Hi Guys:
For those with a PlugIn interest, I figured I'd post some things going on..and the results of current testing..
I have modified the SDK to have one Include folder for the MachIncludes that all plugins will share. I have found some pointer inconsistancies in the .h files and have been fixing them as I go. Shoudlnt be a terrible problem to fix any PlugIns in progress, but Ill re-release the entire SDK when the ModIO plugIn is done.
The ModIO plugin is now running extremely well with Inputs, Im just finishing up MPGs, and config dialogs. I have added multiple ModIO slave capability so in future more than one ModIO may be daisychained together for more IO. Slave 1 is pins 1-8 Slave2 is pins 9-16 ..ect.. This all seems to work well within the constraints of testing the singular ModIO capability at the moment. As each plugin is done, its likely the .h files will have to be updated for past plugins, but we'll fix that as we go.
A user in Brazil is working on a remote control web based addition to allow one mach user at home to run a machine in the wilderness. It sounds like an interesting project , and he's also working on a plugin to allow getting IO from a computer somewhere else as well. Another interesting wrinkle, might be good for remote support type of functions. (Be nice if I could simply take over your Mach3 and see the screen.. probably help quite a few..)
The PlugIns so far are proving to be pretty powerfull to program in. The ModIO is almost flowing together, so Im pretty happy. ModIO will be a good open source example of how to deal with generalized input and MPG inputs from various sources.
After ModIO, which should be done by late week, the biggest challenge starts. It will be the G100 plugIn. Its a major and complete rewrite, open source, of adding a G100 to control Mach3 from the ground up. Now that I understand the G100 enough to program for it, the work begins this weekend on it. My plan is first to clean up the connection and IP attaching proceedures to make it all automatic, then IO functionallity including the encoders, then Jog, and finally movement and homing routines. Expect it to take a while, but I will be releasing versions of the plugin probably weekly for testers. The initial testing of course will be limited to checking your IO, then Jogging, but then the movement will come and we'll be back to where we were, but with the difference that the source code will be available, and others can check my work to see if perhaps they can improve on it, or suggest or add additional functionallity. Its going to be an exciting project from the programming side. This should lead us to a working Mach3/G100 version within a couple months of starting, but with pretty good functionallity by mid July.
All this is truly necessary, and I appologise for the delay it meant to the G100 version of MachIV, but it seems a no-brainer to open it up, so i can get help when I need it, both from Steve, who writes the firmware, and from other C experts out there who may be able to see my logistical errors as I go. The ModIO Plug is well documented, and this doc'ing will follow to the G100, in hopes of making it bulletproof in a much faster time frame than before. Brian will be helping me with testing and code at the same time, so the code should fly faster than before as well.
The new ModIO Plug has a 40hz refresh rate internally, and seems very much faster than the macropump could ever be, this makes a big difference in performace. My Modio is running as I type this , at 40hz refresh on all registers, input and output. My CPU load on screen is currently 15% and it doesnt change if ModBus is on or off. Those who have used the old ModBus, know that 40hz refresh on the ModIO could jack up system usage as high as 40% , as I said, a huge reduction is systemic load on your system.
We've noticed this of course, could impact on many macropump type of operations. Rons encoder interface for example, would make an easy plugin, that would raise its performance to near closed loop speeds. (but not quite. :) ). It seems this new technology will allow Mach to do many things it could never do before in terms of IO devices. A drop from 40% CPU load to 15% (which is normal non-Modio load level) is pretty staggering statistically from a running ModIO card. I suspect we're only touching on the true capability of PlugIn vs MacroPump. I can see eventually a generalised Toolchanger PlugIn that will operate much more robustly, and Ill give that some thought as the G100 is done.
Anyway, all signs are that we're on the right path for drastically increased reliability, and many more functional items over the next while that were previously impossible. I thought Id fill you in on what Im seeing so far. Im amazed and very pleased at the prospects this is opening up..
Thanks,
Art
www.artofcnc.ca
Videos And Support Forums
Users Map:
