Need help with designing driver for 445nm diode

[Jmillerdoc]

One other thing. You need to isolate the diodes one from another. Meaning you do not use the case pin in any circuimstances, you need to obtain the battery negative connection through some other way, which may prove difficult.

Maybe I am misunderstanding something here, I took this to mean the casing of the diode was tied to the negative pin.....hopefully I misunderstood, if the case of the diode is not tied to the negative pin than I see no problems.

I certainly did not intend to school the teacher....you can see where I was coming from hopefully...believing the diode case was one with the negative lead would certainly cause problems with electrical isolation in a common heatsink.

So, if I misunderstood...now I really don't understand the above quote...please elaborate.

Thanks again for all the explanations,
Jeff

 

[Jmillerdoc]

Maybe a definition of "case pin" is in order here. Please forgive my ignorance.

 

[Jmillerdoc]

I think I just answered my own question here:
http://laserpointerforums.com/f65/pinout-445-nm-diode-a130-52402.html

The case pin is jist a pin connected to the case...it is neutral in respect to the other two pins and can be disregarded. Seems it is just a solder point that allows a connection to the body of the sink, or host, for that matter. Just a convenient solder point for use of a body ground layout.

Ok, back to the series build...I am glad that's cleared up. In terms of battery connections...that can be worked out in many different ways as far as I can tell.

Does the Maglight switch have on of it's legs tied to the body of the flashlight? That would restrict it's use to a body ground design unless it could be disconnected from the housing and made to float again. Not sure of how the switch works in the Maglight but if it is tied to ground on one leg I could now see how that quote above makes sense.

 

[Eudaimonium]

I feel like this is the twilight zone. What am I missing here? Were it not for all the Thanksgiving drinking, I'd put my life on the fact that there's no isolation issue.

445s are case isolated, why would there be any issue with diode isolation?

Am I having a brain fart? Eud is more of an authority than almost anyone here, so I must be missing something.

My bad I should have been more clear,

Yes, there is no issue in isolation, as long as you take care not to connect any case pin to anything, except one. I'd reccomend cutting off the other two completely because, if second one gets connected, everything goes.

And I'm not much of an authority, I just try to provide notes and tips

 

[rhd]

Maybe a definition of "case pin" is in order here. Please forgive my ignorance.

Quick note - You should EDIT your posts, and add new information, instead of posting multiple times in a row.

My bad I should have been more clear,

Yes, there is no issue in isolation, as long as you take care not to connect any case pin to anything, except one. I'd reccomend cutting off the other two completely because, if second one gets connected, everything goes.

And I'm not much of an authority, I just try to provide notes and tips

I still think that the case pins are probably irrelevant, even if all three are used. Suppose he ties all three into the driver's (-) input, what would be the downside (apart from the added complexity of a few extra wires)?

Considering the fact that this will be a nearly 30W driver, it might not be a horrible idea to feed that power from more than one case pin.

 

[Eudaimonium]

I still think that the case pins are probably irrelevant, even if all three are used. Suppose he ties all three into the driver's (-) input, what would be the downside (apart from the added complexity of a few extra wires)?

Well, the fact that they would not be connected in series anymore. And EVERYTHING changes.

Anyway, this will be one tough build, especially for somebody who had the difference in series and parralel JUST explained to him. If it weren't for the fact he already ordered the custom heatsink and all (as far as I've figured, right?), I'd say - baby steps. Try an easy DIY setup first.

 

[rhd]

Another thought -

You could do this build with just two cells, and still stick to the linear driver concept. If you used two of these:
http://www.lighthound.com/AW-IMR-26500-2300mAh-LiMN-rechargeable-lithium-battery_p_3046.html
- and then built 3 linear drivers each supplied in parallel from the two cells, but outputting to their own individual diodes, you'd be fine.

Anyway, this will be one tough build, especially for somebody who had the difference in series and parralel JUST explained to him. If it weren't for the fact he already ordered the custom heatsink and all (as far as I've figured, right?), I'd say - baby steps. Try an easy DIY setup first.

I agree 100%.

