Welcome to Laser Pointer Forums - discuss green laser pointers, blue laser pointers, and all types of lasers

How to Register on LPF | LPF Donations

[DIY] Driver for 635nm diodes, 200mA~ Cheap! And my PCB making tutorial, Easy!

gillza & comradmax,
I've been confused on the LM317 issue too till now. Not from Rafa's post, it actually clarified it for me. The regulator is being used for a current regulator not a voltage regulator. To run the LD you need enough voltage for the LD and be above the dropout regulator voltage. The 1.25v is a reference voltage needed and its not the same as the regulators operational voltage. To make it work you have to have a voltage at or above the LD & regulators needs. Using the LM1117 with a drop out of 1.1v + the LD's needs of 2.2v totals 3.3v to work. So a single Li-Ion 3.6v(4.2 off the charger) works. If you used a LM317 with a dropout voltage of 2.2v, add that to the 2.2v for the LD and now you need a min of 4.4v so a single cell won't work. You need 2 cells and that with Li-Ions is going to be 7.2v(8.4v off the charger). That would work with the LM317, but the trade off is all the excess voltage is converted into heat that has to be dealt with. 2 3v CR2's would be a better choice. I was always adding the 1.25v reference voltage into the total needed. If this is wrong, someone please correct me.
 





One thing I would like to interject on the PCB tutorial. The iron on PCB thing is fine for one offs heh. But if your going to be doing a lot of boards, invest in a UV lamp, a roll of negative dry film resist, a laminator, and laser printer transparency sheets.

It's a similar process. You laminate the pcb with the resist film. Print your PCB layout on a transparency, Sandwich the transparency against the PCB then expose it to UV for the required length of time, then etch. These leads to much cleaner etches than the toner/iron on method, can reproduce smaller traces, and most of all, your templates are reusable. :D

If your only doing a single project, it's probably not worth the investment, but for people like me who end up making a board or two a month it seems, it's worth it in the long run.

Another thing worth mentioning is there's a much cheaper alternative to ferric chloride as well.. Cupric chloride. Plus you can make it yourself and it's 'rechargable' so you can use it over and over again indefinitely.

Only downside is that the first few etches should be done outdoors until it gets enough copper dissolved in it that the HCL concentration goes down to the point it doesn't fume chlorine gas :p After a few etches it'll turn a really nice green heh, and the acid content will be low enough that it's safe to use indoors, since the cupric chloride will be doing most of the reacting, instead of the acid.

Here's how to make it.

Stop using Ferric Chloride etchant! (A better etching solution.)

and as the warning says.. don't get this stuff near any metal you want to keep. especially stainless steel. They mean the warning.. trust me.. heh
 
Rafa said:
Yeah you should not add Vref to the total needed.
Click to expand...

When I had the workings of LM317 hammered into my non-electrical-engineering head the solution for Vin was:


Vout + Vlm317_drop = Vin
Where:
Vout = Vresistor_drop(Vref)+Vdiode_forward.

lm317bluray.jpg

I'm not using the numbers from the diagram, just using it to visualize what's happening.

The line of reasoning in layman terms is this: lets say we have a diode with Vf = 4V. LM317 has an internal voltage reference of 1.25V, the chip "senses" the potential at Vadj and drops the Voltage across the resistor connected to Vout by 1.25V (adjusts potentials between Vout and Vadj so the difference is 1.25V less?) according to its internal reference. But the potential at Vout has to be a sum of Vdrop_across_diode (4V) and Vdrop_across_resistor(Vref) (1.25V) otherwise if Vdrop_across_resistor(Vref) is not included the potential difference across the laser diode will be 1.25V short.
So to power the diode with forward voltage of 4V using simple lm317 based driver we need Vin= : 1.25 + 4V + 3V(Vlm317_dropout) = 8.25V

Am I totally wrong and talking out of my A$$, if so could you explain why I should not add Vref into this equation?

Thank you in advance!

P.S. I am not trying to argue, I'm really trying to understand it.
 
