Soft Start DDL?

[Meatball]

I'm interested in finding a way to make a soft start DDL driver with the typical LM317. I've searched the forum, but I only got results for the flexdrive v4 with that particular feature.. But I'm curious as to how it's done on the component level. Anyone have any schematics, or ideas as to how this might work?

Thanks

-Tyler

 

[pseudolobster]

Put a really big capacitor on the output maybe?

I've got a 2.5v, 1f aerogel cap I salvaged from a dead xbox and I can confirm it takes a several seconds to charge from an AA battery.

 

[digital_blue]

IIRC, as capacitors charge, their "resistance to current flow" increases. Perhaps a PNP transistor biased with a cap might work? In theory, as the cap charges, the PNP transistor(Normally closed) will allow more current to flow as the capacitor's resistance to current increases; eventually your cap will charge completely and your transistor should be driven beyond linearity(It will basically behave as a switch). Don't quote me on anything though...

EDIT: Looks like I'm right...
Here's a link that might provide more info:

http://www.zen22142.zen.co.uk/Circuits/Power/softstart.htm

 

[HIMNL9]

For a ramp, you need also to put a resistor in serie to the diode, before the capacitor, but this then affect the diode current and dissipate power

I'm trying to develope my own driver with softstart and protections, but NOT with an LM317, anyway, that is a linear regulator.

Actually, i'm testing a 3400 and a 3429, and waiting for see if i can get also a 6200 and a 3202, for some other tests ..... already tried a 4538, but is dangerous, cause if the diode get disconnected, the output voltage rise a lot, also if it integrate a current controller that isn't so bad (but get components here is a PITA ) ..... ofcourse, for get a decent driver that can use a single LI battery for power also blue rays

I've seen that there's something similar here, from drlava, also if i don't know what regulator it's using, but it's not exactly cheap, as DDL ..... and, more of any other things, i'm a DIY maniac

 

[pullbangdead]

So what's up with "soft start" for laser diodes? I don't really get it, to be honest.

Diodes, in normal operation, are pulsed in optical storage drives, and they are pulsed HARD. They're pulsed to higher powers than in CW, and they're pulsed at very high rates. They go from zero to max power and back to zero in less than 1 millionth of a second, and people somehow think that they need a multiple second ramp-up time for their laser diode?

For example, here: http:
//www.laser66.com/html/6146-301.pdf. It's only rated to 50mW CW, but it'll run up to 100mW total when pulsed, with a pulse width of < .1 microsecond. The whole pulse is less than a millionth of a second, and it's still rated to run twice as high doing that half a million times a second as it is rated to just leave it on.

Maybe you're using these "soft starts" for things other than laser diodes?

 

[Benm]

I've always figured the same thing really - but soft start somehow feels 'safer' for no good reason

There is one thing that has merit though: The optical output of laser diodes decreases with temperture. So when you start the diode at full current while it is still cold, it will produce more light than after it has been running for a while. This brief period could just cross the damage treshold... and soft start would warm it up a little before reaching full power.

If that is the case however, i'd think that such laser would be just as easily damaged by operating it in colder ambient temperatures.

 

[HIMNL9]

Mainly, as soft-start, i use 1 second or so ..... principal reason is for avoid "spikes" at the start moment, that if too high, can damage laser diode .....

Softstart it's not my main goal, anyway ..... i'm trying to build something that actively check the load current and prevent diode burning for that overcurrent, more than other ..... linear regulators are ok, if you have a battery with a voltage higher than the needed, and use them in regular way and well filtered, but for use a single cell, that is less the voltage of the diode, and portable use, you need forcedly to use a boost or step-up module, and very few of them have current regulation ..... basically, as far as i've understood, all that you can do, til now, with existing step-up modules, is set a fixed current with a dummy load, and then soldering the diode in place and hope ..... but, as example, what if you use a wrong dummy load that cause a voltage drop of 3,5 / 4V (like the one with 6 diodes), and then use it on a power IR or red diode with a voltage drop of 1,3 / 1,6 volt ? ..... without active current monitoring, probably your diode say you "bye bye" in very few time

Still experimenting, anyway, and i don't have too much time for that, cause i also run my own pc shop, in the meantime (damn real life, that always interfer in our online time / hobbies with boring problems like bills, food needs, and taxes )

 

[Benm]

There are switching converters with constant current mode, or ones that can at least be wired to work like that. Most switched designes have outputs with a little bit of noise on them however, fluctuating below and above the current set between switch cycles.