At the very least, I'd cut the current from 1.8A down to 1.5A or maybe 1.6A if I was building this. Jmillerdoc, you should probably be prepared for the possibility of loosing one of your diodes if you run all three at 1.8A. I would order four instead of three.

In fact, if this is one of your first builds, you might want to be emotionally prepared for loosing all three. Everyone looses a few diodes in the beginning just from build errors, undischarged caps, etc. If you've got three diodes tied to the same system, then that first potential "oops" moment, could be triple-fatal.

The other thing that occurs to me is that it's going to be fairly near impossible to get any sort of parallel beam output from the three diodes. If you don't have the ability to micro-adjust each Aixiz module's seating, then I don't think you can expect an equilateral spacing between the beams past a few feet out.

 

[Jmillerdoc]

Well, the fact that they would not be connected in series anymore. And EVERYTHING changes.

Anyway, this will be one tough build, especially for somebody who had the difference in series and parralel JUST explained to him. If it weren't for the fact he already ordered the custom heatsink and all (as far as I've figured, right?), I'd say - baby steps. Try an easy DIY setup first.

You assume too much about my lack of knowledge...I have known the difference between series and parallel connections since highschool. You should check out some of my amplifier builds From scratch at diyaudio.com. I am sure you would change your mind about my ignorance.

I agree with RHD, if the case pins are isolated from the +/- diode pins, all three, two, or one can be connected...this will not alter the series connection, as long as the case pin remains isolated from the +/- diode pins. My confusion earlier stems from believing the case pin and - pins were BOTH - pins.

Also, with 25-30W of current flowing through the circuit I agree, a large ground layout would be preferrable. In audio we tend to avoid multiple ground points to eliminate ground loop feedback noise...I doubt that is an issue with
laser diodes.

The function and application of these laser diodes is not that difficult to understand...I am more concerned with my lack of knowledge about their dynamics and common unforeseen pitfalls. How hardy are they? Like some Audio ICs, are there critical tolerances I need to be aware of? For instance, is it okay to run a few mA over the rated max or will they blow up immediately or just wear out at a faster rate...does the voltage across the diode need to be critically controlled or is a little leeway ok?

And no, i have not ordered any parts yet...still trying to decide the ups and downs of each deisgn choice I have available.

Using three different drivers will allow the posiblity of three different diode types down the road should I decide to do something like that.

I am leaning towards the series layout due to it's simplicity. I think I will go with the 1084 over the 1085 due to it's hight voltage rating...some of the 1085's at Digikey are rated to 12v and others slightly higher.

I have an RC battery from my days of RC airplanes...it is long and skinny and probably would fit into a Maglight barrel. I believe it is rated at 8200mA/h and is 14.7v. It was designed to run the electric Motor of the plane and from what I can remember I could get about ten minutes of flight time running wide open. Any reason why this couldn't be used in place of the customary cells you guys use around here?

Thanks again for the advice,
Jeff

 

[rhd]

I think Eud is right - you really should do some more reading and experimentation before jumping into this particular build. Regardless of experience in other fields of electronics, you're missing some of the keys to this one. I don't think Eud was insulting you. He was just picking up on your level of experience with laser circuitry, and making a helpful (and wise) suggestion.

He's right - you should do some additional reading before jumping in here. You're missing some of the nuts and bolt basics.

Like some Audio ICs, are there critical tolerances I need to be aware of? For instance, is it okay to run a few mA over the rated max or will they blow up immediately or just wear out at a faster rate...

Yes and no. Yes, it's ok to run a few mA over the rated max... since the rated max for these diodes is like an Amp. No, there's not really a singular specific critical tolerance. Some diodes die at 1.5A, some at 1A, some at 2A. Yes, some will blow up (well, die out) immediately. Yes, some will also just wear out faster.

does the voltage across the diode need to be critically controlled or is a little leeway ok?

This, in my mind, is one of the questions that really suggests that there would be value to tackling a basic build first. We're not generally creating voltage regulating drivers at all. Of the ICs that we've discussed using, and the build approaches we've been talking about in this thread - none are voltage regulating drivers.

Using three different drivers will allow the posiblity of three different diode types down the road should I decide to do something like that.

