why not LM317 instead of Flexdrive?

I was just browsing around and happened to stumble upon some 1.25A - 1.5A 445s on here and noticed a bunch of people debating what to use for drivers. Some were saying why not use a LM338 or LM317 instead of, for example, a FlexmodP3.... the argument was mainly that the driver had other safety features in it that were helpful.

What other safety features are we talking about here because 40 bucks and a lot of controls that arent typically used on handhelds such as balance and (i forget the other adjustable feature). I know it has some sort of ESD feature, but this is my point: I've heard that the LMXXX series regulators have a built in shut off if they over heat, so that is in some form a safety feature in regards to the LD right?

Not only that but the constant current from a LM317 or LM338 is 1.25A when using a 1 ohm resistor. Does that mean that a small DDL board with this setup would be safe for a decent powered handheld??? Or is there just something extra special about these drivers that I'm missing?

(no idea what capacitor(s) would be idea bc I still dont understand the function)

The driver has build-in capacitor, diode and a pot and it is small enough to even fit the Axiz module. None of this has the LM317.
It can be done with a LM317 but you need other components.
http://www.rog8811.com/laserdriver.htm

sometoymachine said:

I was just browsing around and happened to stumble upon some 1.25A - 1.5A 445s on here and noticed a bunch of people debating what to use for drivers. Some were saying why not use a LM338 or LM317 instead of, for example, a FlexmodP3.... the argument was mainly that the driver had other safety features in it that were helpful.

Click to expand...

Do you have any links to that shows people discussing this?

sometoymachine said:

Not only that but the constant current from a LM317 or LM338 is 1.25A when using a 1 ohm resistor. Does that mean that a small DDL board with this setup would be safe for a decent powered handheld??? Or is there just something extra special about these drivers that I'm missing?

(no idea what capacitor(s) would be idea bc I still dont understand the function)

Click to expand...

The LM series current regulators used in conjunction with a diode (to protect against reverse polarity) and a capacitor (to help protect against current spikes) will work just fine for a decent powered handheld IMHO.

The only downfall I can see is size. It is difficult to find a host with enough room to fit a non-SMD LM series based current regulator.

I don't know why I can't seem to find it now :-/

I have noticed though that the flexmodP3 is pretty large and ends up using a typical resistor for jumping v+ and m+ since modulation wasn't used on the particular pointer I saw in discussion. Also has three pots and what seems to be a decent size capacitor on the board. If packed tight enough the DDL would likely be relative in size I would think.

For some reason I'm wanting to say that it was Jayrob's 2D Maglite host that was
The host in the discussion.

The flexmodp3 is a modulated driver intended for scanner use, etc. If your not using the modulation there isn't any reason to use it. It just adds size and cost.

I personally have never used a flexdrive/flexmod, but usually a basic LM317/w/e is a linear regulator.

One major disadvantage IMO is the fact that a LM317 based driver alone cannot boost the voltage such as a boost converter or other switching drivers do.

(I think this is what microboost driver does via a boost converter) using a boost/switching driver has allowed me to have a low voltage supply and still power my diode (did this for an A-140 @ 4.5V) while running from a 3.7V battery.

I have used the LM317/350/338 before in the past, and it works fairly well for lab builds.
but has some downfalls as stated earlier by others, is size, efficiency, and limited to what you can do with just an LM series regulator.

I hope this helps you out, and if anyone finds errors here, feel free to point them out.
:beer:

EDIT:

for the LM series, and a lab, I usually use this circuit:
which is nearly identical to the DDL, so far Ive had no issues with it, but only used it for the A-140 diodes.



quickly through the parts:
D1 - diode to prevent me from killing the whole thing through reverse polarity connection.

TVS - Transient voltage suppression diodes, used for ESD/spike prevention (I used bipolar ones, but unidirection in this schematic would work too.)

R1 - the set resistor to adjust current
R2 - Bleeder resistor

C1 - soft start capacitor (shows ramp on scope), also helps against spikes as stated before in the thread.

the transistor doesnt need to be there, but I put one there since I use it to do PWM sometimes.

Not sure what PWM is or a bleeder resistor, but you mentioned a boost driver boosts voltage. Do they affect current as well?

sometoymachine said:

Not sure what PWM is

Click to expand...

Gah, I hate MS paint, but hey.. it works..



