Page 2 of 5 FirstFirst 12345 LastLast
Results 11 to 20 of 44

Thread: Analog laser diode drivers?

  1. #11
    Join Date
    Mar 2007
    Location
    Germany
    Posts
    3

    Default

    We don't sell no more. The project BlogDrive is frozen. We work on other projects at this time and have no time for this.

    And the speed is enough for a 2,5A driver.

    But we start an open scoure analog laserdiode driver (probably only >500mA but enough für DVD-LD) soon.

    Sorry for my terrible english but my german is better

  2. #12
    Join Date
    Apr 2006
    Location
    Wirral, UK
    Posts
    191

    Default

    If you want to drive a DVD writer diode the standard die4drive works well.
    *sound of own trumpet blowing*
    http://www.die4laser.com/dvd-rec/Die...riveRev1-2.pdf
    Rev 1.2 has a soft start, limiting modulation to 30kHz.
    It also has an offset adjust to set threshold current.
    This gives light out proportional to input voltage.
    Removing the soft start capacitor you can modulate to 200kHz but this is not recommended as it may reduce diode life.
    Bare PCBs and populated boards are available.
    Regards,
    Robin
    Robin.

  3. #13
    Join Date
    Mar 2007
    Location
    Germany
    Posts
    3

    Talking

    Your die4drive is a nice work...

    And a friend has a die4thing. very impressive. *thumpsup*

  4. #14
    Join Date
    Dec 2006
    Location
    Pflugerville, TX, USA
    Posts
    1,977

    Default

    Robin - What are you selling the die4drives for now? (populated and unpopulated)

    Can you comment on your design a bit... If I am not mistaken you have the diode in series with the modulating transistor which will cause the P/S load to fluctuate. Other designs have the modulating transistor shunting the laser diode so that P/S current is always constant to keep the load steady. I am not an electronic engineer so I don't really know if it is true but I have heard that the shunted diode is safer. Is that a bunch of hooey?

  5. #15
    Join Date
    Jan 2006
    Location
    Charleston, SC
    Posts
    2,147,489,446

    Cool

    Carmangary;

    You are correct. Marconi was talking about this a while back. The Die4Laser driver works well, but you don't want to push the diode really close to it's limits using that design. If you do, a little current fluctuation in the modulation stage might be enough to kill it.

    Marconi's Maxyz modules use a different arrangement, plus they also have a hard current limit built into the output stage of the modulation circuit. (This is separate from the current limit of the main driver.) It's more expensive that way, but you can modulate the diode at 50Khz right up to the max current and not worry about killing the diode. Of course, back-reflection is still a concern, but the power supply design isn't going to help you there.

    Bottom line: high speed modulation is risky if you're operating right at the limit of the diode. If you are willing to dial back the current a tad (and loose some power in the process), then your modulation stage can be fairly primative. But if you absolutely need all the power you can get, you want to be very careful with your modulation stage.

    Adam

  6. #16
    Join Date
    Mar 2006
    Posts
    2,478

    Default

    Many diodes are run CW at strong ratings. As most DVD diodes have pulsed ratings for 240 mW and such, modulation isn't what weakens them when pushed hard, as if the CW doesn't kill them, the modulation certainly won't unless the transients are very fast. That said, there are likely stressful combinations of near-CW pulse width ratio and actual mod speed even with slower transients, but those can kill a diode quite independently of any kind of driver design. Bad drives might make it worse, but the best can't avoid this fact.

    The real risk is overcurrent due to overshoot (and brief oscillation) in any inductance or capacitance present. Any driver has this risk, even a shunt modulator, though with those the risk is less, especially as the regulator will likely respond faster than the active modulating device, correcting the overshoot, because the modulation speed is always restricted anyway, as part of a compromise between mod speed and transient protection for the laser diode. As a guide, an LM317 can be a shortwave RF oscillator! That should tell you that it can more than cope with overshoot if mod speeds are limited to 30 KHz or less.

    In practise, you only need ONE hard current limit if you use a shunt mod. The only other precaution (apart from the usual extremely clean supply) is to avoid inductances and capacitances that can cause over-current pulses.

    Don't just take my word for it. Study that Laserbias.htm I linked to, first post. That guy gets by with mod speeds capable of video transmission, apparently. His design is a tad messy-looking, and lacking one or two finer points of detail, but don't let that fool you into thinking it's a bad design, it's not.



