Machsupport Forum
Mach Discussion => Mach4 General Discussion => Topic started by: Mauri on January 12, 2017, 07:54:36 PM
-
hi,
I am having major differences between CV with Mach4 simulator and CV with Mach4 and Controller on the Toolpath Display and Timings to completion.
I have tried CV = 0/10/25/50/100 with the Mach4 and Controller running and have viewed each time the Toolpath Display and reported the time taken.
I have compared these against CV with Mach4 Simulation Mode on the Toolpath Display and recorded time taken to complete.
I have found that all CV’s are faster to completion with Mach4 and Controller, but are visually worse even comparing CV=0 with Controller and CV=100 with Mach4 Simulation mode on the Toolpath Display.
Could it be that the data is flowing faster to the Controller than the Toolpath Display and hence loosing data for display on the angles?
If not can anyone explain why?
I have many examples if some one needs them to explain.
Regards,
Mauri.
-
Data is flowing to the controller first, then we have to wait on a position update from the controller to update the position on the tool path. We depend on the motion controller to give us updated positions and since there is nothing real-time about it, there will be positional "gaps". So it really depends on how often the motion controller sends Mach position data. No two motion plugins are the same. For instance, this update rate is configurable in the Galil plugin and can go as low as 8ms if the PC will handle it. But a lot of them may be as slow as 100ms. A slower update rate will have the effect of chopping off a corner in the tool path display as Mach basically draws a line from the last reported position to the current one.
Steve
-
Steve,
Thank you for the information, I had a feeling that this was happening.
On my Controller Plugin the Polling Frequency is 10Hz to 250Hz and I have the setting at 100Hz as recommended with the Controller information.
Previous versions of the Polling Frequency had 2Hz to 250Hz.
It talked about Significant Network Traffic and CPU load, however if my PC is fast and directly connected it should not be an issue.
I will enquire with the manufacturer of the card more about this setting.
At least now I can calculate a cut time that was close to Mach3 with its CV setting (CV Distance Tol = 0.01 and Stop CV on angles >49), which equates to about 75 on all CV 180 angles with Mach4 and this reports a reasonable good Toolpath finish using 30Degree 0.05mm Radius cutter on a 2.5D surface.
Thanks Steve,
Regards,
Mauri.
-
Steve,
I have been advised that HiCON Integra Plugin can provide faster updates.
From Marc of VitalSystem a very helpfull Company with great CNC products.
Hz, or Hertz, is a unit for frequency (cycles/second).
To get the Poll interval:
PollInterval = 1000 / UpdateFrequency
This means, with an update frequency of 200Hz, you will effectively have 5ms update intervals.
I would really only recommend going up to 200Hz which is probably more than enough responsiveness as far as PC-based motion control goes. There are times when even more responsiveness is required, which is why we have the HiCON and DSPMC macros.
So since I had 100Hz that would have been equal to 10ms.
I will give the 200Hz a go and see what happens.
Regards,
Mauri.
-
Hello Mauri,
I've read some of your posts and am impressed with you understanding and knowledge, particularly with the communication of M4 to controller. I'm hoping you can help.
I'm deciding between Galil DMC-4080 and the HiCON for my industrial router with nema42 steppers. Since 2010 I've been on Smootstepper ESS, It's been a punishing 7 years.
Moving to M4, but am still on M3, command speed varies motion rate greatly. This is most evident on CV moves. The SS has something they call “controller frequency”, it’s adjustable and has a max of 4khz. This controller frequency directly affects Mach’s CV (constant velocity) feedrate output.
Motor tuning aside, an example is on toolpaths with thousands of tiny 0.002” moves of 5 degrees or less (at F140) a controller freq of 1khz produces an actual feedrate of appx 10 ipm.
Bump the SS controller freq to 4khz and feedrate output jumps to appx 60 ipm with no added jerk at the machine. Motor velocity and acceleration settings and performance are unaffected, just commands are occurring at a higher rate providing an increased feedrate output. The SS plugin crash rate dramatically increases at 4khz.
Better gCode is a solution, but the M3/SS bottleneck exists even with direct cat5 connection of fast computer to SS. The SS plugin and/or controller are to blame*.
My big question is… do you think the HiCON (with M4) will outperform the SS with regard to processed commands per second?
Any of your thoughts are deeply appreciated. Thanks!