Last edited:
@Quemefox: I make, like you, 3 or 4 boards a month. I like insolation method but I think I dont need it, for now. Toner works for me, if I need to make a board with smaller traces or something I will buy all the things needed to use that method. I already tried to make them that way but it's way more complicated, and you have to do lots of extra things. I already bought the revelator for the boards, but hey, as I said, toner works for me.
I have tried what you said, the Cupric Chloride. I thinks it's dangerous, because of the fumes. And the ferric chloride for me it's way cheaper than $3/liter (note that the plastic container I used has been reused lots of times, so whats inside it's not what it originally had. I refilled it with cheaper Ferric Chloride bought at a chemics store.) Do you make boards only for laser-related things?

@Gillza: You don't have to add it to the equation, because as you said, it's just a voltage reference for the resistor to set the current that the LM317/1117 gives.
 
qumefox said:
One thing I would like to interject on the PCB tutorial. The iron on PCB thing is fine for one offs heh. But if your going to be doing a lot of boards, invest in a UV lamp, a roll of negative dry film resist, a laminator, and laser printer transparency sheets.

It's a similar process. You laminate the pcb with the resist film. Print your PCB layout on a transparency, Sandwich the transparency against the PCB then expose it to UV for the required length of time, then etch. These leads to much cleaner etches than the toner/iron on method, can reproduce smaller traces, and most of all, your templates are reusable. :D

If your only doing a single project, it's probably not worth the investment, but for people like me who end up making a board or two a month it seems, it's worth it in the long run.

Another thing worth mentioning is there's a much cheaper alternative to ferric chloride as well.. Cupric chloride. Plus you can make it yourself and it's 'rechargable' so you can use it over and over again indefinitely.

Only downside is that the first few etches should be done outdoors until it gets enough copper dissolved in it that the HCL concentration goes down to the point it doesn't fume chlorine gas :p After a few etches it'll turn a really nice green heh, and the acid content will be low enough that it's safe to use indoors, since the cupric chloride will be doing most of the reacting, instead of the acid.

Here's how to make it.

Stop using Ferric Chloride etchant! (A better etching solution.)

and as the warning says.. don't get this stuff near any metal you want to keep. especially stainless steel. They mean the warning.. trust me.. heh
Click to expand...


I buy positive coated copper board, which comes with the photoresist already laminated and it's quite cheap. Instead of using toner transparencies I use inkjet transparencies, I print them at photo quality and right after the printing is finished I coat them with talcum powder, remove the excess and dry them with a hair dryer. The quality I get is almost professional.
As etchant, I started using HCl + H2O2 but then I discovered ferric chloride. It's for me a better etchant as it's slower and doesn't emit too much fumes. Slower means better tolerance and control and that's why I like it.

My total investment for PCB making:
39€ for 3x 8W UV tubes (I only use one of them for most of the boards I make)
3€/year for ferric chloride
0.50€/year for the sodium hidroxide required for developing the board
2€ for a couple of 30x20x2mm glass panels with rounded edges
1€ for 4x of these: http://www.exchange3d.com/images/uploads/aff586/aa.jpg
17€ for a pack of 60 inkjet transparent sheets
15€ for a tube of soldermasking UV curable paint
0.60€/month for latex gloves
1€ for plastic trays
0.90-2€ per drill bit

I do up to 5 boards/month during the year, and I can get almost commercial quality etchings, with 8mil or even a bit less perfectly clear traces or 0,20mm BGA pads, with spacings of 8mil between traces. I do also soldermask them.

The funny part of the story is that I started making PCB's in a way more complex and expensive way (using an insolator an that) and now, using household, extremely cheap tools, I get PCB's comparable to commercialy made ones. All what's left for me is to get a reliable way to through-plate holes and vias.
 
erdabyz said:
As etchant, I started using HCl + H2O2 but then I discovered ferric chloride. It's for me a better etchant as it's slower and doesn't emit too much fumes. Slower means better tolerance and control and that's why I like it.
Click to expand...

