Rise times and fall times

[fuzcub]

Ok all, I've had a bit of time to mess around with measureing the linearity , rise time and fall times for the lasers in my RGB.
I'm picking off about small % of my final mixed beam, them filtering out any IR, then splitting the beam back into R-G-B and feeding them into 3 photodiodes.
I'm planning to use this to drive a hybrid light/delay/threshold circuit to try to match everything up.

Anyway, I thought I should share my results....

Red is a 635nm 300mw from Lasever ( Magenta trace in graphs)
Green is a CNI 500mw driven with a Lasever driver inc crystal TEC (Green trace in graphs)
Blue is a Laserwave 473nm 150mw. ( Blue trace in graphs)
Modulation signal is yellow in graphs - note that the laser drivers seem to flatten the tops of the signal .

Note that these sensors are not calibrated to any particular measurement - only so that once the lasers have warmed up and stabilised that they don't saturate the photodiodes & their amplifiers.

First of all the linearity over a 1 second ramp up:




Then the rise times:


And again, zoomed in on the start, and averaged to clean up the blue noise:


Last of all the fall times:


Cheers,
Barney

[allthatwhichis]

Wow, you might want to get with DZ on this. He is working on the same thing although a bit farther along I think.

[mixedgas]

Ouch, no wonder a PCAOM looks so much smoother!

What would happen if we turn the blue on to full power at each point and then let it follow the analog? Ie Spike it with a differntiator?



Steve

[dave]

Wow,

Not even the diode is linear!

[drlava]

Nice graphs, well done. If your red is direct injections it should be linear.. something's amiss there. You've come to the right place, members are working on solutions here, a solution for your timing differences and 0 bias differences already exists in DZ's correction board.

[dave]

Nice graphs, well done. If your red is direct injections it should be linear..

Not according to Eye magic

http://www.eyemagic.gr/laser-iris-colorsafe.html

Even direct diode lasers (such as red and the newly developed 445nm blue) benefit from the real time correction as we have noticed through the power feedback they do not respond linearly to the incoming current, mainly because the variable duty cycle of the projected patterns creates a variable heat load on the diode that cannot be tracked and corrected by the peltier cooling.

[DZ]

Wow, very, very nice work Barney. I love the 1 second ramp up graphs. The delay board I have can compensate for 2 of the 3 problems, however the 3rd problem is why the delay board has a difficult time correcting light output when the laser is at anything less than full output. I also plan on having a fix for the 3rd problem soon. I've just finished a small projector that has some serious modulation issues, I figure if I can make the output from this projector look good, then it will be a product worth selling. It consist of a half dead LW blue, a maxy module (which is in pretty good shape) and an unknown green laser being driven by a lasever driver.

May I ask what equipment you used to capture the output like that?

[mixedgas]

Considering how bad the stock Maxy driver is, are you not you dooming yourself from the start? Or did you change it out?

The red, violet, and direct blue diodes could benefit from using a fast on/off driver, a 8 bit flash A to D, and a two chip PWM circuit. I did that for a TTL AO in the 1990s for getting intensity, and to the eye its very linear. It has to be by definition. If you have problems with brightness levels, you can shift the total drive current and the offset. I had 6-10 complete PWM cycles per point at 12K. You did not see the pulses when scanned. This ditches the light feedback for those diodes. The diode is either full on or full off, and has the bandwidth to do it. Risetime approaches a few nanoseconds or so with a proper driver. Since the PWM clock was not synced to the grahics card, but came from a crystal oscillator, the points were slightly averaged at the edges, resulting in beautiful blending between points.

It stood up to review at a professional LEM with very favorable comments. I stopped building it when Harris/Intersil dropped the cheap flash converter chip, but other companies make them for video now. Other laserists built them and had equal sucess doing RGB with 3 surplus ON/OFF aoms.

This would shift your issue to just the green and blue, and only possibly need a small delay for the red. That delay could be digital, using a cheap shift register chip, curing one of the adjustment issues.

If the blue DPSS were inherently polarized, would it not just be easier to have CNI make a AO cell in the laser and go digital now?

I'm doing a bit of a Keeperx here, just stiring the pot.

PS, could somebody scope the blue TEC current while modulating?

Steve

[drlava]

Not according to Eye magic

http://www.eyemagic.gr/laser-iris-colorsafe.html

This is true, but that effect would be no more than a few percent I would expect. His graph shows much more than this.

[DZ]

That's kinda what I expected for red. I've said several times that it seems red lasers have 2 knee's in thier modulation range. That first graph shows it well. It rises quicklly, then stays steady for a while before it finally spikes up.

Considering how bad the stock Maxy driver is, are you not you dooming yourself from the start? Or did you change it out?

Nah, stock maxy driver. It seems to be working just fine. I'm watching the output of the photodiode on a scope, when red turns on, it's rather smooth, green is all over the map though!