simple drive modulation problems

[swamidog]

i've got a new projector for show programming. it was a commissioned build by another PL member. single mode diodes doubled up via pbs cubes and ct2615's with gigantic psu's.

he did a beautiful bit of work and it's a great little projector.. except

the diodes are driven by simpledrives v2500 r4. one driver per diode. a total of 6 drivers. nice and neat, but there's something really weird going on with the modulation.

modulation on the low end is constantly flickering. it kind of looks like the standby beam suppression is cutting in and out or the sbs threshold is drifting. the other issue i see manifests in closed shapes. the trailing lines fade to black before the shapes close. it's like the diode modulation drops off way too soon. i've tried adjusting the bias current on the drivers and upping the low end setting in software, but the problem is still there. it also acts like the green modulation is slow than the red and the blue. this is really weird since it's an all diode system.

here's a shot of the color linearity test patterns for each channel in LSX.



each of those little circles should be closed. ignore the missing circles on the blue wheel. that's a camera artifact.

i also tested in quickshow and beyond. same issue.

is there anything i can tweak on the drivers? has anyone experienced issues like this? i'd very much like to see color test patterns from other simpledrive users.

thanks!

[mixedgas]

First step is to rig up a battery powered test rig using a 9V battery, a 7805 Voltage Regulator, a filter cap, and a two to ten Kilo-Ohm Potentiometer. You'll also need a good digital voltmeter with a 200 mV scale. I suggest battery power because many modern, low cost, bench power supplies have large spikes as you change the voltage.
~
The goal is to see where the beam suppression kicks in, and kicks out, and if it has hysteresis in the circuit. Hysteresis is often designed into comparator circuits to prevent glitches around the trip point, usually hysteresis circuits have a different turn on and turn off point. The turn off point will often be lower then the turn on point.
~
Apply the variable voltage to each driver in turn. If you have a power meter, even an un-calibrated one, check the linearity and turn on/turn off points of the beam output.. Do this for each driver, and do it several times. Often there are surprises in where the suppression circuit kicks in and kicks out. Bonus points if you use a oscope to look at the switching points for glitches. While I have never tested the Simpledrive, I done this with other drivers.
~
What I see suggests that the diodes have vastly different operating curves at the low end. A simple test would be to move the bias current on the driver, if it has a bias current setting to get past diode lasing threshold.
~
Generally I avoid using DACs and ramp test patterns for this test, to avoid software modulation artifacts.
!
Steve

[Bbe]

It seems this requires additional tuning of the quiescent current. V2500 is too strong for single-mode diodes, the pots operate in a very short section ~1/20 of the full range.

What you need to do:
Set the bias pot to minimum.
Adjust the zero pot to a position where the quiescent current crosses the 0mA mark. Adjustment can be very thin, due to the above.
Turn off the power, then close ntc pins on the driver and turn on the power - the driver must start in the bias current setting mode.
Set the threshold current for each diode to half of that specified in the datasheet. External modulation signal is not required for this.
Restart the driver, and adjist the operating current using the gain pot.

Again, all the adjustments will be very thin, you need to be careful.

The question of the quality of the supply voltage is also relevant

This driver has software-defined beam suppression circuit design, in addition to the hysteresis, this suggests some time lag.

The goal is to see where the beam suppression kicks in, and kicks out, and if it has hysteresis in the circuit. Hysteresis is often designed into comparator circuits to prevent glitches around the trip point, usually hysteresis circuits have a different turn on and turn off point. The turn off point will often be lower then the turn on point.

[swamidog]

thank you BBE,

would these be better drivers to use?

http://www.ebay.com/itm/500mA-analog...wAAOSw~y9ZLbpg

It seems this requires additional tuning of the quiescent current. V2500 is too strong for single-mode diodes, the pots operate in a very short section ~1/20 of the full range.

What you need to do:
Set the bias pot to minimum.
Adjust the zero pot to a position where the quiescent current crosses the 0mA mark. Adjustment can be very thin, due to the above.
Turn off the power, then close ntc pins on the driver and turn on the power - the driver must start in the bias current setting mode.
Set the threshold current for each diode to half of that specified in the datasheet. External modulation signal is not required for this.
Restart the driver, and adjist the operating current using the gain pot.

Again, all the adjustments will be very thin, you need to be careful.

The question of the quality of the supply voltage is also relevant

This driver has software-defined beam suppression circuit design, in addition to the hysteresis, this suggests some time lag.

[Bbe]

I did not say that the v2500 is not suitable for single-mode diodes, just it's like to use a sledgehammer to crack a nut )) You will need more time to set it up.

Yes, they are easier to adjust and more stable when used with low-power diodes.
Just imagine that the steering wheel in your car has a quarter turn instead of three turns

would these be better drivers to use?