Losing shades of grey in raster

[Jean Paull]

I work mostly with raster images on glass. Although I use my own mixture, the procedure is similar to that used with Cermark - application to the glass, lasering and then washing off what has not adhered.

My initial laser tube expired and was replaced in June. After re-alignment I was getting great results for a couple of months. In the last few weeks however I have encountered a problem where my lighter shades of grey (10% to 30/40%) are washing off. The darker shades stay attached. Quite fine text in 100% grey tone stays attached and shows good detail.

I prepare the halftone dithered images in exactly the same way. I've checked and refined alignment and even re-done the tube connection. Over a 2-3 hour testing session it may seem like a modest speed or power adjustment is improving the situation but then a test run at the same settings in the same position again shows serious wash-out problems.

The tube is the most recent change, could it not be functioning properly? The Chinese laser has been working reliably for 4 - 5 years and has not been heavily used, what else might be causing the problem?

[Rich Harman]

Hi Jean, welcome to the forum.

The original tube lasted almost five years? Are you using a Reci tube?

[Mike Null]

Welcome to SMC.

I suspect that the problem is the the grays. Your machine driver scales back the power in specific amounts depending on the shade of black you are using. Since your compound, like Cermark, relies on heat for bonding a reduced power level won't produce sufficient heat to permit it.

I think you'll have to do considerable experimenting with your power settings or speed to solve this issue. You may find that changing to half toning will solve your problem though I doubt you'll like the screen you get.

[Dan Hintz]

In addition to Mike's suggestions, consider checking the focus. A poor focal point might easily result in full blacks getting enough power to bond, but low-level grayscales not.

[David Fairfield]

Something to be aware of on the Epilog driver (probably not any help to you but...) when color mapping is switched on the greyscale no longer functions, the software will interpret greys as solid black. I've tried this with a test gradient, everything from 1% to 100% is engraved as 100%. When the color mapping function is off, it returns back to a gradient. So, at least theoretically, a driver issue can interfere with rastering a gradient.

Dave

[Lee DeRaud]

...In the last few weeks however I have encountered a problem where my lighter shades of grey (10% to 30/40%) are washing off. The darker shades stay attached. Quite fine text in 100% grey tone stays attached and shows good detail.

I prepare the halftone dithered images in exactly the same way...I'm apparently missing something here: are these images greyscale or are they halftone/dithered?
If halftone/dithered, everything is lasered 100% black: it's a 1-bit process.

[Jean Paull]

Thanks for the replies.

The tube is glass CO2, Rich.

I check the focus before each run, Dan. It is done manually using a template. I removed the lens and cleaned it well. It does have a small scratch on it but this has existed since it arrived and has not previously been a problem. Could something have altered with the lens/focussing? It makes sense that a focus deterioration could cause the problem and the lens quality is probably not particularly high.

I hadn't considered a power reduction for lighter shades of grey being a factor, Mike. However, the procedure was working fine previously, even for awhile after the new tube was installed - just fewer, more dispersed dots for the lighter shades, but still securely attached.

To clarify the type of image being sent to the machine, Lee. It will only accept a bitmap. In Corel I convert a greyscale image to bitmap, choose black and white mode, apply a dither, save as a bitmap and then import the file in my laser software (Laser 5).

I really appreciate the assistance with troubleshooting this one.

[Lee DeRaud]

I hadn't considered a power reduction for lighter shades of grey being a factor, Mike. However, the procedure was working fine previously, even for awhile after the new tube was installed - just fewer, more dispersed dots for the lighter shades, but still securely attached.

To clarify the type of image being sent to the machine, Lee. It will only accept a bitmap. In Corel I convert a greyscale image to bitmap, choose black and white mode, apply a dither, save as a bitmap and then import the file in my laser software (Laser 5).If I read that correctly. it means that there isn't any power reduction for "shades of grey", because everything is 100% black at the output of the laser. So the only difference is that small isolated dots ("lighter" greys) aren't getting fused properly, while larger dots ("darker" greys) are. That sounds more like some kind of issue in the laser itself, rather than the software/driver. (Which makes sense, since the hardware changed and the software did not.)