Boost regulators are inherently soft-start since they usually need a few switch cycles to bring the output voltage to the desired level anyway. Most designs can step up the voltage, or step it down, but not both using the same layout. If you wanted to power a laser from a single lithium cell, you'd need boost for bluray, and buck for ir/red... though linear is probably easier for the latter.

 

[HIMNL9]

Yes, linear for a red/IR one is more easy to build ..... i'm doing this for blu-ray, anyway ..... got some decent result with a 3429, but still not got how to make a self-regulated current driver, with it ..... still experimenting, anyway

 

[Bionic-Badger]

I like the soft start because I don't want the laser going off when I, for example, am replacing batteries and the case bypasses the switch and lights up the laser momentarily. I've had that happen before and it never really occured to me that I should wear protection goggles while changing batteries.

If you're hell bent on voltage regulation, you can use the same LM317 driver to regulate voltage as well. The only problem is that the second LM317 will drop another 2-3V, which may not play well for your power supply. The examples in the LM317 datasheet show you how to do that.

 

[Meatball]

thats a good point BB..... it could be the difference between sudden accidental blindness, and "whoa that was a close one, good thing i saw it coming!" lol

But for real, do we have any sample circuits on schematics out there that i could try? I'll be testing with my blu-ray test load and a multi-meter.... just in case you guys are REALLY bad with circuitry or something... but lets see some ideas on how to make this work!

digital blue, I'll take a closer look at that link you provided, thank ya!

 

[Meatball]

K it looks like that particular circuit is designed around a soft start when the IC is configured as a voltage regulator. I'm assuming it still has a slower start to the current as well, but I'm looking further just in case..

Now, I DID find this youtube video of an LM317 being used as a slow starting regulator... I've already PM ed the guy and asked him about the schematic he used.. no clue when or if I'll here back from him though, its not like people regularly check their youtube inbox..

http://www.youtube.com/watch?v=7Zn2766OkDQ

 

[Benm]

Example circuit.. well, here is one: http://www.fancon.cz/led-lamp-driver/led-lamp-driver.html

Circuit seems sound as a constant-current boost regulator, text below it is in czech, so i'm lost there as well... it uses the commonly available 34063 switchmode regulator ic.

R3 is the current-setting resistor in this circuit, and should have 1.25 volts over it, so value calculation exactly matches that of lm317 setups (10 ohm = 125 mA, 5 ohm = 250 mA and so on). Note that it is a boost converter, so it can work a blu ray from one lithium cell, but not from 2!

 

[Meatball]

the circuit looks only a little more complicated than our typical ddl, however, I'm not sure my local radioshack carries 150uf inductors! :/ I never found where it was labeled as a soft starter, but perhaps that simply a well- known characteristic of the 34063? Does the "switchmode" ability refer to a soft start?

google translator reads the text below like this:

The rapid development of LED technology is reflected in the production of integrated exciter, often referred to as LED Drivers. The market is a great many different integrated circuits for this purpose. There are circuits for alarm serial or parallel cascades LED Step-Up and Step-Down Converter, charge pumps. However, when you need them to implement any involvement, and usually encounter unexpected problems. Many energizer LED (LED Drivers) is designed for a specific range of consumer electronics and often lacks the characteristics required of us. To them can be included and any requirements for precision settings propustného current diodes (Forward Current), the possibility to easily change the number of connected LEDs, a wide range of voltage, resistance to failure in the disconnection of the load and short circuit. Finally, they selected about each, but there is opposite the fundamental problem. This is the availability of circuits, the datasheets can be found easily on the Web and great number. The vast majority of these IC LED drivers in individual quantities of virtually unavailable.

