Project

General

Profile

Getting there - Homing problems now...

Added by drogon over 7 years ago

So after being sent the wrong SD card holder I've now removed it and fitted a full-size one and gone through as much testing, setup as I can.

My HPC3020 seems different from the others mentioned here - 4 JST connectors - X, Y motors, Laser power and the end stops. The X & Y ones are identical in all respects other than the white & yellow coil wires swapped on the X axis.

The laser on/off seems the same as the diagrams.

The end-stops are 3 wires into a 5-way JST on the HPC board, have connected them into the LaOS board in what I think is the right way - however they appear inverted. When running the test program, the inputs read 0 when the end-stops are reached. I set the inverted flag in config.txt.

However when I execute the home operation on the LCD, it keeps on going and doesn't stop even when it hits the end-stops, so I've no idea what to do next. If I change the polarity in the config file then it does nothing on a home operation as it thinks it's already at home - unless it's really at home when it drive both motors against the mechanical stops until I hit reset.

However it does do something - I have been able to download one of the tests via visicut into it and actually start the laser cutting (well scoring as I had the power cranked right down)

So almost there - why aren't the end stops working )-:

Does the config file polarity actually work?

Do I need to wire them the other way? Right now, I have the common from the end-stop switches going to 0v with the Vsensor jumper (J43) going to +5v. The limit switches are normally open, so the opto isolator outputs will read high until the limit switch trips when the LED will light, the transistor will conduct and the output will be 0v.

So am I missing some magic in the config file?

Posted here for reference: http://unicorn.drogon.net/hpc.config.txt

Clues welcome..

-Gordon


Replies (7)

RE: NOT Getting there - Homing problems now... - Added by drogon over 7 years ago

Really frustrated now.

I've re-written the config file according to the page on the wiki as opposed to the one I downloaded.

I changed the motion.enable to 1 which turns off the steppers.

I did a home operation on the controller and manually moved the head. The controller returned to the menu and printed "Home done" on the console as soon as it hit both end-stops.

Changed the config to enable the motors.

Did it again - and it fails to read the end-stops and I have to hit reset on the MBED.

If I manually jog it to the home (top right) position then do the home operation, it returns and says "Home done" on the serial port.
If I manually jog it to the opposite corner then do the home operation, it moves towards the top right then doesn't stop and overruns.

So the end-stops work, they just don't work when the motors are moving.

Why?
What's going on?
What's going wrong?

So frustrating.

Other frustrations:

The head moves in the correct direction when going home, but the joystick Y axis is the wrong way round. X is fine. Obviously no correlation between the joystick
and the config file )-:

Lack of support. Where is the community? HPC have abandoned me - where is the support here? Is no-one interested in the HPC I have which appears to be different to all others?

Thought I could be clever by manually moving it to the home position until I hear the limit switches click - then execute home on the panel. Seemed to work. Then used VisiCut to load in a design, then tried to cut it. Head zoomed off to the bottom left corner and hit the endstops. I'm getting good at hitting the reset button now.

What does "Origin" on the display do?
What does "Referrence point" in VisiCut do?

-Gordon

RE: Getting there - Homing problems now... - Added by peteruithoven over 7 years ago

I've notices that the endstop polarity setting did work at one time. Maybe do a quick check with a multimeter?

(Don't use the jogger, you shouldn't need it and it poorly implemented at the moment)
(The joystick is indeed reversed, not sure why) (Maybe make a issue about it?)

Not sure what Origin on the dispay does.
In visicut, using this "Referrence point" you can change the 0,0 position. Which is hard to explain in text.

RE: Getting there - Homing problems now... - Added by drogon over 7 years ago

Well I've got there now, thanks. Broken wire on the HPC cable.

I used the jogger to work out the bed size (and where I want 0,0 to be) and entered that into the config file.

The Origin command on the controller - it causes subsequent move command (joystick) to just move on it's own. Who knows what that's doing.

And another frustrating hour or 2 and I've worked out visicut too. At least I think I have.

The last thing I need to do is enable PWM and since my HPC appears unique, I guess I'm on my own there too.

Oh, the [cancel] function doesn't work on the display either. Good job that reset button is handy.

Shame the LaOS board is a different size to the one inside the HPC - maybe an exercise in cutting another panel to make it fit is needed. Ah well.

-Gordon

RE: Getting there - Homing problems now... - Added by depronman over 7 years ago

drogon wrote:

Well I've got there now, thanks. Broken wire on the HPC cable.

I used the jogger to work out the bed size (and where I want 0,0 to be) and entered that into the config file.

The Origin command on the controller - it causes subsequent move command (joystick) to just move on it's own. Who knows what that's doing.

And another frustrating hour or 2 and I've worked out visicut too. At least I think I have.

The last thing I need to do is enable PWM and since my HPC appears unique, I guess I'm on my own there too.

Oh, the [cancel] function doesn't work on the display either. Good job that reset button is handy.

Shame the LaOS board is a different size to the one inside the HPC - maybe an exercise in cutting another panel to make it fit is needed. Ah well.