However, if so, it should also show up in very small text as a kind of "astigmatism", where things like the crossbar of a "T" bond ok, but the upright does not, since it looks like a smaller "dot" along the raster scan direction.

You should also be able to see a difference between using halftone (where the dot size changes) and other forms of dithering (where the dot spacing changes).

[Mike Chance in Iowa]

There is one more factor to consider Jean. You mention "Although I use my own mixture, the procedure is similar to that used with Cermark" so my question to you has there been any change in your own mixture? Are you working with the exact same mixture, or have you mixed up a new batch, or purchased more? It's possible something has changed with that.

[Chuck Stone]

I'm thinking that even though you have checked the focus before each run, that still doesn't rule out
a focus problem. (not that I know what I'm talking about, I'm just sort of thinking out loud here)

You mention the scratch on the lens, you mention a new tube. And shortly after replacing
the tube this issue starts happening.

Is it possible that the new tube's power has affected the scratch in the lens? In the coating?
If it scatters the beam more, focus can become an issue. And that doesn't mean that your
mark will necessarily be out of focus, but it could be that there is an interruption of the power
being applied to your medium and it would be more noticeable on lighter areas. (which are not
really lighter at all, just more widely spaced black dots) There is a certain amount of overlapping
when the head travels back and forth. And in the lighter areas, you're not getting this overlap.
It is possible that the 'black' areas were getting too much power, the lighter areas not enough..
and somewhere in the middle was just right. But now that there has been a change in the beam
(for whatever reason) now the scale is shifted. The black could still work because it is getting
multiple passes of the beam.. but it may be that it is no longer enough for the lighter areas
because the dots are more widely spaced.

If you bump up the power, do the lighter areas come out better?
Have you run a focus test? You might try making marks at various heights, regardless
of where the focus probe says you should be.
Can you turn the lens around on your machine? If so, does it affect the problem at all?

ok, that's all my guesses..

[Jean Paull]

I appreciate your suggestions for testing and am working through them.

You described the problem with greater clarity than I was able to manage, Lee.

Nothing has changed with the materials used for coating the glass - no new packages or batches. Since the problems started I've become more methodical with the mixing and application also.

The first test I ran last night was a dismal failure with even the darker tones washing off. The chiller had been running for awhile and I had test fired to confirm alignment in four positions but the test was the first work that had been run. Settings were speed 400, power 25%.

I then cut a few small circles and squares from 1/8 in. acrylic just to confirm the tube was firing properly. They came out just fine with the usual setting of speed 9 and power 100%.

I went back to the glass testing and was getting great results - as in no problem with adhesion even with the lighter greys.

I'm wondering if this was random or if pre-firing the tube at a higher power before starting detailed raster work contributes something. A search of the board didn't reveal anything but I might not have been using correct search terms.

I'll see if I can replicate the result with my testing today.

[Mike Chance in Iowa]

I remember reading a post some time ago about a person having problems with their laser every time they turned the laser on. They had to do a test cut on scrap material before they were able to run their real jobs. I don't recall what type of laser it was or which forum it was, but you may be on to something. I'm curious to see if you can replicate the solution.

[Chuck Stone]

I'm wondering if this was random or if pre-firing the tube at a higher power before starting detailed raster work contributes something. A search of the board didn't reveal anything but I might not have been using correct search terms.

tickle voltage, maybe?

[Dan Hintz]

If you're seeing random power fluctuations, the likely causes are tube gas getting weak (leak) or the power supply slowly giving up the ghost.

[Jean Paull]

I was not able to replicate the prior result of an improvement after running a higher power job.

While I initially thought that small sized text in 100% black was adhering adequately I tried a font that had more variations in line width within each letter and I am losing the narrower lines so that is showing the problem as well.

Results of running a halftone version beside a dithered bitmap have varied. There are instances of each being better than the other and also being about the same in levels of grey retention.

Turning the lens upside down resulted in a somewhat worse result.

Is there a way to test for gas leakage from the tube (other than comparison to a new tube)?

How would I test the power supply?

I really appreciate the comments.