    Edit: Buffo, a close look at Robin's circuit shows it's safer than you suggest. The LD is given a 5V regulated supply, so current is dependent entirely on the assumption of a short where the MOSFET is, followed by a current limiting resistor. That's your hard limit. Then the MOSFET resistance (varied from near zero ohms to >300 Mohms) can only adjust that current downwards. It's free of problematic capacitances, the only ones being those across the diode to protect it. There are no inductances other than the tiny ones in the MOSFET and the connecting wires, and being in series with the load and a current limit resistance, they will be very well damped. I prefer a shunt mod, but that's just because the logic of it is nice.

    EDIT again:
    Last edit was wrong. That's a sense resistor only. I'd include a resistor as I described, though, as a hard limiter. Probably on the supply side of the laser diode.
    Last edited by The_Doctor; 08-06-2007 at 00:45.

  7. #17
    Join Date
    Feb 2005
    Location
    Florida
    Posts
    802

    Default

    Quote Originally Posted by Buffo
    You are correct. Marconi was talking about this a while back. The Die4Laser driver works well, but you don't want to push the diode really close to it's limits using that design. If you do, a little current fluctuation in the modulation stage might be enough to kill it.
    Actually what I meant . He didnt have any protection on the input..where overvoltage (more than 5v) on the input where someone may tie the mod line to the input supply rail (+7 to 12v) being more than 5v will cause the driver to exceed the set current. this is only true if you are using a 5v reg on board and feeding the reg outboard as most will dol with + 7 to 12volts.
    Otherwise, It is the best circuit I've seen.

    Quote Originally Posted by The_Doctor View Post
    Many diodes are run CW at strong ratings. As most DVD diodes have pulsed ratings for 240 mW and such, modulation isn't what weakens them when pushed hard, as if the CW doesn't kill them, the modulation certainly won't unless the transients are very fast. That said, there are likely stressful combinations of near-CW pulse width ratio and actual mod speed even with slower transients, but those can kill a diode quite independently of any kind of driver design. Bad drives might make it worse, but the best can't avoid this fact..
    It most definitely is...because you are close to the diodes operating bounderies . The best drivers are designed to avoid
    transients by using negative feedback to overcome these. The load (diode) is a resistive and capacitive load , there will be some reactance that needs tended too.. Take another look at Robins driver. He has my vote.
    But you are correct, most applications lately stem from overdriving the DVD diodes.

    The real risk is overcurrent due to overshoot (and brief oscillation) in any inductance or capacitance present. Any driver has this risk, even a shunt modulator
    Very true

    though with those the risk is less, especially as the regulator will likely respond faster than the active modulating device
    not that I have seen with a scope

    because the modulation speed is always restricted anyway, as part of a compromise between mod speed and transient protection for the laser diode.
    Actually , why comprimise...the diode has an almost unlimited bandwidth.

    As a guide, an LM317 can be a shortwave RF oscillator! That should tell you that it can more than cope with overshoot if mod speeds are limited to 30 KHz or less.
    The LM317 may well oscillate at HF but that is open loop bandwidth. and not much more than several megahertz, This was done on purpose to keep the regulator from becoming an oscillator while used as a regulator. It also has poor load regulation.. about 1 percent. I do like them as simple regulators, tho.
    Ive built many supplies from them, even mod'd them for below 1.25v capablity.
    The devices that Robin and I use are rated at least 10x that. Mine are good to 120mhz. Good for video and any sidebands you are likely to throw at it.
    But, Using faster devices require more attention to reactance. but they are more easily controllable, if you use a limited bandwidth device, you have limited control.
    Dont forget, the LM317 is made up of bipolar junctions. We use mosfets. much faster. The main problem is not having negative feedback to control reactance while using the LM317. At least in any circuit Ive seen..Current yes, but not frequency compansation.

    In practise, you only need ONE hard current limit if you use a shunt mod. The only other precaution (apart from the usual extremely clean supply) is to avoid inductances and capacitances that can cause over-current pulses.
    I have found that pre-limiting makes it easier to control overshoots., rather than rely on only one device. Why tax the driver.
    I also had problems using the shunt style with being able to react fast enough
    to heavy loads. Thats why you dont see commercial manufacturers doing it this way.

    Don't just take my word for it. Study that Laserbias.htm I linked to, first post. That guy gets by with mod speeds capable of video transmission, apparently. His design is a tad messy-looking, and lacking one or two finer points of detail, but don't let that fool you into thinking it's a bad design, it's not.
    I have tried all of Johns circuits. Also, I know him well..as we share the same ham radio hobby. those circuits while simple..dont have enough control..they can be dangerous to the unexpected hobbiest. Just not enough control.
    But I do suggest reading and making a few as a starting point.

