Advanced Search

Nice! that link is helpful. So to summarize the next steps for a very basic test setup with no mill: 

• Ill need a older PC with a native parrellport
• Load a copy of linuxcnc and Probebasic (ill follow some youtube) 
•  parrellport cable to breakout board (i can prob find soemthing basic on amazon, if that goes well later upgrade to Mesa)Seem thats incorrect and need something a bit more robust like a "7I96 STEP/IO Step & dir plus I/O card"
  • I/O OUT  --> to drive steppers via commercial driver like a stepper online DM542T
  • I/O IN --> to read my two X and Y encoders. These are typically two pins so it can sens direction, so 4 needed Xa Xb and Ya Yb from the encoders. maybe even a Za and Zb if i want to move the Z axis.    
  • Some config file or other tweak to get the "encoder.N.counter-mode" to work from that link...
    
    Probably a little stuck on the last bullet. Not sure what to google. Am i on the right track?
    Its my first post so sorry if this is old news. 
    
    

Hello David
Thanks for writing. As far as I know the hydraulic tool changers were both working until I interfered with them. Now the horizontal one is not working and been like that fo two years. I suspect the springs are broken but I don't know haw to pull the unit apart. There is hydraulic pressure in the lines but maybe the lines need bleeding. I cant see how best to do that.
I need to send you some photos. I tried once before (years ago) and failed.
I am going to be away for a few days. Promise some action on my return
Cheers Chris McMullen
Auckland New Zealand

There is a guy who sells flexihal boards in aus.....i dont know what his site is, but he goes by Jaymis

Some of Mesa's cards have dedicated inputs for hand-wheel encoders.
But this is so slow counting that you can use normal pins and the encoder component to get the hand-wheels inputs.
"normal pins" means parallel port pins, mesa input pins or some other digital io in pin.
linuxcnc.org/docs/html/man/man9/encoder.9.html

This is a weird-ish question, but LinuxCNC is so vast, I thought I'd ask:

I have a build where I'm making a trunnion inside the same area where the XYZ axis operates. Now that trunnion is growing up to have parts attached to it...

Is there a way to define a "no go zone" within the XYZ build box so that I don't accidentally crash into it?

It would need several different sections, ie, a series of boxes could be defined within which the controller is supposed to know that's not OK and send that 'move would exceed limitations' message if I tried to violate it.

I'm thinking the answer is "with great complexity comes great responsibility..."

But I thought I'd ask.

Cheers. 

Hi all,

* Note, the topic name was changed up to 2 times by someone, haha.

Today i added look ahead. This took almost half a day. It's a tiny file, but took a lot off time.
Here is the look_ahead_source_code_file

Now i did a test run tonight using a line fillet algo :  file_path_standard_subseg_line_fillet
This is a simple optimizer algo, that is handy for simple runtests.

The test is done on the extreme  spiral_on_surface file. With tiny segments 0.001mm and deviation P0.0001 Q0.0

Some hal values:

• incements_a_cycle = 2   (planner skips extra gcode segment to keep up with scurve new position)
• ring buffer size = 10
• total loaded segments from program start = 15117
• future buffer, items that are ready to perform, including fillet segments = 3
• cycle dist, is the distance in mm for current servo cycle = 0.0078 mm
• hal traject buffer dist, total todo segments length. ( sum off future buffer ) = 0.1544764 mm

This file run's quite nice for what it is.

 

Git sources are updated.
The first donation is recieved. Thanks ! See the github readme file for more info !

