PWM Laser dimmer

[Apocalypse]

Okay I'll take a few pics later. Thanks for the help!

 

[Cyparagon]

That's not an ideal way to power a laser diode. It's not constant current or even constant voltage. Hell, it doesn't even have a low dropout. To complicate matters further, the pulsing means you cannot measure the peak current without an oscilloscope. There's a good possibility this will just kill your diode.

 

[Apocalypse]

The thing is I'm powering the whole thing from an adjustable power supply. Also I got a benboost driver after it. Would that be ok?

 

[Apocalypse]

So here is the circuit

Here is it annotated (there are some wires beneath the circuit board)

Did I do something wrong?

The TIP31C is usually heatsinked but I took it off to make it easier to see

Also, oops, I annotated the NPN transistor wrong. It's supposed to be E C B instead of B C E.

The pot is 100K, the diodes are 1N5819s and the caps are 10nF caps (103).

 

[Cyparagon]

The thing is I'm powering the whole thing from an adjustable power supply

Then why not adjust the power supply instead of this elaborate fruit cakery?

 

[Apocalypse]

I plan to power this from batteries so it is portable later. I'm using an adjustable power supply as a precautionary measure right now. If I feed the square wave into a drive and then into a laser, isn't it equivalent to turning the laser on and off and the current would still be controlled by the driver? Also, as you can see I'm learning how to build circuit boards lol so any suggestions are welcome

 

[Cyparagon]

The DC current gain of that NPN looks to be about 50, so the maximum current it can pass is 50 times the base current. At the nominal 3.6V of lithium cell, the 555 drops 0.7V and so does the transistor, so your base current is (3.6-0.7-0.7)/10,000 = 0.2mA. Multiplied by the current gain of 50 means a theoretical 10mA can flow through it with a 3.6V supply. At fifteen volts I would expect somewhere around 750mA (although gain tends to be higher at lower currents like this). Decreasing the base resistor or increasing the supply voltage will allow more current.

This circuit doesn't translate well to battery operation. The timing of the 555 will change with the charge level, and so will the base current of the pass transistor. Not to mention your battery cutoff will be much higher due to the voltage drop.

Here's a better option: replace the pot on the benboost with a panel mount pot of the same value. It does have a pot, doesn't it? If not, get a driver with a pot. You're adding all this extra stuff which just decreases efficiency.

 

[Apocalypse]

I thought PWM was the way to go to dim circuits or does this circuit just suck? If I use a pot to dim the circuit, doesn't it generate a lot of heat and wasted power? If the power is the NPN transistor, are there better transistors for this function?

 

[Bionic-Badger]

For your circuit diagram, can you draw it out as a schematic? Trying to figure out what a circuit is doing from a photo of the wired circuit is a pain.

If I were to make a PWM laser circuit, I would make something like this. The max current is fixed and then all you do is feed it a control signal from whatever PWM source you use.

 

[Apocalypse]

The circuit diagram is http://afrotechmods.com/groovy/PWM_tutorial/PWM_tutorial.htm as it is posted above.

 

[Cyparagon]

I thought PWM was the way to go to dim circuits or does this circuit just suck?

PWM is not the best way to dim lights. Smooth constant light is better. If you smooth out those pulses and add feedback, you have an actual driver (buck and/or boost).

If I use a pot to dim the circuit, doesn't it generate a lot of heat and wasted power?

Yes, but you're wasting just as much with this design. That's not what I meant anyway. Replace the pot on the existing driver.

Control the circuitry that's already there, don't force it on and off with an unregulated system that won't fit in a pointer to begin with.

If the power is the NPN transistor, are there better transistors for this function?

Yes, low Rds mosfets would be better here. You'd still have an unregulated system though.



You're making the classic mistake of misunderstanding the root problem. See sig. You're asking "how can I get this pwm circuit working at higher current?" when you should be asking yourself something more general like "how can I dim this laser?"

A more extreme example is that guy who posted a thread asking "where can I find a visible beam <1mW laser?" when his actual problem ended up being "how can I keep non-employees from entering this area?" :crackup: He wanted to use a beam as a warning to people or something, yet the problem was entirely unrelated to lasers to begin with.

 

[Apocalypse]

PWM is not the best way to dim lights. Smooth constant light is better. If you smooth out those pulses and add feedback, you have an actual driver (buck and/or boost).

I don't quite understand what you mean here, can you elaborate? How do I dim a circuit then? Bionic Badger has provided me with a good circuit http://laserpointerforums.com/f51/simple-linear-mosfet-dimming-circuit-pwm-70812.html .

However, it also relies on PWM to dim the circuit unless I'm mistaken.

 

[Bionic-Badger]

PWM does provide the most linear method of dimming and is the reason why LED drivers utilize PWM for their dimming rather than direct current regulation. There are linear ranges of mW/mA for LEDs and lasers, but factors such as threshold current, mW/mA curves, etc. may result in non-linearities in the visual light output for some input values. For example, the threshold current for a laser diode defines the minimum light output the laser can produce (besides being off), and this may actually be pretty bright depending on the laser diode (like >100mW for a 445nm laser). A PWM signal, however, could have a duty cycle of 1% or less and that would be far less bright than the minimum threshold current output.

The key for PWM-based dimming is to ensure that whatever you're rapidly turning full-on/off with PWM has its maximum current set properly so that it doesn't fry your LED (laser) if you have your PWM at full duty cycle (100% on).

The circuit I linked you regulates the maximum current using feedback in the form of the BJT transistor connected to the MOSFET gate. It works by restricting the maximum amount of current to 0.58V / Rsense (the resistor on the bottom).

What makes that circuit more "safe" (as far as current-regulation is concerned) than the one you've built is that the maximum current is fixed, and won't change (much) regardless of the input voltage (the 12V on your diagram). The low-side current regulator is pegged to a ground reference--rather than the high-side voltage reference like the current-regulating resistor in the circuit diagram you provided. If you were to raise the voltage of your power supply, it would increase the current being supplied to your LEDs in your current design.

Do note that in the circuit I linked you, the MOSFET acts as a voltage-controlled resistor. It doesn't matter what its Rds_on is because it is always acting as a resistive component. The voltage drop across the MOSFET, multiplied by the current passing through it will be how much power it has to dissipate--which can be a decent amount. It is for this reason I usually only use the circuit I linked you in lab lasers that I want to have analog/PWM control over. If you want something more portable and energy efficient, you should look into a buck/boost type circuit.

 

[lasersbee]

I would also go along with BB's suggestions for solving your
requirements.

BTW... it seems that the original circuit asked for 1K (1000 Ohms)
resistors for R1 and R2..
I see the colors of your resistors as BRN/BLK/ORA... That is 10K
(10,000 Ohms)...
This may also be a reason why your output current is very low.
R2 should be 1K or the TIP31 will not conduct enough.


Jerry

You can contact us at any time on our Website: J.BAUER Electronics

 

[Cyparagon]

BB's circuit looks okay. Except of course the timing will still change with battery charge if you use a standard pwm generator. That's not what you're building though. Either build his circuit or for the third time, use the pot on your benboost.