    Buffo, a close look at Robin's circuit shows it's safer than you suggest. The LD is given a 5V regulated supply, so current is dependent entirely on the assumption of a short where the MOSFET is, followed by a current limiting resistor. That's your hard limit. Then the MOSFET resistance (varied from near zero ohms to >300 Mohms) can only adjust that current downwards. It's free of problematic capacitances, the only ones being those across the diode to protect it. There are no inductances other than the tiny ones in the MOSFET and the connecting wires, and being in series with the load and a current limit resistance, they will be very well damped. I prefer a shunt mod, but that's just because the logic of it is nice.
    There are some inductance / capacitance and resistive reactance cuased by the diode, while Robins circuit addresses most of this..he has made it somewhat universal..it still is diode dependant..Each diode type will react differently, It is up to the user to account for the rest.
    My drivers are made to match the diode, that is why you dont see them advertised for sale.

    I've been making drivers for lasers more than 5 years now...There isnt much out there that I would trust.
    even some of the commercial stuff out there just dont cut the mustard for us lasershow freaks.
    A scope is your best friend.
    Last edited by marconi; 08-05-2007 at 09:51.
    "My signature has been taken, so Insert another here"
    http://repairfaq.ece.drexel.edu/sam/laserfaq.htm
    *^_^* aka PhiloUHF

  8. #18
    Join Date
    Jul 2007
    Location
    Melbourne, Australia
    Posts
    70

    Default

    This discussion has become somewhat too technical for me , but I get the basic idea though. So, is it safe to say that Robin's LD driver is good and safe enough to be used and modulated at around 30KHz? Are the previously mentioned LD drivers actually better than Robin's? Thanks guys for your help.

  9. #19
    Join Date
    Feb 2005
    Location
    Florida
    Posts
    802

    Default

    Yes, To be safe, I would recommend Robins driver.
    Get one for each diode you plan to use.

    There are none that I know of that I would trust better to someone just starting out.

    A commercial unit like a Wavelength Electronics are very expensive and require hands on experience..also modification.
    "My signature has been taken, so Insert another here"
    http://repairfaq.ece.drexel.edu/sam/laserfaq.htm
    *^_^* aka PhiloUHF

  10. #20
    Join Date
    Mar 2006
    Posts
    2,478

    Default

    I agree, a scope is vital. I'm using one now to look at a shunt using an IRF630 MOSFET. It's not ideal, for a start, it switches, it's no good for linear. Also it has higher gate capacitance than is ideal. This helps to show something though...

    A scope does show that a regulator can recover fast enough to avoid overshoot, but it seems dependent on not overdriving the gate hard, so as you say, pre-limiting is helpful. There are overshoots at the onset of shunt too, at gate voltages just capable of turning off the laser, but those are ok, they just switch off the diode harder, and that's ideal for precisely timed blanking.

    I wouldn't use a Wavelength electronics unit. Maybe if I had money to burn and if every diode cost $$$ I might, but not otherwise, it's just as easy to build from raw parts and see what works with the diodes you have.

    You mention diodes' inductances, but I think the MOSFET's have more problems with their gate capacitance, and the other characteristics of the MOSFET will have more effect than the diode, I think. I'm going to try to find some others to try though. RS were out of IRF510's at one point and recommended the IRF630 as replacement so I got a load really cheap on eBay, but they're not really compatible at all. For basic switching, they might be, but the differences (mainly linearity, capacitance) really affect a laser modulator.

    I suspect an additional stage of isolation with an optoisolator really helps too, to prevent an overdriven gate causing spikes. I intend to try optoFET's and bought twenty, but Fairchild apparently decided to make no more, right after I bought them. NOT helpful, and there was no warning either, they'd just made them RoHS compliant, so there was plenty of reason to assume they'd continue. Any optoisolator will do though so long as it's stable over a wide range of input drive. They also make it easy to use balanced ILDA signals, and to isolate from static. Most simple diode drivers I've seen don't use one, and I think they're missing out by not doing so. The only drawback I can think of is that they're current driven, not voltage driven, so maybe not ideal on long signal cables unless you use the Darlington output types.
    Last edited by The_Doctor; 08-05-2007 at 11:05.

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •