClearBit(153);
			ClearBit(154);
			MsgBox("Y ServoDrive Fault",MB_OK|MB_ICONEXCLAMATION);
			ClearBit(144);//Enable/Disable MainContactorRelay for servo drive main & logic power.

i forgot to add the state variable for switch debouncing ....

int ysdelast=0,ysdelastsolid=-1,ysdecount=0;

Now when drive faults i get the pop up message window stating "Y ServoDrive Fault" and after i click the OK button on pop up my Main Contactor for servo drives disables. And the pop up goes away. If i dont clear bit 144 after i click OK i get a loop of my pop up window because the ALM pin of drives toggles state of kanalog input .

At some point it would be nice if KMotionCNC had a status line somewhere, like at top of window in the title bar. And either have the message scroll across or simply flash.

But anyhow, ALM input is working.

Now iam working on getting the code for my pendant into Main C code. Currently i have the MPGsmooth code(which works better than the new Hurco at work) in my Main C program and working. The pendant has a + and - button for continuous jogging, that i cant figure out how to code. The 2 buttons should act like an axis Jog button on the KMotionCNC screen. And my selector switch on pendant would select what axis to jog.
I have looked at the examples JogWithPot and JogWithTogglesAndPot c codes. But cant figure out how to combine the 2 and use the Jog Rate DRO box i have on KMotionCNC for jog speed. Curious...Is there code for using Hotkeys that are assigned to on screen buttons that could be used ?
Anyone got any clues?

Thanks,
Troy

[TomKerekes]

Hi Troy,

Now iam working on getting the code for my pendant into Main C code. Currently i have the MPGsmooth code(which works better than the new Hurco at work) in my Main C program and working. The pendant has a + and - button for continuous jogging, that i cant figure out how to code. The 2 buttons should act like an axis Jog button on the KMotionCNC screen. And my selector switch on pendant would select what axis to jog.
I have looked at the examples JogWithPot and JogWithTogglesAndPot c codes. But cant figure out how to combine the 2 and use the Jog Rate DRO box i have on KMotionCNC for jog speed. Curious...Is there code for using Hotkeys that are assigned to on screen buttons that could be used ?
Anyone got any clues?

I would prefer not to use Hot Keys as that would unnecessarily involve Windows and the PC and be less responsive.

Maybe something like this:

#1 save the JOG Rate DRO value in a program global variable so it is accessible from all code in the Init Program. (move the variable "double rate" from inside the UpdateJogSpeeds function to before main()). Variables inside functions are temporary, only usable from within that function, and live only from the time the function is entered until the function is exited. Where variables outside functions are usable from any function and live for the duration of the program.

#2 define a new function called something like ContJog(int dir) that will handle continuous jogging in the specified direction.

#3 define a new function called something like StopContJog() that will handle stopping jogging.

#4 debounce the + Jog button. When depressed (ie result==1) call ContJog(1). When Released (result==0) call StopContJog().

#5 debounce the - Jog button. When depressed (ie result==1) call ContJog(-1). When Released(result==0) call StopContJog().

#6 write the ContJog function: Check JOB_ACTIVE and if active don't Jog. Determine which axis to Jog based on selector. Determine Resolution for the selected axis. Do the math to convert Jog Rate ipm (rate) to counts/sec for that axis and with the appropriate direction. Issue a Jog to that axis

#7 write StopContJog function. Stop the selected Axis by commanding Jog speed 0. Maybe better to all axes in case operator does something crazy like changes the selected axis while holding continuous jog button.

Let me know how much makes sense

Regards

[Need TECH Help!]

Hi Tom,
That sounds a whole lot more robust than what I was trying to coble together. Big plus not relying on PC. And blocking user error with selecting another axis during jog.

[mmurray70]

If you change the code to the following it should only display message once. Im guessing these bits you are clearing are your servo enables? In any case check for result = 1 and some other bit that you expect to be on and you will be clearing will make sure it will run only once. Good idea actually to check status of a bit whenever you turn one on or off, no need to continuously set it over and over again.

// Handle Y Axis Servo Drives Error Line
		result = Debounce(ReadBit(YSDERRORLINEBIT),&ysdecount,&ysdelast,&ysdelastsolid);
		if  (result == 1 && ReadBit(153))
		{
			DoPC(PC_COMM_ESTOP);
			ClearBit(145);
			ClearBit(153);
			ClearBit(154);
			MsgBox("Y ServoDrive Fault",MB_OK|MB_ICONEXCLAMATION);
			ClearBit(144);//Enable/Disable MainContactorRelay for servo drive main & logic power.

Be careful with the MPGsmooth program. It can cause a runaway under certain conditions. Id strongly reccomend making this small change here: http://www.cnczone.com/forums/dynomo...er-caused.html

[Need TECH Help!]

Hi mmurray,
Wondered if there was something simple that would also look for a bit .Thanks for the heads up on the runaway. Will change code and see how works.
Thanks,
Troy

[mmurray70]

Performance will be same as before. Tom said its fine with that mod up to 5khz which is 3000rpm for a standard 100 count per rev MPG. Most people wont be spinning that fast

Heres a vid of a guy on youtube with the exact problem i had:




Pretty serious issue IMO. Tom, might not be a bad idea to mod the version supplied with new releases of kmotioncnc. Pretty scary to see your machine take off randomly when everything appears to be working properly.

[Need TECH Help!]

Hi murray,
Ya, i have had enough run aways to last me a while when i was trying to tune servos analog. Made the code change and MPG still operates nice and smooth. Thanks for the heads up.

On another UP note. DMM-TECH and CNC4PC has fully reimbursed me for the 3 replacement drives i bought And DMM is keeping the blown drives i sent back. Big thanks for them doing that. It puts my budget back on track.

[Need TECH Help!]

mmurray,
On your retrofit did you control your VFD with RS232? Or a different control method? I plan to down the road to have a tool changer and from what i can understand the RS232 protocol would be the best way of controlling VFD.?
Thanks for any info.
Troy

[mmurray70]

Good to hear they are replacing your drives. That makes me much more likely to buy from them in the future. Hopefully they fix the errors in their manual now too. Especially about the max input voltage at least.

Im controlling my spindle VFD with the analog outputs from kanalog. One relay driver output turns the spindle on clockwise, another turns the spindle on CCW and the analog out 0-10v sets the speed. Fairly simple in my case. I have no idea about anything RS232. I think i would probably avoid it if you have the option to go 0-10v. But honestly I dont know it, it might have its other advantages that i know nothing about.

All my tool changer is controlled by Kflop too, no plcs or anything. I think i had 5 sensors to connect to verify each step and 6 outputs to control all the tool changer motions. In my case i just connected everything to a Konnect board and used the solid state relays originally in the machine. Its super easy to control everything on a Fadal, couldn't ask for an easier machine to control with Kflop.