PWM is an acronym for Pulse Width Modulation. It is basicly turning on and off the power to the diode, very rapidly.
for example:

I took this scopeshot (oscillioscope) of one of my drivers so you can see the waveform of the raw PWM signal, my purpose was to limit the intensity of the diode, since I wanted different brightness levels, so instead of analog control via voltage/current, I had a constant current setting, and different PWM modes (by varying the duration of Ontime/offtime) to control when the diode will turn on and off, and for how long. Note this is happening faster then your eye, so the laser just appears dimmer, If I where to change the frequency low, the it will appear as a strobe.

in the circuit, the signal, which is a square wave is fed into the transistor turning it on and off when its on, it will conduct the driver to ground and when off, wil act as normal and supply the diode with power.

sometoymachine said:

bleeder resistor

Click to expand...

A bleeder resistor is a resistor that is placed in parralel with a capacitor.
its a resistor that will drain the capacitor's left over charge stored in it.
basicly an alternative to manually shorting the capacitor everytime.

sometoymachine said:

but you mentioned a boost driver boosts voltage. Do they affect current as well?

Click to expand...

Yes, current can be controlled with boost drivers, simular to what I said about PWM earlier, although there are ICs there to do it for you, you can control the current output by the duty cycle on the switching device for a boost circuit.
the duty cycle you see in the scope shot is 50%, meaning the On time = Off time.
varying this duty cycle will change how long the switching transistor, IGBT, etc, will turn on/off.

hope that clears some things up.

Definately new concepts to me but very interesting. The most Useful to me and my simple design would be the bleeder diode bc the DDL model doesn't include this. I know that after the test load, shorting the capacitor is required before connecting the LD. However, does it need to be shorted after every power off? If so, how would this diode be integrated into the circuit-- just as the diode pre-capacitor?

sometoymachine said:

Definately new concepts to me but very interesting. The most Useful to me and my simple design would be the bleeder diode bc the DDL model doesn't include this. I know that after the test load, shorting the capacitor is required before connecting the LD. However, does it need to be shorted after every power off? If so, how would this diode be integrated into the circuit-- just as the diode pre-capacitor?

Click to expand...

Bleeder diode?
you mean bleeder resistor?
well, anyway, the resistor is connected directly across the capacitor, its value you will need to calculate to suit your own needs. the capacitor should be always drained before connecting an LD, as a charged capacitor can give a bit of a spike, even in the slightest to damage your diode, while I never had this happen with an A-140 (tough diodes), it will probably harm some of the more sensative ones.
generally upon the first connection to the diode, you need a drained capacitor, after that, you dont need to short it after everytime you power up your laser. I just do to be on the safe side, as I disconnect my diodes alot.

the bleeder effectivly rids of this charge without having to short them yourself. but again, I am speaking in terms of blue diode experience, Ive yet to try this concept with other LDs.

the diode? Ill assume you mean the TVS one I drew, I solder actually 2 diodes usually, 1 TVS diode on the output of the driver board, and one diode already soldered right on the laser diode's pins (small SMD TVS).

the TVS on the driver takes care of spikes
the TVS on the diode takes care of spikes and ESD (TVS diodes, will conduct and act as a short once you exeed thier "clamping voltage", a zener will work too in stead of a TVS, but TVS is far faster, can catch those very quick spikes and transients zeners will miss.)

if you clear up a bit to what your asking, I will try to provide a more accurate answer, I didnt really understand your question. unless I just answered if offcoarse :yh:

sorry about that.

Yes I meant resistor. And the second time I mentioned diodes, I WAS referring to the TVS as you described it. That clears up a lot. I have never heard of using SMDs on the LD itself. Clever. Thats assuming that something in the driver fails to regulate the voltage correctly I guess.

When using home built or non-standard drivers (multi-mode LED flashlight drivers for example), extra protection never hurts.

sometoymachine said:

sorry about that.
Yes I meant resistor. And the second time I mentioned diodes, I WAS referring to the TVS as you described it. That clears up a lot. I have never heard of using SMDs on the LD itself. Clever. Thats assuming that something in the driver fails to regulate the voltage correctly I guess.

Click to expand...

its a bitch to solder on, but thats the general idea. I think a 0.25 cent part is worth investing for a 50$ diode
most of the drivers I use are home-brew boost converters, even with both regulation and on-board protections, a little extra as said doesnt hurt.

when you pick out TVS, just make sure the clamping voltage and reverse recovery voltage is just slightly higher then what you will be running your diode at, else youll blow out a TVS.

TVS are mostly designed to be used in transient mode, a CW activation of a TVS by overvoltage, especailly with some decent current behind it, will pretty much blow an SMD TVS diode in seconds.

(I tested out my last TVS I used for my build, at 5V the diode leaks some current through it, 6V the TVS diode fully conducts, at 6V 1A, the diode split in half in about 2 seconds. the laser diode was bieng run at 4.5V, where the TVS did not conduct.) these types of conditions are dependant on whatever the datasheet specifies.

main reason I see benafit in TVS diodes soldered on the LD is ESD protection.
the spike/surge suppression is a bonus.