Whoops. Big error on my part!
It's easy to forget stuff at my age. I should have checked stuff before I spouted off..
The missing pressure switch in the hydraulic in the photo is 3F1. It runs the pump directly. Otherwise the pump works as mentioned.
Contactor 5K1, which operates the tool unclamp, solenoid is part of a complex set of interlocks that stop you releasing the tool while the machine is running and stops the machine running if the tool is being changed. It also defeats the gear change jogging.
I'm not sure exactly how it works if you have the Philips controller, but I think that if you pulled any of that wiring by removing the controller the tool unclamp wont work and probably also the auto tool changer.
A lot of the interlocking is relay logic so it's probably possible to sort through the logic and strap it where the controller stuff no longer counts.
If I had an auto tool change on mine it would have been done completely on the back plane of the Heidenhain controller. But even then the relay logic would be involved and 5K1.
A key starting point is: does the hydraulic pressure pump run? It's noisy on mine. It usually runs for about 5 seconds on mine if the machine has been off for a while. On mine the light in the power button illuminates if the hydraulic pump isn't running. It's directly wired to the other contact in the 3F1 pressure switch. Seems an odd way to do it. And the bulb sometimes blows from the solenoid kick.
You will have to change gears using a spanner. On mine the half inch drive extension of a socket set works. The manual has a chart of what the coloured dots on the selector cams mean.
I think I put the logic drawing of this up earlier. It's so complex that it took me days to understand. And as you can see I often get confused by it. I went so far as to draw it in each possible state by erasing the other stuff. And even then I don't know if I got many of the possibilities.
A simple test might be to just press on the tickler on 5K1 contactor and see if the tool unclamps? Make sure to catch the tool if one is fitted.
 

As the message says it's not really an error, I think the only way to stop the message would be to modify the hal_parport code, or I not sure if this would work but add a debug level when loading the driver. First would work but means compiling Linuxcnc one's self the second option is just a guess.
If the machine works I wouldn't bother too much about it. Yeah it is a little annoying. But then again you have the convenience of CNC

USB to parallel port not supported at all, don't an option.
Any other USB motion control are not supported.
All supported Mesa boards can be found on their website. You can also browse the forum to see what boards are popular at the moment.
For the DIY route Ollie's Linuxcnc- RIO project is fantastic, well supported and in active development as is Scott's Remora Project.
I've also been working on a setup that uses FPGA Dev board available via AliExpress that runs Mesa's hostmot2 firmware, ATM it only uses SPI interface for the Raspberry Pi or EPP for the PC platform, tho I'm currently working on an ethernet plugin board.
In all honesty if it wasn't so expensive to get items from the US to Australia I'd go with Mesa in a heartbeat. What board just requires at little time researching, coming up with a couple of choices then asking a few questions, the 7i76 boards pretty much have it all.

Seems like an issue with languages/locales, or missing translation to Russian.
Try to set locale to UTF-8, see if that helps, if not, then short of installing it again in English i am out of ideas.
I did test on my Laptops, Axis works, even the parallel port config works despite not having a parallel port.

You can test everything, even run machines, without installing LinuxCNC, just boot from USB in Live mode and run the sim configurations included, or use another USB or the internal HDD if you want to actually run the machine, again without installing anything.

I am still working with version 2.8.4. I wanted to test what version 2.9.4 can do better and installed it in parallel.

The good news is im open to hardware options!

PC and Parallel - I have an older PC with the Parallel port that i could use... (I read that a USB to 25pin is no good for latency, so not going that route, this has the port!) 

Mesa - I know of it. seems good if the setup is more straightforward. Looks like they run about $100 to $300. Do you happen to know a PN or feature so i can look that up?

Cheap stuff maybe just to test on? there is also some USB styles... but i dont think they place nice with Linuxcnc?
www.amazon.com/Controller-100Khz-Breakou...08BJTB1GQ/ref=sr_1_6

What kind of io hardware are you using? Parallel port, Mesa cards, or something else? For example many Mesa io boards offer settings to use some of the regular inputs as hand wheel encoder inputs.

Yes --> "hand wheel inputs with encoders to "manually" move your stepmotors" via linuxcnc. and if it misses a step or two from my hand_encoder to the linuxcnc/DRO, no big deal as me (the human) will stop turning when the DRO shows the number im after. I get there may be backlash depending on the ball leads I use. question for another post. Unless its a deal breaker now.

for this one --> "adding encoder feedback to your axis to feed back the position" my plan is to use steppers with motor_encoders for feedback to the stepper driver. So closed loop driver-to-stepper but open loop to Linuxcnc to driver. I may be off here.

so silly question, but if i were to turn the X and Y at the same time, it will do a 45deg ish travel? they are both non blocking?