No. Once you start talking about different diode types, you start talking about different pinouts. Swap in a LOC and now you've got the isolation issues Eud was alluding to. Swap in a green module (so, IR diode) and you've got polarity issues. Even with three separate drivers, you'll still have problems with case (-) vs case (+) if you use different diode combinations.

And careful with your IC choices. I've also seen 1084s (for example the cheap ones on Mouser) that have a max VIN of 7V.

I think everyone here is just trying to help. Build a single-445 handheld, with a DIY driver, first. You'll thank us later

 

[Eudaimonium]

Look man, don't take this the wrong way.

It's exceptionally difficult to word something other than "exceptional compliment" in Internet without making it sound offensive. You wouldn't get that defensive in real life conversation, no need for it here.

All I was saying is that, I would not attempt a laser with 3 heads in a Maglite. Definitely not in Maglite. There's that Fenix light or whatever, and if you're making three laser heads, might as well go for 3 colors since all 3 same color, are essentially like 1 laser, only more clumbsy. When you get it working, you'll see what I'm talking about.

It's just my opinion though, I'm sure you'll enjoy it when it's finished.

Sorry for offending you.

 

[Jmillerdoc]

Thanks for the input...I am not insulted, just don't want you to think I am a electronics virgin.

I understand fully the power supplies used here are not voltage regulated...the voltage regulators used are used as current regulators...have done this many times in creating CCS drivers for input tune plate loads, etc....I guess I should have been more precise...when asking about the voltage tolerances across the diodes I am referring to the minimum and max voltage they can tolerate and still function within an acceptable range. Clearly the diode acts as a type of transfer function turning voltage and current into optical power...something like P=(IV)t...
I will have to admit I have not yet referred to the manufactures data sheet on these diodes...mostly because I have no idea what brand and model numbers these 445nm diodes are. I am being lazy in that regard. I was hopping to get a quick and easy answer here. I guess what I am asking is there a generally accepted range of operating voltages I should adhere to.

Since batteries are being used I can see where voltage regulation would have little benefit unless you wanted the supply to be way over voltage requirements and the regulator was used to drop the voltage some...that would seem to be a waste though really....the other thing is I see no benefit in ripple rejection either unless you were using a high ripple DC power supply....not the case with batteries.

Back to my presupposition the diode acts as a optical power transfer function...it would seem to me there would be the possibility of using a lower voltage, higher current supply vs. A higher voltage, lower current supply. My guess is there is both a minimum and maximum operating range for both current and voltage.

So back to my original question...for three diodes in series, what would be the maximum supply voltage one would not want to exceed. In Eud's example (or was it RHD's, can't remember at the moment) it was recommended to use four 3.7v batteries (4.2v fully charged?) for a net supply voltage of 14.8v (16.8v charged). It was also mentioned the diodes are 4.5v devices so in series 13.5v...with the proposed supply we have 1.3-2.3 excess volts....so I must assume this is OK and if I assume this is OK then my deduction is that there is likely some maximum supply voltage somewhere that should not be
exceeded. Again, I own RHD's build now that has a single 445 diode running off two 3.7v batteries....7.4v (8.4v charged) for a 4,5v device, almost double.

OK, I will concede however, this proposed build may not be a good starting point...but for the reason RHD mentions about parallel beams...I assumed the tolerances would be good enough to keep the beams parallel to infinity...that was terribly naive....and it sucks too that it's not true! Bummer. Yeah, now that I think about it I can see how the tolerances would have to be extremely tight to obtain parallel beams to any extreme distance. This will be something to tackle on a later build. I am curious to know how some of these guys build dual diode handheld lasers that produce a single beam from 2.5w up to 3.5w on one I saw.....they didn't explain how it was done on the threads I found.

Thanks again for the help...please refer me to some good sticky threads you believe may help me...I am perfectly willing to read as much as I can get my hands on...problem is knowing what I need to be reading now has been by stumbling through the threads...not very efficient.

Jeff

 

[rhd]

Since batteries are being used I can see where voltage regulation would have little benefit unless you wanted the supply to be way over voltage requirements and the regulator was used to drop the voltage some...that would seem to be a waste though really....