If you have any more questions, id be happy to address them.
:beer:

GBD said:

I personally have never used a flexdrive/flexmod, but usually a basic LM317/w/e is a linear regulator.

One major disadvantage IMO is the fact that a LM317 based driver alone cannot boost the voltage such as a boost converter or other switching drivers do.

(I think this is what microboost driver does via a boost converter) using a boost/switching driver has allowed me to have a low voltage supply and still power my diode (did this for an A-140 @ 4.5V) while running from a 3.7V battery.

I have used the LM317/350/338 before in the past, and it works fairly well for lab builds.
but has some downfalls as stated earlier by others, is size, efficiency, and limited to what you can do with just an LM series regulator.

I hope this helps you out, and if anyone finds errors here, feel free to point them out.
:beer:

EDIT:

for the LM series, and a lab, I usually use this circuit:
which is nearly identical to the DDL, so far Ive had no issues with it, but only used it for the A-140 diodes.

quickly through the parts:
D1 - diode to prevent me from killing the whole thing through reverse polarity connection.

TVS - Transient voltage suppression diodes, used for ESD/spike prevention (I used bipolar ones, but unidirection in this schematic would work too.)

R1 - the set resistor to adjust current
R2 - Bleeder resistor

C1 - soft start capacitor (shows ramp on scope), also helps against spikes as stated before in the thread.

the transistor doesnt need to be there, but I put one there since I use it to do PWM sometimes.

Click to expand...

Hey GBD,
In regards to PWM.. I have a couple of different PWM drivers on my breadboard, at the moment, but my implementation is different than yours.

The PWM transistor you're using.. just a regular NPN?

How are you driving the PWM transistor? 555, ect? Can you post a schematic of your pulsing circuit?

If I'm understanding this circuit correctly... The LMxxx is always on and the current is dumped to ground when the NPN is ON, creating an OFF state for the LD?

My implementations, which are probably the wrong approach, use either a PNP or NPN to modulate the V+ to Vin or GND to all GND connections. The advantage I hoped for was far better efficiency as the driver is completely off during the OFF state and obviously less waste heat. I have a distinct feeling that my circuits won't look nearly as nice on a scope.. I'll find out next week. At low speed, it works for sure and as expected.. I just don't know at what speed things get muddy.

Thanks in advance for any info you have!!!

Hello,

sampson said:

The PWM transistor you're using.. just a regular NPN?

Click to expand...

The transistor is a regular NPN, PNP will not work. also, N-channeled deviced such as mosfets/igbts will also work, but are not very practical as IGBTs have slow turn off, and mosfets (well, both really) will require some decent gate charge to actually turn them on.

an NPN transistor is a simple way to this, they are fairly easy to drive.

and your right, when the transistor is on, it feeds the output of the driver to ground, putting the LD into off state, when the transistor is off, the LD recieves its current as normal.

sampson said:

How are you driving the PWM transistor? 555, ect? Can you post a schematic of your pulsing circuit?

Click to expand...

I use 555s for simple and quick setups.
you can use a uC if you want, but I find a pair of 555s more then enough for basic modulation.

I drew up a circuit real quick:



EDIT: fixed schematic

its fairly straight forward, a pair of 555s, far one on the left is set as a variable frequency oscillator, the second set for pulsewidth control (triggered by the first one)

an External PWM source can be from a uC or anything else.
the External RST is a reset function that causes the driver to output nothing if its enabled.. neat thing for a sequence strobe effect, although requires an additional oscillator.

the schematic I posted is not accurate on what values of resistance/capacitance to use, as I was using this for other things before lasers.

sampson said:

My implementations, which are probably the wrong approach, use either a PNP or NPN to modulate the V+ to Vin or GND to all GND connections. The advantage I hoped for was far better efficiency as the driver is completely off during the OFF state and obviously less waste heat. I have a distinct feeling that my circuits won't look nearly as nice on a scope.. I'll find out next week. At low speed, it works for sure and as expected.. I just don't know at what speed things get muddy.

Click to expand...

well, generally you can modulate the driver itself to turn on, yes, the whole thing will be off, but I dont think the LM series handle pulsed modes very well, and will probably cause spikes upon turn on. this is just a theory, I havent tried modulation this way.

generally the faster you go, the more chance of ringing and spikes to occur, im almost certain directly turning the driver on and off will manage to wreck it past a certain frequency.

this is why i opted to just short the output to ground via a transistor, and have the driver constantly running.

sorry for the thread jack there...