Messages

7501

General Mach Discussion / Re: new computer can't jog faster than 50 ipm?

« on: November 27, 2015, 02:41:54 AM »

Hi all,
my take on the timing variation is

When a stepper is running at max speed, lets say 1200 rpm at 500 step per rev (I use Vexta 5-phase steppers) corresponds to 10,000 pulses per second.
If the pulse generator (ie the PC) is stable the period between rising edges is 0.1ms. If the PC isn't doing as well then the period between pulses will vary.
If we assume just 10% variation the periods will vary from 0.09ms to 0.11ms which corresponds to a variation in speed of 1333rpm and 1090rpm.
Further these speed variations can occur in rapid succession and constitute VERY high accelerations/ deaccelerations and is no wonder the steppers can't
keep up.
My 'big' scope, while old, is quite good and has a histogram function which allows me to measure % variation on a nominally constant pulse stream. I have found
that more than 2-3% severely limits high speed running. At much lower cutting speeds variations of 10 and more % can be tolerated.
I don't believe that the PP, breakout board or stepper controller have that much influence provided they work of course!
As comments in this thread and others make clear an external motion controller is really the only way to eliminate pulse speed variations. I use a single board
Atom based PC for my controller and it works fine so I haven't bothered with an external controller.
I also use an embedded OS (Windows 7) which allows me the choice of what modules run in the machine. I don't run any Ethernet or Ineternet capable modules
as they have security checking processes which 'clogs' things up when trying to run Mach3.

Craig

7502

Third party software and hardware support forums. / Re: Yet another drive compatibility thread. Maxon 4-Q-EC DES 50/5

« on: September 05, 2015, 09:56:43 PM »

Hi,
looking at the manual suggests to me that you could control these units from mach but it may not be straight forward.
I think that you could use Modbus to communicate to the drives over RS232. It would require you to write Modbus routines
that the DES can read. It would be a challenge.

Craig

7503

General Mach Discussion / Re: Missing program

« on: August 05, 2015, 07:43:26 AM »

Hi All,
Keith is right, there is at least one realtime Linux distro, the one I am thinking of is Ubuntu Realtime. As Keith says it sticks to the business at hand,
if you nominate your CNC software as business that's what it does. Despite its billing its not in fact a real 'realtime' OS.

The genuine contenders for realtime OS have to guarantee a response time and there is a very specific technical standard to which it must comply
and is required for any number of important systems; medical, automotive, aerospace.... If you search FreeRTOS you will find more info.
Yes as the name implies you can have a genuine realtime OS for free!!! Downside is at least from our point of view is that it doesn't apply
to PC type OS's. If you want a realtime scheduler for a microcontroller running the ABS system in a car, one that has documented garantees
you're in luck.

Craig

7504

General Mach Discussion / Re: feedrate in feedbackloop

« on: August 04, 2015, 03:31:06 AM »

Hi BR549,
thanks for your reply.
The spindle has no indexer built in but several ideas occurred to me that might work.
For the rpm involved, less than 5000 say, it might be possible to drive a small motor by belt from the shaft above the clamp nut
as a crude tacho. Might also be feasible to use an optical counter such as RC guys use.
The last idea I had was that as I recall it was possible for the VFD to measure slip as a near to linear approximation to demand torque.
Either way I think it possible to reasonably determine rpm and or torque which would be sufficient for my purpose.

After I first posted I started looking at the OEM DRO's and buttons and noted quite a few which could offer some control over federate
under software control. As an aside it is a bad habit of mine to start thinking after I've already opened my mouth.
Your reply suggests that it may to reduce/stop feed live rather than delayed by the buffer and that interests me. My guess is that it
would be necessary to reduce the federate within milliseconds to avoid a stall with the rotational inertia of the spindle being the only
(time) saving grace.

Thanks again, I will RTFM. If anyone has any ideas to focus my search please post...

Craig

7505

General Mach Discussion / feedrate in feedbackloop

« on: August 03, 2015, 07:41:39 AM »

Hi,
I have just started to experiment milling steel with my homemade mill. It is equipped with a low torque high speed spindle
which is far from ideal for high torque requirements of steel milling. On two occasions over the weekend I broke two carbide
tools because I overloaded the spindle, it stalled and the axis overdrove the tool and 'ping'.

I wondered if it might be possible to have some sort of feedrate override so that when the spindle is near stalling (sensed either
speed lower than critical or high current) the federate could be automatically reduced in sufficient time to save the tool.

Some of the imediate questions that occur to me are:
1)When a line of code is in the instruction buffer is it too late to alter the federate at which it occurs?
2)If the tool load reduces and the feedrate is reset to the programed federate would that constitute instant acceleration and stall an axis?
3)If some sort of feedrate control in a feedback loop is not possible could a 'federate halt' be signalled and would it be quick enuf to
prevent a 'ping'

I'm trying to avoid an Estop event as it seems that several instructions get lost and have had difficulty in restarting at the right point,
this maybe my own inexpertise but suspect that a software triggered Estop may still be too slow to prevent the 'ping'.