This documentation may be freely used only for non-commercial purposes. AUTHOR PROVIDES THIS DOCUMENTATION "AS IS", without WARRANTIES AND ASSUMES NO LIABILITY FOR ITS QUALITY, FUNCTIONALITY OR ANY DAMAGES FOR USING presented facts. RECOVERY COMPENSATION the author is therefore eliminated in full.
Object:

When construction UV LED lamps for fishing Scorpions (Flashlight for Scorpions Hunting) I had a basic requirement to run a serial cascade UV LED (LED cluster) at the maximum permitted current diodes. It was me on the warranty service with a maximum allowed continuous power LED, the LED efficiency criterion energizer not paramount. When I used the construction of good, available and cheap circuit MC34063. Scheme is the involvement of step-up converter, sometimes called Boost converter. Involvement is suitable for UV - Ultraviolet LED and White LED, but it can obviously be used for virtually any LED. For example, in the case of ordinary red LED can be connected to 16 diodes in series. Converter - converter is able to operate at about 2 - 4 battery cells or 3 - 4 NiMH batteries. For supply voltage less than 3V, it is no longer guaranteed performance management circuit and the converter efficiency significantly decreases. It was me to maintain maximum light performance in the broadest possible range of supply voltage and the strong decrease of the voltage sources. LED current stability is not normally provide limiting resistors. The circumference of the converter with the MC34063, therefore, provides an increase in tension at the optimum value as well as effectively manage the flow, through the LED.
Function description:

Involvement MC34063 as spínaného Step-Up / Boost converter with a high inductance is described in the documentation circuits. Change is in that district in that engagement does not work as a source of tension, but as a source of power. Output current, which feeds cascade LED (LED cluster), is designed for the value of resistor R3. It is therefore possible without any přenastavování connect to different numbers of LEDs in total tension decreases (Forward Voltage) LED size 27V. Because this type of converter is generally short-circuit-proof, I added in case of circuit failure or short-circuit output overcurrent protection. Composed PolySwitch fuse at position resistor R1. For the role of converters, however, this policy is unnecessary and non-linear element PolySwitch is possible to replace the appropriate resistor values. Even slightly increase the efficiency of the device. Neither the zener diode ZD is not necessary, its role is to protect the converter against overvoltage in the case of the load - LED. Capacitors C1 and C2 provide the blocking voltage, capacitors C4 and C5 filtration output voltage.
Used parts:

Circumference MC34063 there in the implementation of a DIP or SO-8. If you do not need to miniaturize, probably reach the MC34063AP1. R1 is PolySwitch RXE 050E 0.5 A / 1.17 Ohm as protection against short circuit. May be replaced by about 0.2 Ohm resistor. If this value is not the resistance, can be used without problems parallel combination of two 1 ohm resistors in miniature design. The resistors R2 and R3 are the little bear, it is appropriate to use metallised version. Inductance L is realized by radial chokes FASTRON 09P-151 with the 150uH / 0.32 ohm / 700m. Rectifier diode D must be sufficiently fast and have the smallest loss of voltage. Is readily available 1N5819 Schottky diode, and matches. Capacitors C1, C3, C4 are normal - ceramic. C2, C5 are electrolytic capacitors with low internal resistance (low ESR). If you are unsure, as an indicative guide it is possible to check the temperature range listed in obtaining a capacitor. Those with lower losses are usually indication of the body temperature at least 105 degrees Celsius. Z-diode is 27 V and power dissipation at least 2W.

 

[HIMNL9]

I already tried the 34063 ..... it's not a bad regulator, but it require big inductances, due to the 100KHz frequency, and can't be miniaturized too much ..... sure, not for fit in an aixiz module or similar


Some of the new models works at 1MHz or more, and requires much less external components and smaller inductors, and can be miniaturized a lot more

You can try with this circuit, if you want to experiment a bit the same one i'm doing at the moment ..... and you probably have much more time than me, for do that ..... ltc / mc 3429 also don't require sckottky diode

sorry for the bad quality, drawed it on-the-fly ..... also, the values are indicative, i'm still trying to find the better ones.