Well actually - that's exactly what we're doing with linear regulators, even in constant-current mode. Using two lithium ions, you're supplying ~8V into the IC, dropping all but Vf + 1.25V at the IC. Effectively, you are using the IC (as well as the resistor) to drop 3 or so Volts for a 445 build.

Back to my presupposition the diode acts as a optical power transfer function...it would seem to me there would be the possibility of using a lower voltage, higher current supply vs. A higher voltage, lower current supply. My guess is there is both a minimum and maximum operating range for both current and voltage.

I think you're over-complicating things. LDs generally have an IV curve that is somewhat consistent from diode to diode (more so than the Power to Current ratio would be). Voltage tends to be a function of Current.

So back to my original question...for three diodes in series, what would be the maximum supply voltage one would not want to exceed.

Again, you're framing your question in Voltage-regulation language. If you're regulating current with a linear driver, then exceeding a particular Voltage isn't a concern (beyond the basic need to know if you've got enough to cover your driver's dropout and the total Vf of all three diodes).

Again, I own RHD's build now that has a single 445 diode running off two 3.7v batteries....7.4v (8.4v charged) for a 4,5v device, almost double.

Important distinction - in that build it's the driver, not the diode, running off ~8V charged.

It was also mentioned the diodes are 4.5v devices so in series 13.5v...with the proposed supply we have 1.3-2.3 excess volts....so I must assume this is OK and if I assume this is OK then my deduction is that there is likely some maximum supply voltage somewhere that should not be
exceeded.

Depending on the specs of the IC, you could theoretically use 10 cells, and 37V of input power. Your IC would just get really hot, really fast.

OK, I will concede however, this proposed build may not be a good starting point...but for the reason RHD mentions about parallel beams...I assumed the tolerances would be good enough to keep the beams parallel to infinity (...) I am curious to know how some of these guys build dual diode handheld lasers that produce a single beam from 2.5w up to 3.5w on one I saw.....they didn't explain how it was done on the threads I found.

That's one point that I don't actually have first-hand knowledge to back up. Once you account for tolerances in the pressing of the diode, the threading of the Aixiz module, the fit of the module in the heatsink, the set screw, and then the internal alignment of the lenses themselves, it just doesn't seem realistic to the beams could be terribly parallel.

That said, I could be wrong. If you're getting Jayrob to make this heatsink, then I'd say you have the best shot at good results. If he could make it press-fit, even better. That said, we constantly see expensive single-laser handhelds with off-center beams.

If you happen to have a 445 laying around that is in a perfectly cylindrical host, turn it on, and then roll the laser. Does the dot stay at the same elevation on the wall 5 feet away? Some of mine aren't bad, some of them move up and down by two or three inches.

 

[Jmillerdoc]

Ok, now I am starting to see my folly....I've been thinking of the LD as a type of black box that can only have a certain voltage across it...what you're saying os it acts like (I know this sounds stupid) a diode! I use regular red diodes in some of my tube audio circuits on the cathode of the tube to get a certain drop I am looking for, sometimes two, sometimes four or five....the top of the diode chain may be 40-50v relative to ground so let's say five 2v red diodes will give me a 10v drop at a constant current....I am starting to see the similarities here.....what you're saying is the LD has a drop across it of 4.5v, just like a regular red diode may have a 1.8-2v drop across it. The DDL keeps the LD from sucking up too much current...it's needed to keep the diode from drawing over 1800mA (or whatever you want it to). Looks like the big difference here is an LD is current Hungry compared to a regular old red diode.

Am I on track here? I hope so...

So RHD, do you recall the resistor value and the current you set my 1085 with?

I also didn't realize these 445 diodes may easily burn up with 1800mA....I was getting the impression from what threads I have read that 1800mA was pretty reasonable and easy current for them to handle...now you've got me a bit frightened that I may waste $40 pretty quickly trying to run 1800mA on a potentially poor performing LD. Maybe I will start with 1500mA and kick it up over time. What is the highest mA you've heard these have handled?

Ever used an LM350 before for a DLL?

Thanks!