Jump in X-axis when engraving

Added by AmmO almost 9 years ago

Hello all,

Thanks to forum and wiki, I have changed my Moshi HPC-3020 (German edition) 40W laser into a liberated LAOS laser.
Will post photo's of connections and wires here later.

I've used the latest LAOS firmware (18-03-2015) and the stable 1.7 version of VisiCut with Inkscape 0.48.

My home position is in the top left hand corner (x=0, y=225000).
Orientation is correct when I cut achsentest_v2.svg
When cutting this (@ 1% power, 50% speed), I get a nice shape with accurate size.

The burned line is a little thick, but that could be because I need to up the resistance on the 5V line from the laser PWM.
(I've seen values of 470 Ohm to 1 kOhm reported, can anyone confirm accuracy?)

When using the engraving setting, the story is different, however.

After about 0.5cm of correct engraving, the X-axis suddenly shifts right, about 50% of the width of the shape. The laser test-fires once (see the dot (at the green arrow)) and then maps the rest of the shape to the right. This happens with both shapes and text-fonts (see pictures attached). When I try to engrave achsentest_v2.svg, it keeps engraving in a straight line.

Could it be engraving only works with bitmaps, not vector graphics?

Anyone have any tips?


Replies (8)

RE: Jump in X-axis when engraving - Added by jaap almost 9 years ago

Hi Ammo,

These pictures look like typical stepper motor skipping.

You can check this by engraving the same form twice. Does it jump at exactly the same spot? Probably a software problem. Does it jump at arbitrary places? Stepper motor skipping.

You can adjust the power to the stepper drivers with the little variable resistor on the Pololu's. Or you can lower the engraving speed. Cleaning the bearings can also help.

Too low power on the steppers usually gives random skips. Too high power gives noisy movement. Balancing in between is part of the fine art of laser mechanics ;-)

RE: Jump in X-axis when engraving - Added by AmmO almost 9 years ago

@jaap: Thanks for the fast response.

The skip point seems to be fairly consistent, in the different tests. It also does not occur with cutting, perhaps because the head movement is more regular (not weaving to and fro like with engraving).

After I rule out whether software (consistent) or hardware (random), I'll try tuning the Pololu's.

Just for my understanding: the Pololu's are transistor/capacitor based?

I've read a lot on the forum about 'optocouplers', is the Pololu an optocoupler or not?

(I'm from a software background, so the finer distinctions of electronics are not my specialty.)

RE: Jump in X-axis when engraving - Added by hpux735 almost 9 years ago

The pololu's are not optocouplers. They don't need to be. The coils in the stepper motors are isolated (at least they should be) from everything else, so there's no benefit from isolating them, they already are. The purpose of the optocouplers in other places is to allow signals to change voltage domains or protect on system from voltage spikes or surges in another. In the case of the laser cutter, we're a bit suspicious of the laser power supply (thousands of volts and chinese engineering, all in one convenient package!!).

I've had many battles with x-axis skipping, and I've tweaked many things. Ultimately what finally fixed it for me was a firmware update. I'm not sure if you're already on the newest one or not, though I assume Jaap would have mentioned it.

Another possibility is hardware. My cutter had a inconsistent sounding (and feeling) x-axis stage. I finally tore it apart to see why, and I discovered that the pulley wasn't round. I have it in this video:

Good luck!

RE: Jump in X-axis when engraving - Added by AmmO almost 9 years ago

Fine informative videos. Learned lots of things about the internal construction.

(Though being of a 'if it ain't broke don't fix it' mentality, I don't want to take the laser apart until I absolutely have to.)

The x-axis feels smooth enough. Y is a little edgy.

I've got some voltage regulation between my power supply and the laos board. It came with the German security wiring built into the thing. It cuts the power to the laser beam when the case is opened, and has a key and confirm-button sequence to start the laser. It takes several seconds to cut the beam, since the laser keeps running on the reserve current in the power supply, so I still want to buy some laser glasses. 8-)

Will check the gcr code and the visicut files for errors. It could be an engraving-algorythm problem (tricky optimisation, or trace artefacts that throw the rastering algorythms?). It doesn't seem to happen on cuts, only in engraving mode. And it is remarkably similar every time I make a different engraving.

