ILDA Analog Modulation Laser Control 0vdc-5vdc

JBWolf

Howdy everyone!
I have a project that could use some expertise in analog modulation control.
So far I have combined 3 lasers with pbs & dichroic mirrors into a nice white.
This is a non-scanning system (no mirrors/motors), just a straight beam.
Dealing with high frame rate recording has been a problem that leaves PWM/TTL unsuitable for good visual results.
I now have an analog modulation controller with 3-inputs, each 0-5vdc for 0-100% output with each laser diode... ILDA compatible.
I am not sure of the amperage limits, but would be within ILDA specs.

How do I create and control three 0-5vdc outputs on the propeller so I may vary the output power for each laser safely?
Thanks!

Phil Pilgrim (PhiPi)

Use the Prop's DUTY mode counter output, driving a simple RC low-pass filter. Feed the filter output to an op-amp configured for a gain of 1.5. The op-amp's output will range from 0 to 4.95V, which you can use with your modulation controller. Since you have three channels, and each cog in the Prop has two counters, you will need more than one cog to pull this off.

-Phil

Tracy Allen

What is the bandwidth or rate of change required, and number of levels? Fastest would be three ten-bit DACs using 30 pins of the Prop1. DUTY mode is easier with just 3 pins, but it takes on the order of 20 microseconds to resolve to 10 bits.

DigitalBob

This is a 8 channel DAC board I made several years ago using the MCP4922 DAC (12 bit) it work pretty well, 0-10 volt scaleable output with zero bias.

JBWolf

I think 256 levels on each channel would be sufficient so I can use a standard palette for colors... which would give a max of 16.7mil combinations.
As for rate of change, If 216khz can be achieved that would be perfect, otherwise I'll take what I can.

Peter Jakacki

Would you consider using the P2 since it can do this standing on a cog and three of its smartpins? Other than a SPI Flash, you don't need to add extra chips except the opamps for 0..5V output.

Tracy Allen

DUTY on prop1 with clkfreq=80MHz can give 256 steps at 312.5kHz (80/256). However the cycle is a variable frequency. At the extremes it consists of one pulse 12.5ns wide followed by a period of 255*12.5ns in the complementary state. In the center, it is a 40MHz square wave, 50% duty. So for full range the followup filter needs to handle that full range. Can your application avoid the extremes, skip the 1% and 99% levels? At step 4 or step 252 the simple filtering will be much easier, at over one MHz.

True PWM is usually thought of as a constant frequency with variable pulse width. The Prop1 can do that at high frequency by overlapping the phase of the cog's ctra and ctrb, directed to a single pin. So 3 channels would need ctra and ctrb from 3 cogs, and the output variable width at 312.5kHz. A higher order filter might allow 216kHz throughput without too much bobble. Bump Prop clkfreq to 100Mhz with a 6.25MHz crystal.

Easier with an 8-bit R/2R network using 24 pins. Or a DAC or DDS chip that can be updated fast enough.
Or, Peter, what smartpin modes on Prop2 would be best for this?

Phil Pilgrim (PhiPi)

You could use two DUTY-mode counters with minimal filtering to increase frequency response. Each would feed into a summing junction using resistors in some n:1 ratio. That way neither would have to operate at the low-frequency extremes of their duty cycles.

-Phil