-Gordon

Hi Gordon,
I have very recently converted a K40 laser cutter to LaOS Rev 5 board.
On mine if I set the 'origin' using the control panel, then move in an direction using the joystick it crashes in to the top of the machine.
The Y axis is the wrong way around on the joystick (X is correct)
The cancel does not work - due to the panel being disabled once cutting is started (was causing errors in the cutting, so was disabled once cutting starts and re-enabled once cutting is finished)

Once I got everything working correctly I have never needed to abort a cutting job, so the lack of cancel is not a major issue, but it would be handy if you could cancel the job say when you are trying the cut with the laser off, as currently you have to let the job finish which can take some time. When I had the moshi controller board about 1 in 20 jobs would cause the laser to crash in to the machine side (if you didn't switch it off before it made contact :)) so LoAS is a MAJOR improvement.

The LaOS Rev 5 board was slightly larger than the moshi board that I removed, so had a few holes to drill in the mounting plate, then used some plastic standoffs used in desktop PC's to mount the LaOS board in the same place as the moshi board was.

I have a DXF file for a new mounting plate, to accommodate the display, buttons and all the original switches and buttons, I cut mine from 3mm MDF. Let me know if you want the file, may save you some time even if you just use the 'elements' from the DXF.

I'm using corel draw to generate the image to be cut, exporting as a DXF file, then import this into Inkscape (doing nothing extra with the file), then use the link in Inkscape to send it directly to Visicut.
Once in Visicut I then select the appropriate material and thickness, set the position on the bed where the image is to be cut and execute it. Finally push the joystick to start the cutter running.

I got pwm to work and included the original power knob, so that I can override the visicut power settings if required - works good. I can send you the info of what I connected to what if it will help (not sure if your m/c has a similar laser psu to mine)

BTW I'm using a windows 7 laptop to 'drive' my cutter.

I've not had any luck getting 'visicam' to work, would be interested if you get it to work.

Cheers,
Paul (Depronman)

RE: Getting there - Homing problems now... - Added by drogon over 7 years ago

Thanks for your reply - so Looks like the issues I'm getting are well known - origin/cancel, etc.

My LaOS v5 board is smaller than the board I took out the HPC, but 2 of the mounting holes work. I'll move the others at some point.
I'll need to fit a grommet in the case where the old USB cable went to make sure the Ethernet cable is going to be OK in the long-run.

All my stuff is Linux - my primary reason for wanting the LaOS in the first place - there simply isn't any alternative - other than to be a Windows user and pay an extra £1000 for the updated software from HPC - simply not an option for me. (I confirmed on the phone that the hardware was the same, so I was happy there)

I'm getting to grips with it now, but haven't hooked up the PWM yet. Would be interested in your method if it allows the potentiometer to work too - or act as an overall maximum setting with the software set to 100%... I think my Laser PSU is fairly standard, but can't get the manual for it. There is also a 2nd PSU in the case with many connections. I've not checked them yet. (2 are the +24v and the +5v for the motors & control card though) Will buzz them out a some point.

visicut is OK - it's crashed a few times and the original author isn't replying to email )-: I'm using Inkscape, saving to disk, then using visicut to open the saved file rather than launch visicut directly from Inkscape. Seems to be slightly more reliable that way.

I'd like to be able to manually jog the head to a location on the bed, then push start with the design starting at the laser position rather than where it is on the visicut page - it would make making many of one thing easier, or using some scraps to cut a small piece out of an unused area.

Anyway, I think I've now gotten to grips with it and it'll do for now.

Cheers,

-Gordon

RE: Getting there - Homing problems now... - Added by depronman over 7 years ago

I was pointed in the direction of visicam as an alternative solution to jogging to a position then starting the laser from this position.
In theory visicam would take a picture of the laser cutters bed (with the partly used stock showing in the picture), you would then drag the job in visicut to an area where sufficient stock was available, then start the cutting in visicut.

This is all fine and good in theory, but in practice I can't get visicam to work under windows, neither can I get any help in doing so.
I've not given up on the visicam solution, but for the moment it alludes Me.

If the Jog and Origin feature worked correctly then visicam would be less of an issue.
Mail me directly (paul at whittlefamily.co.uk) and I'll sketch out the pwm wiring solution I used. I does indeed allow you to vary the actual power the laser o/p's by increasing or decreasing the knob value, mine is set to 10 as the default, but I can reduce to 9, 8, 7 etc or increase to 11, 12, 13 etc (you obviously can't increase the lasers power beyond maximum power :)) The laser test and laser on/off buttons also still work

My K40 also has two power supplier, the laser power supply (larger of the two) and a PSU that reduces mains voltage to 24V and 5Volts (well I think it provides 5 volts)

Cheers,
Paul

RE: Getting there - Homing problems now... - Added by jegb almost 6 years ago

The joystick issue:
-Logically, the + location is up on the joystick. If your center is top left, the +Y is moving downwards, which is odd, but works as specified. There could be a flag variable on the config file that reverses the UI control for jogging Y.

The homing cycle discrepancy with 0,0 coordinates:
-When I home, the 0,0 is on top left.
-When visicut runs, 0,0 is on bottom right and then rest position is top right.

I have to check the config.txt:
-scale is +ve on X and -ve on Y
-Min, Max are both set to 0,200mm each axis, rest on 10,10mm

Not sure what is wrong, but requires some thinking...

    (1-7/7)