Thanks Robin, Thats what I was looking for ..
Rock on...
Thanks Robin, Thats what I was looking for ..
Rock on...
"My signature has been taken, so Insert another here"
http://repairfaq.ece.drexel.edu/sam/laserfaq.htm
*^_^* aka PhiloUHF
Do I sense some high power DPSS MaXYZ's Modules in the pipeline?
What are the critical specification I should look out for? I'm going to try to find a suitable replacement MOSFET on my own. Is it just Vgs(th), Rds(on) @ 5V, Ciss, Crss and its transfer characteristics graph? Any other factors I should worry about? Thanks guys.
Have any of you used Norm's drivers? He was selling them for awhile on Ebay as a kit. I have a couple of those. He's basically using an LM317 for regulation of the constant current and a TIP31C in parallel with the laser diode for shunting. How does that compare?
What do you guys think about this one?
http://www.vishay.com/docs/70804/70804.pdf
I think everything fit the requirements except for its Ciss, which is at 2.7nF typically. Would this be good for <40KHz?
I'd go with one of those that Robin picked and mentioned here last night.
Gary, that TIP32C might be good. In shunt mode, the 0.6V drop won't matter as it's less than the laser diode's own drop, and it should be nice for analog mod too, having better linearity than a dedicated switching device. I'm not sure why power switching MOSFET's are so often chosen even in shunt modulators for analog, where arguably a bipolar transistor will do better. If you want the isolation the insulated gate of the FET provides, it's better to get it with an optoisolator anyway, given that plenty of nasty things can happen by overdriving that gate directly. The more I look at it, the more I wonder if FET's really offer a special advantage in laser drives.
The Vishay SUP45N03-13L would just about do if you have a bucketful lying around but don't go out of your way to buy one.
2.7nF is quite a significant input capacitance.
Don't get too hung up with 'linearity' of devices. To get linear performance you use the loop gain of an op-amp combined with a sense resistor.
A constant current source to the LD with a programmable current shunt will work just as well as a programmable current sink.
When properly designed only the advantage of having one end of the laser diode earthy distinguishes them.
The one advantage of MOSFETs is that they do not saturate in the same way as bipolars - so recovery from large signals is usually much faster.
You could build the die4drive circuit with a bipolar instead of the MOSFET but the performance would be awful.
Weird, I overlooked that ability of op-amps to act as a servo to impose its own linearity on a device controlled inside its loop. Doesn't a nonlinear device with a very narrow range between 'off' and 'on' risk a noisy signal though? That could limit how close you can go to a maximum safe drive.
Yes - it does but not as much as you would think.
The o/p voltage noise of the opamp ends up as noise current.
Inoise = Vnoise x MOSFET transconductance.
So for a transconductance of 20 mhos & an opamp with 100uVp-p noise on its output you would get 2mAp-p noise current.
Not an issue for a 300mA drive but it would be if you were trying to drive a 30mA load. Luckily if you use a smaller MOSFET for the smaller load the transconductance is an order of magnitude smaller too.
The main problem is stability.
You have to provide feedback in a couple of ways:
1) DC feedback from the current sense resistor - to control your o/p current.
2) AC feedback from the o/p of the opamp - for stability.
Without (2) the control loop is unstable & you have an oscillator which will likely burn out your diode.
Hi all
Originally Posted by uk-laser
That would be interesting, WHY...