Heh. Your missing all the fun then. The initial HCL+H2O2 etches REALLY fast and yes and it fumes.. but that's just the initial mix. The trick is to keep using it until it turnes bright green(if your in a hurry, you can just get some copper wire or something to dissolve in it). When it turns bright green, it becomes cupric chloride, which is what your really after. In this state, the cupric chloride is doing the etching, most of the HCL has been converted so it doesn't fume. When it has so much copper in it that it turns brown, you just bubble air through it and it turns bright green again, and it's good to go once more.. no having to dispose of it like used up ferric chloride.

Once it reaches cupric chloride, it etches a little slower than ferric chloride.
graph_time_vs_temperature.png


Etching with Air Regenerated Acid Cupric Chloride

But, if your only doing one board every now and then, this probably isn't worth the trouble. Just use ferric chloride. I was mainly posting this info for people who do or have interest in doing lots of boards. I think most commercial PCB makers use cupric chloride because of the ability to regenerate it.
 
Last edited:
Wouldn't it be just as easy to drop a penny or a length of copper 12/2 house wire in the HCL and sit it aside till it's all converted where the fumes weren't an issue? Of course some will say this is a waste of the initial fast etch, but you end up with cheap copper chloride solution and no fumes to deal with.
I made it this way for copper plating solution. ;)
 
It is, but you need some type of oxidizer to start the initial reaction or else it takes forever. HCL doesn't really react to copper very well by itself. That's why you have to add some h2o2.
 
I was never in that big of a hurry. lol
I just drop it in and leave it in the garage. When I need it, it's there.
It's just a pretty simple and economical way of making it.
You can use a penny as the anode & cathode, add a little dc voltage, and it'll go green pretty quick.
 
Rafa said:
@Gillza: You don't have to add it to the equation, because as you said, it's just a voltage reference for the resistor to set the current that the LM317/1117 gives.
Click to expand...

Rafa

Thank you for your reply. I am still unsure about that. Well I don't mean to trash your thread with irrelevant information so, I'll try to explain my point a little differently one last time and in the mean time will try to search this forum for more information.

So there is a Vref = 1.25V that is produced withing Lm317 (probably as part of the Lm317 dropout voltage) and used to compare to drop across the resistor in the circuit. That voltage is not taken into the account when calculating necessary Vin. (Vin = Vout + Vlm317_droput)
However, when within the driver circuit (that LM317 is part of) the voltage across the resistor is dropped by 1.25V to set the current, LM317 measures this drop and compares to its internal ref. produced within Lm317 (mentioned above).
So Voltage drop across the resistor in the circuit is set by comparing the drop across the resistor to internal reference (Vref), essentially the values Vresistor_drop = Vref = 1.25V but they are two different quantities.
So when I was explained this by a person I bugged before you :) he said that Vout from LM317 = Vdrop_across_resistor (that is equal in value to Vref) + Vdrop_across_laserdiode.

My apologies that I keep coming back to this, I just thought I'll try to explain myself a little differently and maybe better.

Thank you!

P.S. Thank you again for your tutorial. It is by far the easiest I've seen around.
 
Last edited:
Vaporizer said:
I was never in that big of a hurry. lol
I just drop it in and leave it in the garage. When I need it, it's there.
It's just a pretty simple and economical way of making it.
You can use a penny as the anode & cathode, add a little dc voltage, and it'll go green pretty quick.
Click to expand...

Ahh. Adding energy to the mix would probably have the same effect as an oxidizer. heh
 
Hey guys I finally went to the store, got some components and found that actually this doesn't work....
Gillza may be right, I don't really know why it doesn't work but if I connect it to a lithion, 3.8V, its not working, but if i connect it to a 5V PSU works just right.

At leas the etching tutorial works right.... :crackup:
 
hi
pulled this from fairchild semiconductor LM317 datasheet (others looked at are more or less same)

vdroplm317fairchild.jpg


it looks to me the more you draw the more the voltage drop is, just guessing more circuitry is active to stabilise output and temperature.
 
Last edited:
First, It's not an LM317 waht I used. Hoy many times do I have to say it?
It was an LM1117. And that graph is not what you say. The more current you draw the more voltage it drops. The more hot the IC is, the less Drop-Out it has, up to 500mA.
 
You must log in or register to reply here.


Back
Top