Scanner Jitter?

[planters]

It is funny how independent experimenters invent the same basic techniques. Each of my lasers has a different near field beam pattern; 20 dots blue, 16 dots red and single dot DPSS telescoped up to match the approximate, but not exactly 8x10 mm red and blue beams. My existing alignment method is: First, overlap on each dichro and align in the far field. Then, using smoke (I really enjoy this part) co-align each beam so that any additional width of a particular beam is balanced on each side of a narrower beam (more sensitive than dichro) and align in the far field. Finally, using a machinists square balance the distance to block each wider beam on each side of a narrower beam and once again re-align in the far field. The tweak I have to make to co-align the lines in the far field (12m) is no more than 10mm (1/2 the line width) and is stable so this seems to solve this. I'm more concerned with the hard won, tight 8mm static beams widening out to 20mm when scanned.

Steve,
There is one optical issue that might remain. I am using large scanner mirrors (14mm wide x 19mm long) and I am filling them. If they are flexing/twisting they might explain defocus. The fact that all three beams are suffering a similar 1.2 mrad additional spread might support this. One easy check would be to compare the far field line when the scanner is producing more/less acceleration. Does simply decreasing the K number ie 30K to say 20K actually decrease the instant to instant accelerations?

The PSUs I am using are from Mean Well and have an advertised rating of aprox 4x my peak draw by the Flex Mod P3s that are driving the diodes. I will measure the far field green line without modulation.

As I mentioned in another thread I am as yet without my replacement oscilloscope, so I can't look at the PSU output directly.

Marc,
You ground everything to one point and use Eye Magic scanners have you measured the static beam (before any scanner optics) vs. far field beam width?

My projected graphics are still very nice. A line width the size of an American Quarter at 12m may be typical, I don't know.

[kitatit]

It is funny how independent experimenters invent the same basic techniques. Each of my lasers has a different near field beam pattern; 20 dots blue, 16 dots red and single dot DPSS telescoped up to match the approximate, but not exactly 8x10 mm red and blue beams. My existing alignment method is: First, overlap on each dichro and align in the far field. Then, using smoke (I really enjoy this part) co-align each beam so that any additional width of a particular beam is balanced on each side of a narrower beam (more sensitive than dichro) and align in the far field. Finally, using a machinists square balance the distance to block each wider beam on each side of a narrower beam and once again re-align in the far field. The tweak I have to make to co-align the lines in the far field (12m) is no more than 10mm (1/2 the line width) and is stable so this seems to solve this. I'm more concerned with the hard won, tight 8mm static beams widening out to 20mm when scanned.

Steve,
There is one optical issue that might remain. I am using large scanner mirrors (14mm wide x 19mm long) and I am filling them. If they are flexing/twisting they might explain defocus. The fact that all three beams are suffering a similar 1.2 mrad additional spread might support this. One easy check would be to compare the far field line when the scanner is producing more/less acceleration. Does simply decreasing the K number ie 30K to say 20K actually decrease the instant to instant accelerations?

The PSUs I am using are from Mean Well and have an advertised rating of aprox 4x my peak draw by the Flex Mod P3s that are driving the diodes. I will measure the far field green line without modulation.

As I mentioned in another thread I am as yet without my replacement oscilloscope, so I can't look at the PSU output directly.

Marc,
You ground everything to one point and use Eye Magic scanners have you measured the static beam (before any scanner optics) vs. far field beam width?

My projected graphics are still very nice. A line width the size of an American Quarter at 12m may be typical, I don't know.

Hey guys I'm probably bringing nothing to the table here but when i was talking to Dave and Adrian about power supplies for my Eyemagics, he said they had some trouble when they first used them.

He said if you use Switch mode power supplies with them, you need to add an extra component. I think it was a diode somewhere. I don't know what the symptoms were but they pulled out a bit of hair trying to figure it out at first.

If Dave or Marc can't shed some light on the problem, dsli_jon has a lot of experience with Eyemagic scanners to.

kit

[dave]

The diode prevents the negative 15v rail on the scan amp locking up if the positive supply rises first IIRC.

This would stop the scanners working at all. Eric's problem sounds like noise pickup maybe?

[taggalucci]

Eric's problem sounds like noise pickup maybe?

...and coz there aren't enough Aussies on this topic... I've seen wobbly, noisy behaviour from scanners where the grounding scheme is not well implemented.

I'm interested to know whether this observation is made only when the scanner is in motion or also occurs when the scanner is pointed at various locations. Eric, have you tried static beams at different positions to see what effect this has?

I'd hope your mirrors aren't flexing, though I can understand why you might think that.

[planters]

That is an interesting question. When the scanner power supply is offline the spots on the projection screen are tight. As soon as the PSU is energized the spots shift substantially and broaden. This is without any intential input to the scanner from the computer. Also, when in this mode I have accedentially contacted a mirror with one of my paper masks and there was a noticable vibration. I can't rembmer if with a DT scanner I used several months ago, if this shift was also accompanied by broadening and vibration, but I don't think so. Of course at that point my beams were not at this higher level of quality. Hmm

[andyf97]

Try isolating the scan block from your enclosure with some heat conductive sheeting.

[laserkits]

Simply disconnect your red and green lasers and try the test with only the DPSS. If the beam width is still increased then you know it has nothing to do with near field collimation and something is going on with the scanners or the scanner mirrors.

[laserkits]

That is an interesting question. When the scanner power supply is offline the spots on the projection screen are tight. As soon as the PSU is energized the spots shift substantially and broaden. This is without any intential input to the scanner from the computer. Also, when in this mode I have accedentially contacted a mirror with one of my paper masks and there was a noticable vibration. I can't rembmer if with a DT scanner I used several months ago, if this shift was also accompanied by broadening and vibration, but I don't think so. Of course at that point my beams were not at this higher level of quality. Hmm

A ground loop can and will cause this... I would start with Bill's Star grounding scheme doing one mod at a time until the problem goes away. Do you have ILDA pin 25 connected to the "DC ground" of the scanner power supply? If It is a + and - power supply it will have three outputs, +, -, and Gnd. If it uses two power supplies together to get the + and - outputs, then the + from one supply will be wired to the - from the second and this forms the ground...

[logsquared]

Sounds like your scanners are overdamped and ringing. Or your power supply has bad caps and the switching frequency or some harmonic of it are "riding" on the DC output. On the older eyemajics you had to to tie the scanners amps together with a low Z wire to help crosstalk and ringing. Not sure if the newer amps need this?

[planters]

logsquared,
I think you mean underdamped and ringing. Right?
laserkits,
Yep.. I think your right and I will look up Bill's paper and start down this path.
In general, is any broadening unusual and the goal should be to see a static spot width equal to a line width of a scanned pattern?