Messages

81

General Mach Discussion / Re: worm gear cad design

« on: December 29, 2013, 07:38:42 AM »

Hi John:

I too use BASIC to generate cams with uniform acceleration into a constant velocity curve with deceleration at the end. To deal with the cutter comp I have the program write a script to be read by AutoCAD or BobCad. Once in CAD I fit the spline and offset the curve by half the cutter diameter before generating Gcode.

I am comfortable in BASIC and wasn’t looking for an easier way just a newer way. I’ve tried a C++ consol application but can’t find a way to open and write to a file. The same for VB without doing a windows app which takes me forever. Is there a simple way to do this in C++ or VB?

I’m sure you noticed my Gcode output was completely wrong. It is a constant pitch and has nothing to do with what Andy was looking for. Oops!

Tony

82

General Mach Discussion / Re: worm gear cad design

« on: December 28, 2013, 08:37:33 AM »

Andy never said how many turns the worm was to make in the 24” but attached if a file that will generate a ten turn ¼ to ¾ variable pitch in 24”

You could edit the following BASIC code to generate anything you wanted.


Here’s the BASIC code I used to generate the file:

OPEN "VPITCH.TAP" FOR OUTPUT AS #1

X = 0
VPITCH = .5 / 3600 '=.0001333 pitch increased per degree
INC = 23.5 / 3600  '=.0065  pitch increased per degree
 CLS

 FOR I = 1 TO 3600 STEP 1

X = X + VPITCH + INC

'PRINT #1, "TEST"

PRINT "G01 X "; USING "####.####"; X;
PRINT " A "; I

PRINT #1, "G01 X "; USING "####.####"; X;
PRINT #1, " A "; I


 NEXT I

CLOSE #1

Off subject. Is there any new easy way to write and execute code like this? This BASIC is at least 25 years old.

Thanks

Tony

83

General Mach Discussion / Re: worm gear cad design

« on: December 27, 2013, 02:21:15 PM »

Thanks John, now I see what you mean in incremental mode.

I also have a very poor resolution rotary axis for pipe coping with plasma but it works quite well in absolute coordinates.

 40.2777778 steps per degree.

G01 A .1 gets. .0993
   10 = 10.0055
   20 = 19.9862
        3600 = 3600.0000

Sorry to have gotten off subject.

Tony

84

General Mach Discussion / Re: worm gear cad design

« on: December 27, 2013, 11:59:02 AM »

"The equation to calculate STEPS PER is:
Motor steps per revolution (200) * ratio of timing belt drive (2) * ratio of Rotary table (40)  = Total number pulses to go 360 degrees = 16000
Divide this number (16000) by 360 degrees = STEPS PER 1 degree = 44.44444

You cannot have a fractional STEPS PER except if it is an even fraction equal to micro-steps, like 1/2, 1/4, 1/8 micro-steps, then a fraction like .5, .25, .125 will equate to a whole number when factored with the multiplier of 2,4,8 in the final calculation."

Won't Mach work with the 44.44444 but only stop on a whole number? Like 6 degrees (266.6664) would stop at 267 or maybe 266

85

General Mach Discussion / Re: Signal ground / Earth ground

« on: November 27, 2013, 11:51:29 AM »

Thanks Tweakie,

That is a help. I don't appear to have any noise issues with my plasma machine but do see where I don't have perfect "star" connections and someday should make them better.

Tony

86

General Mach Discussion / Re: other way to turn off torch

« on: November 27, 2013, 08:36:48 AM »

do you have a spindle delay in Config/ Ports and Pins/ Spindle Setup?

87

General Mach Discussion / Signal ground / Earth ground

« on: November 27, 2013, 08:13:03 AM »

Signal ground / Earth ground

In reference to ground loops as a source of “noise”:

Quote from Tweakie.

“Also check for GND loops - in particular the LPT cable shield should not be connected to GND at both the computer end and the machine frame if they are both, in turn, connected to earth etc.”

“Quote
Interesting Tweakie, what about the signal grounds like pins 18-25 on a DB25 connector? can that be grounded at both ends? I hope.

Tony

Hi Tony,

Yes, the 18-25 is common signal GND between the PC and the controller but remember that this is signal GND, it is not Earth.

Tweakie.”

I believe the LPT signal grounds are connected to the computer chaise which is also the -5V and -12V power supply GND which in turn is connected to earth ground through the AC power card. Is this correct?

Is it OK to have the signal grounds be connected to earth ground at both the computer and a controller which is grounded at the machine?

Thanks for any help.

Tony
   

88

General Mach Discussion / Re: computer keep on switching of when the plasma ark(spark)

« on: November 26, 2013, 04:38:33 PM »

Also check for GND loops - in particular the LPT cable shield should not be connected to GND at both the computer end and the machine frame if they are both, in turn, connected to earth etc.

Tweakie.

Interesting Tweakie, what about the signal grounds like pins 18-25 on a DB25 connector? can that be grounded at both ends? I hope.

Tony

89

General Mach Discussion / Re: Backing off homing switches

« on: November 17, 2013, 03:59:09 PM »

I have a problem with auto homing.  I installed a set of npn prox switches. I used the pmdx 126 boart, the 134 mother board, and gecko 203's. I configured everything correctly
and get all the lights for x,y, (z is not being homed)  The machine begins with the Y axis homing and reaches the switch and stops and backs off. Then the x axis homes towards
the switchs, the A axis (slaved axis) hits the switch and backs off, the x- hits the switch, backs off, and continues to ram in a back and forth fashion?    What am I doing wrong?

Maybe

From Gerry: " You need to make sure to UNCHECK "Home Slave with Master" in General Config. You may also need to modify your Ref All script to use RefCombination() to get the two sides to move together."

90

General Mach Discussion / Re: Backing off homing switches

« on: November 17, 2013, 03:42:34 PM »

Another way to deal with crappy limit switches is to increase your debounce time in General Config to maybe 500 units at 40 microseconds per unit equaling a .02 seconds delay. That may be just enough time (distance) to clear the switch. If your slow speed zone speed was 25"/min you would be .008" off the switch.

Tony