-Haik
*SS controller freq, number of lines to precalculate and M3 look ahead lines, CV settings, motor tuning and more have all been thoroughly exercised.
-
I would not use a Galil to control steppers.
Steve
-
There are just so few choices for quality motion upgrades. Vital Systems and Galil take what they do seriously. The "hobby" approach and perpetual beta of others have had huge costs for me. I'm getting closer to just buying a new Biesse or Thermwood system. My old Therwood C5 (w/atc) is mechanically upgraded and is very tight.
It would be great to have a router with 3ph AC servos, I'm just not looking forward to the service contracts and s/w addon costs.
I am awaiting some info from Vital Systems. I've asked about Spartan3 FPGA cababilities on this topic.
This is my last hope for reaching a quality bar that includes stability and some resemblance of performance.
I just looked at several of your other posts, there's a lot of depth and detail that is also helpful. Thanks for the reply!
-
Haik,
HiCON controller was basically the first that would run properly with Mach4, It was used with Mach3 very successfully.
Our HiCON has had no issues with running Mach4 and if any come up with the Mach4 updates the people at HiCON quickly provided an update to the firmware and plugins.
They have a number of add-on cards if required to hook up any configuration of equipment motors and drivers.
They work with servos or steppers.
They may cost a little more but I am sure you will be happy.
Remember that good motors and driver are also the heart of smooth motion.
Make sure if you go ahead and purchase one that you gather all the information you need to make your controller.
Most of this is on their website, so download it and print it and read it.
Regards,
Mauri.
-
Thanks for the reply!
If I may ask... What other controllers run Mach4 properly (up to $5k, cost is not a factor)?
Kind Regards,
- Haik
-
Haik,
All controllers out there are continually making changes to suit the newest version of Mach4.
HiCON has been the one that has worked out of the box.
In the past I used Smooth Stepper, but have always had some issues.
PMDX may be worth an enquiry.
Regards,
Mauri.
-
OK... Thanks for reply'ing.
Yes, PMDX is a consideration for USB products and BOBs, etc. Steve Stallings and PMDX do make the best BOB I've ever seen and used. I highly recommend their PMDX-126 board.
The PMDX USB product is doing well for many and in talking to Steve it's too much of a gamble that USB will maintain connectivity in an industrial environment, it's just fine for 3D printers though. I've been through this myself in the past and the Mach Forum is packed with users having connection/pipe/handshake issues going well beyond what a ferrite bead can fix.
Keeping in mind the USB standard (as per USB-IF at usb.org) is not suitable for communications in harsh or industrial environments.
A glaring example is the non-locking connector implemented in USB cabling.
Separately, PMDX has repeatably promised an ethernet Smart-BOB controller and after waiting for it for 2 years "it ain't gonna happen." :(
Kind Regards and thanks again for helping out!
- Haik
PS- there's a good chance I'll get the HiCON. Indications are that Vital Systems has high regard for standards and best practices. In engineering these things are much more than corporate-speak.
-
To other folks that might read this...
The Warp9 ESS is cruel joke wherever a rotating tool is involved, hardware product possibilities quickly vaporize as the firmware and plugin development fail egregiously. An example is rotating the toolpath in the M3 toolpath display while running a (gCode) program crashes the plugin almost every time.
Last year while resuming M3 from an ESS crash their plugin sent corrupt buffer data resulting in a runaway Z axis... driving my HSD Spindle into my 4th axis costing me over $5000 in spindle replacement. I'm lucky I wasn't injured.
I hope others read this and understand that just losing time dealing with home-brew motion control is only the lighter side of the costs. Tool, stock and machine damage constitute another realm of loses. Worst loses are safety and health.
-
Be careful, don't blame what might be a PC issue on ESS. That tool path will consume lots of CPU when rotating it in Mach3 and cut the flow of data off to the ESS. It depends a lot on the CPU and graphics card. I can cause the same issue when running Mach 3 on my Matsuura mill with a Galil (it has an Atom CPU). Mach 4 runs the tool path differently and it is much more efficient. Not an issue even on the Atom board.
From what you have described, Mach 4 will take care of most of your issues. I also have a machine with ESS and Mach4 and it flies. No problems at all. It just runs and runs. I have a HiCON too and I like it. IMHO, ESS and HiCON are both good devices. They do about the same thing, which is run steppers or position controlled servo drives.
Steve