Craig

7506

General Mach Discussion / Re: Missing program

« on: August 03, 2015, 03:13:26 AM »

Hi Keith,
CNC'ing has proven to be a long learning curve, first mechanical design and calcs to achieve your goals, Mach3 or
what have you as controller, motherboard/CPU/GPU and OS to run it sweetly, CAD/CAM to have decent code to run...

I built my machine initially to make PCB's but always intended that it should be able to mill steel. I made a lot of choices
along the way in order to do this, cast iron beds, steel table/column, ground ballscrews, preloaded linear rails, low backlash
planetaires with 5-phase steppers.

Have been making PCB's, aluminium and brass parts for a while now. The weekend past I tried my first steel (1045, 160HBn)
parts. It worked but NOT easy, two broken one blunted carbide tools. What I had thought rigid even overbuilt has proved to be
otherwise and my luvly German made high speed spindle looks very marginal when cutting steel. Suddenly things that I never
really had to bother with in aluminium have become important; surface speed, chipload, tool torque, tool deflection, tool wear
and the list goes on.

As I say CNC is just one long learning curve!!!

Craig

7507

Mach3 under Vista / Re: Difficulties in testing gcode strings off line

« on: August 02, 2015, 04:17:14 AM »

Hi  Stein,
when you say it doesn't work what exactly do you mean? It just sounded a bit like the difference between 'run mode' and
'simulation mode'.
I don't have a motion controller but I use my indoor computer for simulation no trouble.

Craig

7508

Mach3 under Vista / Re: Windows 7 -Mach 3 freeze

« on: August 02, 2015, 04:09:21 AM »

Hi,
the CAM program I use throws in a line or two of code which locks my machine up. Your problem may
be similar. In particular the CAM post puts a G91 with a G28 when it is looking for a safe place for me to
change tools and again just before it exits a part program, ie looking for a safe place to park.
Have a look at the code either at the line or several lines preceeding the point where it locks up.
I haven't worked out yet why it happens or what the post thinks its trying to do. I just scan thru the
code at each tool change and at the end of the program and 'comment out' the offending lines before
running it. One thing I have noticed is that if you simulate the program run as I often do then the
danger points become clear as they lock up in simulate mode as well.
Hope that helps.

Craig

7509

General Mach Discussion / Re: Missing program

« on: August 02, 2015, 03:24:53 AM »

Hi Keith,
Microsoft has released a number of embedded OS's over the years, including various compact versions. We had an automotive diagnostic/ scope
at work which was XPCe. Except for a few artefacts you couldn't tell you were operating a Windows system.
It is certainly true that a stripped down OS has a better chance of deterministic real time performance Windows has to my knowledge never
achieved it. By deterministic I mean that the OS is garanteed to respond to an event, usually an interrupt, within a certain time. Windows as it stands
can't do it.
The Mach3 pulse engine is a timer that requires service every 40us (25 kHz). If your computer is running well it will respond with 2-3us. If however another
thread/process/application has a higher priority interrupt service in progress the Mach3 starves and your machine halts. Try a viewing axis change
on the toolpath screen when running a part program and chances are you will stall the machine. The graphics processing means that the pulse engine
can't get serviced ergo Mach3 stops. When the redraw is complete and the pulse engine gets service its too late, your axis can't accelerate instantly.
The pulse engine is assigned the highest priority a application can have but Windows reserves all the highest priority levels for itself, no way around it.
I consider it a miracle that Art Fenerty was able to achieve what we have and blame none that we can't have what we want.
There are a couple of genuine real time operating systems out there, one which I think is very clever is RTX.
It runs a realtime scheduler in one (or more) cores and runs Windows in another. Whatever happens in the Windows core doesn't affect the realtime
scheduler. The Microsoft Embedded Manager in Melbourne is very enthusiastic about it, and well she might, it was going to cost $16000 to get started!!!
An external motion controller like a smoothstepper sounds very much more attractive by comparison.

Craig

7510

General Mach Discussion / Re: Missing program

« on: August 01, 2015, 06:39:06 AM »

Hi Keith,
I imagine NZ supplier would be happy to supply, they are Jenlogix. I can't find on their site any mention of the Unigens product but
that is what they sold me probably as its cheaper, quite a bit cheaper, than their more usual Avantech products.
Advantech has a supplier in Aus as well, search Advantech Austrailia.

The Windows 7 Embedded Standard (WES7) runtime licence is no trivial matter. Officially if you wish to use an embedded MS
product then you have to sign an agreement with Microsoft. This makes you an OEM supplier. You get a runtime code which you
install on every unit you make and buy a little sticker for each unit they call 'certificate of authenticity'. You buy them either singly
or in blocks as you require.

In truth Jenlogix had no right to supply me with either the runtime code or the sticker, but they did. I did pay for it and it must have
been bought originally from MS, it is a holographic thingy. I have only the one machine running WES7 so I don't feel that I have
abused the technical oversight. One could of course with the one runtime code make as many systems as one desired which is
why MS are so protective of their Embedded products.

It is my belief that it is WES7 that contributes as much to the stability of the platform as the crispy new hardware.

Craig