Wondering how many Amperes the stepper motors can draw from the 24V supply?
(Since under/over-powering could be another hardware cause of skipping.)

RE: Jump in X-axis when engraving - Added by hpux735 almost 9 years ago

When I had a lab power supply powering the 24v rail, I remember the current being around 2 amps. Now that I've solved the X-axis gremlins for myself, I've actually been able to move back onto the laser power supply's 24 volt supply.

RE: Jump in X-axis when engraving - Added by KalleP almost 9 years ago

Rastering is for bitmaps, engraving is often used as a synonym for rastering but is really a term to describe surface marking rather than through cutting.

Vectors can be cut or engraved but not rastered (without first converting to bitmap outside of the LAOS software I think)

You need to have smooth motion of the carriages in both X and Y.

Sometimes when rastering you will find that software effects can be detected or mitigated by providing some additional space next to the marking area so there is room for the carriage to get up to speed.
Sometimes placing a line or even a single pixel at the limits of your raster image (outside the work-area even) can result in a workable plot file if the software has some quirk.

If there are power supply limitations remember the laser PSU control circuit shares the 24V supply that is used for your motors. The 5V (and 3.3V) rails are often regulated with linear regulators so should be reasonably immune to small variations.

Your laser beam should cut off in a few milliseconds after the laser on signals are turned off. What may be happening on your system is that the 24V supply that controls the laser supply is switched with the safety interlocks. This means that the laser PSU chip is running on residual power from locally situated capacitors and making a valiant attempt to keep operating. Aside from the slow switch off time I would be hesitant to drive the laser PSU transistors with reduced voltage pulses on a ongoing basis. Using this as a water temp/flow cut-off is OK as it should rarely happen, using it in the cover safety interlock seems sloppy. I would move the cover safety to a series connection with the laser enable signal and it will have immediate effect.

When tuning the Pololus some variants have no end stop on the trim-pot and going past the zero point can cause all sorts of strange symptoms and should be avoided. When first starting out set it to the mid point and work from there. On many of the Pololu style stepper driver designs the wiper of the trim-pot is at the current set voltage and you can connect a volt meter to the screwdriver blade and monitor the set-point as you adjust and write down the values for later if you have to replace of exchange modules at some point. The 2 steppers can often be set to slightly different values, the X motor has a lot less mass to move around though it may be planned for operating faster with rastering.

RE: Jump in X-axis when engraving - Added by depronman almost 9 years ago

I would expect the two steppers on the laser cutter drawing a total of 1.5 to 2.0 amps maximum. Any more than that an you have a axis binding.
I found adjusting the trimmer on the drivers was fairly easy, I turned them to around 75% as a starting point, then with the axis stationary, but power up, I increased the trimmer until the motor started to buzz slightly, then back it off until the buzzing stopped. Repeat for the other stepper driver.

This will provide a basic setting to get you running and will air on the side of overpowering the steppers and should not skip steppers, now run a few large cutting or better still engraving jobs and monitor the temperature of the two steppers, if they are warm to the touch, but not hot after about 10 mins of running they are about right, if hot then tuen the stepper driver trimmer down a little. If you turn the trimmers down too far the axis will skip steps, which is not easy to see on a laser cutter, hence the My idea is set off the the current on the high side and work down.

I have a home brew 3D printer which uses the same drivers and I set them up in the same manor, now the 3D printers works the drivers and steppers MUCH harder than the laser cutter, I have tuned the 3D printer over a few hours of running to the sweet spot where by the motors run barely warm, but they don't skip steps. This took a lot more effort on the 3D printer than it did on the laser cutter due to the laser cutter being far kinder on the steppers.

Best of luck with the tuning

RE: Jump in X-axis when engraving - Added by AmmO almost 9 years ago

Thanks. Will post when fully operative.