Power Supply

[dethlore]

I've spent the past 2 months following threads and getting information, and I felt it's finally time I jump into the forum after exhausting Greg's mind at Stonetek (whom helped me tremendously getting my first laser ever built!!!), so I apologize if I've missed anything about this...

I am wanting to use an AC/DC Converter I've had for a long time to use somewhat as a bench power supply. It outputs 13.5v up to 25Amps. I know if the load is pulling that kind of current, it'll pop instantly, so my question is this:

How might I efficiently regulate the voltage and Amps to safely connect drivers?:thinking:

I've thought about connecting LM317s in series and parallel, but they don't seem to drop the 3v each LM317 like I expected.

Any thoughts how I could get this working? The current project now, after I just recently finished a 20xDVD and 18xDVD laser, is for the 445 Nichia diode.

This is a great forum LOADED with useful information! Thank you for your replies!

 

[lasersbee]

Look into the LM317 Data Sheet... There are Application
notes in the Data Sheet that will show you the proper
way to wire a regulator in Voltage Regulator Mode or in
Current Regulator Mode...

Your power supply seems a bit over powered at 25 Amps..
A smaller 3-5 Amp supply would be sufficient...


Jerry

 

[BShanahan14rulz]

at 13.5V input, you WILL need a heatsink on your LM317, but that should be all that is holding you back... You've looked at the LM317 datasheet and seen the Constant Current diagram?

 

[dethlore]

I have read the LM317 datasheet a few times to make sure I do this right... and if I CAN do this. I understand I need to connect a resistor from adj to out pins to regulate current. Am I mistaken in thinking the input amps are governed on what the circuit will accept (usually more than is safe for it) and I'm not getting a constant supply of 25amps?

I know this PS is overkill... but it's one I've had for a while and what money I DO have I'm trying to scavenge for another Nichia diode.

For heatsinking, I'm thinking of using a Phenomx6 cpu heatsink with fan, mounted to the LM317.

Other issue I'm having so far with it is after being able to limit current enough that it doesn't get too hot, is that the full voltage is coming through the LM317... isn't it supposed to drop it 3v?

Thanks for your fast replies! You guys are awesome!

 

[Toke]

As I recall the LM317 you are supposed to make a voltage divider between out and ground, then connect adjust in the middle, and you will get a voltage out determined by the relation of the two resistor values.

13.5V is pretty high starting voltage, you will need something like 4.2 or 8.4v.

 

[lasersbee]

If you are using the LM317 in a current regulator mode...
you will need to put a proper load (ie LD Test Load) on the output
Then and only then will the voltage be there and you can not
adjust it... you can only adjust the current....

Remember that any voltage over the LD voltage and regulator
dropout Voltage will be turned into heat....

As I recall the LM317 you are supposed to make a voltage divider between out and ground, then connect adjust in the middle, and you will get a voltage out determined by the relation of the two resistor values.

@ Toke.... that is if the LM317 is used as a Voltage Regulator...


Jerry

 

[Toke]

@ Toke.... that is if the LM317 is used as a Voltage Regulator...

Well, yes.
If he wants to connect his supply to a driver it should be voltage regulated to emulate a battery, not current regulated and trying to drive some current through the driver at 13.5V. (10.5V)

 

[dethlore]

With that being said, would I be able to use 2 LM317s to regulate first current, then voltage, or vice versa?
I'm sure I could sink both of them on the heatsink I'm thinking of using to deal with the heat.

 

[lasersbee]

Well, yes.
If he wants to connect his supply to a driver it should be voltage regulated to emulate a battery, not current regulated and trying to drive some current through the driver at 13.5V. (10.5V)

I agree 100%...

He should use an LM317 in Voltage Regulator Mode
driving another LM317 in Current Regulator mode...IMO


Jerry

 

[Cyparagon]

So that means the function of the first is just to drop voltage? The second can drop all the voltage just fine. Even if it can't (if you can't find a big enough heatsink for example), there are simpler ways of dropping voltage. A resistor comes to mind.

 

[lasersbee]

So that means the function of the first is just to drop voltage? The second can drop all the voltage just fine. Even if it can't (if you can't find a big enough heatsink for example), there are simpler ways of dropping voltage. A resistor comes to mind.

Not only that... it will dissipate some of the heat generated by dropping
the 13.5Volt from his PS to maybe 9-10VDC and help filter it and keep it stable. We don't know how stable or filtered his 25 Amp really supply is...
Then the 2nd current regulator could handle any current demands without
also having to drop 8-10volts (depending on LD used)...

But that's just the way I would do it...
There are may ways to skin a cat....

Jerry

 

[dethlore]

The power supply is high end and very stable. I've been able to regulate the voltage down to 9v successfully, and am currently tweaking the current regulating Lm317... I'll let you know how it goes.

For the LD, I'll be using this for every diode I get, with some tweaking for correct voltage and current of course, but at the moment, I'm using a test load I built for the 445 Nichia.

 

[Bionic-Badger]

Why don't you just use some rectifier diodes to drop voltage so that the LM317 doesn't need to eat excess voltage? Some 1N5400 diodes will drop 1V and can handle 3A. Provided you buy a bunch ($11 for 100, $1.50 for 10 at Mouser), you can use the same diodes for your dummy load.

 

[dethlore]

I tried the same thing with N4001 diodes but I let the smoke out of one of em. Guess they're drawing too many amps, but I guess that IS a possibility with the voltage divider... Seems to be many ways to attack this one. I'm going to try them all to see which is most efficient.

EDIT:
I have the correct voltage now using a couple resistors as a voltage divider before the LM317. Not getting hot at all, but that's because the circuit is still only drawing 100mA max. The test load I have SHOULD be drawing more amps than that... I thought in using resistors as voltage dividers the only thing that mattered was the ratio and power rating?

Is it possible that the Ohms on the resistors in the voltage divider are too high and not allowing more amperage to flow through?
Using a 1kOhm in series with 2 parallel 1kOhm 1/4watt resistors for the voltage divider (I know-I know these aren't the best values, but it's what I had nearby and available last night.)

 

[Cyparagon]

Voltage dividers are horribly inefficient under heavy loads. If you don't have a big enough heat sink, use a resistor to lower the voltage. A 3Ω will drop 3V at 1A. Just be sure it can handle 3W.

 

[dethlore]

Voltage dividers are horribly inefficient under heavy loads. If you don't have a big enough heat sink, use a resistor to lower the voltage. A 3Ω will drop 3V at 1A. Just be sure it can handle 3W.

Aye... perhaps I will go back to the original idea of using an LM for volt reg and a second one for current. I think I will be needing 3 ohm resists to allow enough current in... and a big heatsink