Inconsistent engraving problem ?

[twehr]

If the coolant pump is on only while the laser is on, maybe there are bubbles or dirt in the coolant, which randomly disperse some of the laser's power, until the coolant circulates a little and flushes them out, which is when the situation improves.

My pump is running anytime there is power to the machine, not just during lazing, so it is not that, I believe.

If it were mechanical - related to dirt, etc, it would vary with physical position of material on the bed. It does not seem to. Though my rails are often gritted up from jobs, I have repeatedly cleaned and tested with no real change.

Remember my comment about where the banding begins. No matter how many different thought paths I go down, I always come back to the fact that the problems are more predominant when there is a change in the on/offs of scan lines. I have seen it in photos where the facial features cause the lines to begin or end. In nearly every case, I can see a correlation between the features of any image and where the banding is or is not. There is always noticeable start point that is at the same y level as the image feature change, like bottom of a nose or edge of a running stream or edge of wings, etc.

[cvoinescu]

This could be proven with a judiciously designed test pattern. Looking at the engravings again, it may be that the power supply has a lag problem. When the amount of 'laser on' per raster line increases, the power seems to decrease over the next few lines, then restore to normal again. When the 'on' amount decreases, the next few lines are darker. When the variation is gentle, the difference is minimal. Something a little like this used to be common in dying CRT high voltage supplies in black-and-white TV sets, although the timings there were a few orders of magnitude faster (and the power a couple of orders of magnitude lower), and here there seems to be some feedback that tends to counteract this, albeit very slowly.

laser test pattern.png (956 Bytes) Viewed 38497 times

This test pattern should be telling. Engrave it either as white=burn, black=don't burn, or the other way round; I'm not sure which is better. The first shape should show banding, the second one may or may not, the third one should not. The fourth one has a constant amount of 'laser on', but split into different numbers of on-off cycles, so if it's the turning on of the laser that causes the problem, it'd show banding, otherwise not. The last one has the same amount of 'on' and the same number of on-off cycles per scan line, except the bottom band which has many more on-off cycles than the others.

[lasersafe1]

Ior (b) some scan lines are packed closer together and some spaced further apart, which would also vary the total amount of power delivered to the material.

Indeed a real possibility since there was an earlier statement from the thread author that he didn't know how to change the dpi of the image. As Twehr has taught us, this is a vital part of producing good engravings. The dpi of the image must match the stepover distance or be an even increment of it or there will be some skipping. Yet this still does not explain why it would skip in the top portion of the Ruger symbol but not the bottom.

I like those test patterns. Perhaps there should also be a set that have been dithered with a dither mask. This brings up another possible cause. If the original Ruger symbol were dithered with a repeating pattern mask instead of a random pattern, you may get aliasing because the dithering may be at a different pattern rate than the dpi. I always use the Floyd-Steinberg dither. I get the feeling the original Ruger pattern was not dithered at all and this was a pure black and white engraving. Working on a dithered pattern is essentially PPI. Working on a B&W is not at all like PPI.

[SScnc]

Thank ALL of you for your great replies,I just saw them. I'm going to spend some more time re-reading through them and I'll also see what I can do in Corel to improve the image before sending it to LaserCAD.

Thanks again, I really appreciate it !

Steve

I'll post back my results with more pics

[BenJackson]

Another problem I see from the original post. I know my tube won't consistently fire down at 2 or 3 mA. Perhaps this is too low. Can you speed up the raster speed and go to higher currents of 8-10mA?

I'm curious to know what is driving your INPUT (pwm power) pin on your power supply. I'm converting my LinuxCNC config to use the 5i25 (FPGA driver) so I needed to re-calibrate my PWM vs power because I know the new card is putting out different voltage for "high" output. I just did a test where I picked some arbitrary low point (0.1 duty cycle) and fired the laser and got about 1mA with the test button. That's with 20kHz PWM driven by the FPGA. I also got 18mA output at 0.49 duty cycle, which seems quite low to me (was using 0.7 at a *higher* voltage) so perhaps the INPUT pin does not quite work as expected.

(That 1mA really was firing the laser, it lit up and it made a tiny, tiny spot on some wood. Barely marked it, but it was firing)

[lasersafe1]

I was only speaking from memory from 3 years ago when I actually had an ammeter connected. I no longer use the ammeter because my machine "knows" the limits. I use the DSP that sets an analog 0-5V (via PWM) in the power supply and then uses the TH pin to fire the laser with digital on-off. It doesn't matter if you talk about PWM or Analog, both will produce 0-5V in the supply. Everyone here should already realize that you can't do power changes rapidly via the PWM line. It responds relatively slow because it is integrating your PWM. The TH and TL lines are the only lines that will give an immediate power supply response, and they need 5V TTL levels.

