Messages

5711

Mach4 General Discussion / Re: Help! Error Message G81 Cycle, Z Value Unspecified Error

« on: December 08, 2017, 01:01:34 PM »

Hi,
I use a software utility that converts PCB artwork into Gcode and it puts a G81 on each line also and that's when I encountered the behaviour you noted.
I was tempted to call it an error, after all the same code ran under Mach3 just fine, it must be a Mach4 bug, but its not. The Mach4 behaviour is closer
to the correct, or at least 'Peter Smid' interpretation.

Craig

5712

Mach4 General Discussion / Re: Help! Error Message G81 Cycle, Z Value Unspecified Error

« on: December 08, 2017, 12:52:50 PM »

Hi,
try this code:

Code: [Select]

n10 g20
n20 g81 x0.067 y2.121 z-0.223 r0.500 f6.0
n30 y3.496
n40 x2.844 y3.993
n50 y2.100
n60 g80
Couple of other interesting things that Mach4 does which can catch you out:
the G code interpreter changes all letters to lowercase, strips out leading zeros and spaces.

So:
G81 X0.067 Y2.121..... becomes
g81x.067y2.121.....etc

In the vast majority of cases it makes no difference but in some cases it does, I found out when Mach couldn't find a macro I had written and named
M03 or similar and Mach interpreted differently and continued using the built in macro because my substitute was incorrectly named. smuprh pointed
this behaviour out. While uppercase, leading zeros and spaces make Gcode more human readable the Gcode interpreter sees it differently, be aware of it!

Craig

5713

Mach4 General Discussion / Re: Help! Error Message G81 Cycle, Z Value Unspecified Error

« on: December 08, 2017, 12:30:37 PM »

Hi,
sorry should be G81 not 83, same principle applies.

Craig

5714

Mach4 General Discussion / Re: Help! Error Message G81 Cycle, Z Value Unspecified Error

« on: December 08, 2017, 12:27:39 PM »

Hi,
I've encountered exactly this 'error' also. G83, in fact most of the canned cycles are modal, so the first G83 requires all the information to
drill the hole, Z depths, retracts...etc. Thereafter you only need put the X,Y co-ordinates and Mach assumes all the remaining info stays unchanged.
Your code will work if the first G83 line is left as is but each subsequent line LEAVE OUT the 'G83'

You might believe, as I did this is an error, its not. In fact this interpretation of a modal canned cycle is correct whereas Mach3's interpretation is/was
sloppy enough to get away with it.

Craig

5715

Mach4 General Discussion / Re: Selected tool not updating from T1 during toolchange

« on: December 08, 2017, 12:36:25 AM »

Hi,
kool, just as a note for others the setting to which I was referring is Configure/Mach/Tools.

Craig

5716

Mach4 General Discussion / Re: How to make Lua wait for signal state to change

« on: December 08, 2017, 12:32:52 AM »

Hi,
have a look at function calls in:
https://www.lua.org/manual/5.3/

Return is the correct statement to explicitly return from a Lua function.

Craig

5717

Mach4 General Discussion / Re: How to make Lua wait for signal state to change

« on: December 07, 2017, 10:23:42 PM »

Hi,
not sure just yet.
You can use BREAK to exit a WHILE loop for instance.

I think the equivalent for breaking out of a Lua function is RETURN.

Will do a little research.

Craig

5718

General Mach Discussion / Re: G16 and G83 interaction (again)

« on: December 07, 2017, 06:05:47 PM »

Hi BR549,
good to hear from you...

Craig

5719

Mach4 General Discussion / Re: Selected tool not updating from T1 during toolchange

« on: December 07, 2017, 06:04:57 PM »

Hi,
not sure that this will help but there is a setting within Mach4 that determines how the m6 is to be interpreted.
In one mode the current T number is the tool actually in the spindle whereas in the other mode the T number
is the next tool to be loaded. This allows the ATC unit to get the tool ready to be put in the machine and save
some time when changing tools.

Look on Configure/Mach/General page from memory.

Craig

5720

General Mach Discussion / Re: Servo drift due to EM noise. Optical encoders not preventing it.

« on: December 06, 2017, 09:30:09 PM »

Hi,
two mechanisms for noise, conducted and radiated.

Conducted noise is often overlooked. It is where the 'hash' impressed on an electrical supply by a device, in your case a spindle,
which then gets conducted to other devices on the same power supply, a servo drive for instance. As an experiment
run your spindle from an extension lead plugged in as far removed (electrically) from the supply for your PC and servos as you
can.

The other mechanism is radiated where the hash generated by a device is radiated to other devices and circuits by
electromagnetic radiation, ie radio waves. There are numerous techniques for reducing the sensitivity of a circuit to
radiated interference, a search of this forum alone will result in hundreds of ideas. Still the best bet is to try to reduce
the hash generated by the spindle/VFD. Line reactors are not popular with hobbyists, you have to pay for them, shock
horror but they are effective.

Craig