High side driver

[DMills]

The full scale current is determined by R1 and the supply voltage.

As drawn, R1 is 1 ohm, so will drop one volt per amp, and if you use darlingtons then you will have about 1.4V total Vbe, so with say 5V up it, and 2.5V across the laser, the output stage would top out at (5 -2.5 - 1.4)/R1 = 1100mA, which would equate to requiring 11mA down the 250ohm chain or R2,R11,R3, plus 1.4V drop in Q2, so there would be about 0.85V across T2 (so thats all right).

You can impose a hard limit on maximum current at any given supply voltage by playing with R11 as the current flowing in it when T2 saturates will directly determine maximum output (Useful if the modulation input might get away from you).

For accuracy, R1 should be sized to drop as much of the excess voltage as possible, so that small Vbe errors have minimal effect.

I will draw up a rather more refined version tomorrow night, and might even test it over the weekend.

Regards, Dan.

[ElektroFreak]

Very impressive.. can't wait to see this as it's refined. Very nice work!

[DMills]

Just came up with another variant that should modulate stupid quickly, make the high side current source a fixed current and connect a low side current sink across the diode...

Negative mod, but it sure is simple.

Ok, assuming my abomination actually works, what do folks want it optimised for (Current, Supply voltage and diode voltage drop)?

Regards, Dan.

[ElektroFreak]

It would be possible to optimize it across a range of current, right? So that we could set it for any diode we want..

[shrad]

high side current source and low side current sink is a good idea, but you don't have the advantage of grounding the cathode anymore

but keep this idea, it is really interesting and will have double security, and as you said modulate quickly

what about a driver which would have a master regulation part with opamp etc.. and some chainable source/sink modules to be soldered side by side, for multiple reds or blurays?

[DMills]

Actually you do!
The low side sink is in parallel with the diode, and the diode current is equal to the difference between the high and low side currents.

I figure to optimise for say 0 - 1A, 2.5V min across the regulator (so a 5V rail will run a diode up to 2.5V), 30V max (for a whole chain of series diodes!).

Sound reasonable?

Regards, Dan.

[lucerne]

http://www.laserblog.de/2009/11/neu-picodrive-rbld-2/

This is maybe what you're talking about. It's able to handle 6 laserdiodes. Red or blue ones.
The cathode of each laserdiode is connected to GND.
The driver needs only 5V supplyvoltage and it has a buck-boost switching regulator inside. It's responsible to get the right voltage for red or blue laserdiodes.

I've also made a driver for DPSS lasers. But thats another project
http://www.laserblog.de/2009/05/upda...rive-dpss-v11/


Regards, Leander

[ElektroFreak]

^Looks nice! What's the cost?

[lucerne]

Well, the PicoDrive DPSS is in production at the moment. The PicoDrive RBLD not yet (but in a few days/weeks). Costs are about 195€ for the PicoDrive DPSS and 125€ for the PicoDrive RBLD.

Regards, Leander

[JALEO]

Dear Leander
Do you have the picodrive RBLD ready?
Do you have the specs?

Thanks