Anything under some voltage (0.4?)at in the PWM conversion will not give a reliable output because the tube will not reach the strike voltage properly. As a tube ages, the strike voltage can be harder to reach, so people find they must raise their minimum setting.

The ammeter is only accurate if you are in full DC cutting mode. If you are doing PPI or raster engraving across a dithered item, then the tube is firing with a high current strike pulse and then relaxing down to the DC setting before quickly shutting down again. The meter cannot respond fast enough to show you what is really going on, so it gives you an integrated reading. So yes, in this case I suppose you could see a reading of 2mA, because the tube is firing at some higher current but for brief periods as commanded by TH or TL. I was assuming that the first images posted were not dithered, so that is why I questioned the 2mA.

[Gadroc]

Another problem I see from the original post. I know my tube won't consistently fire down at 2 or 3 mA. Perhaps this is too low. Can you speed up the raster speed and go to higher currents of 8-10mA?

I'm curious to know what is driving your INPUT (pwm power) pin on your power supply. I'm converting my LinuxCNC config to use the 5i25 (FPGA driver) so I needed to re-calibrate my PWM vs power because I know the new card is putting out different voltage for "high" output. I just did a test where I picked some arbitrary low point (0.1 duty cycle) and fired the laser and got about 1mA with the test button. That's with 20kHz PWM driven by the FPGA. I also got 18mA output at 0.49 duty cycle, which seems quite low to me (was using 0.7 at a *higher* voltage) so perhaps the INPUT pin does not quite work as expected.

(That 1mA really was firing the laser, it lit up and it made a tiny, tiny spot on some wood. Barely marked it, but it was firing)

As lasersafe said you need to test the duty cycle with out PPI / raster to calibrate your power mA to duty cycle. It is consistent with my experience with the DSP controller that a 5v 20kHz PWM will not require a full duty cycle for your max mA. I now run a limiting resistor between the DSP and the Input line other wise 32% power on the DSP creates max mA for a 40W tube. It's not necessary if you just limit your settings from 0-32%, but it's a easier to translate recommended settings from vendors if 100% is really max mA.

[BenJackson]

As lasersafe said you need to test the duty cycle with out PPI / raster to calibrate your power mA to duty cycle. It is consistent with my experience with the DSP controller that a 5v 20kHz PWM will not require a full duty cycle for your max mA. I now run a limiting resistor between the DSP and the Input line other wise 32% power on the DSP creates max mA for a 40W tube. It's not necessary if you just limit your settings from 0-32%, but it's a easier to translate recommended settings from vendors if 100% is really max mA.

I was using the test firing button held down long enough for the reading to stabilize.

Your 32% comment is very interesting because it pretty much exactly matches the result I just got (your 32% * 5V vs my 49% * 3.3V, both about 1.6V mean). Before, when I was using a software-driven PDM (pulse density modulation, gives a higher frequency for the same update rate) I had to use a much higher setting to get full power. I'm pretty sure I did this test with the front panel knob:

viewtopic.php?f=16&t=557&start=30#p5565

Back then it took more V to get power. It's almost like the steep part of the graph now continues up to 1.6V for full power. I wonder if my tube (or PSU) "broke in" and I should have recharacterized it long ago.

[SScnc]

Very good news !!! as lasersafe1 and cvoinescu stated, it was mechanical. My Y motor is a dual shaft stepper that is offset to the right side. The right hand belt pulley attaches directly to the motor shaft while the left hand pulley is attached to a long shaft which is of course then attached to the other stepper output shaft. The set screws on that long shaft had come just a touch loose, I tightened them down very snug and instantly got what I consider to be beautiful results on several different engraving files. NO more horizontal lines.

My main digital camera is at work, I'll get it and post the results of the Ruger logo ASAP. I did learn how to do the artwork with the correct DPI and I tried many different settings and kept getting those horizontal lines, wish I had listened better about checking the mechanics, I was over confident that all my mechanics were fine. WRONG !!! I'm still learning, and always will be.

Thank you all again for your help,

Steve

[Techbuilder]

Very good news !!! as lasersafe1 and cvoinescu stated, it was mechanical. My Y motor is a dual shaft stepper that is offset to the right side. The right hand belt pulley attaches directly to the motor shaft while the left hand pulley is attached to a long shaft which is of course then attached to the other stepper output shaft. The set screws on that long shaft had come just a touch loose, I tightened them down very snug and instantly got what I consider to be beautiful results on several different engraving files. NO more horizontal lines.

My main digital camera is at work, I'll get it and post the results of the Ruger logo ASAP. I did learn how to do the artwork with the correct DPI and I tried many different settings and kept getting those horizontal lines, wish I had listened better about checking the mechanics, I was over confident that all my mechanics were fine. WRONG !!! I'm still learning, and always will be.

Thank you all again for your help,

Steve

I know this is a super old post but this is happening to me as well and I would be so appreciative for any information.

What settings do you use to achieve the final result if you don't mind me asking?