Lm317 - why???

Hello there...

I am new in this forum this is my first post actually! But I've read a lot of threads here... and I surfed the net for hundreds of hours looking for laser info...

I have even build 2 laser drivers out of scratch ... I even made my own PCBs with the FeCl3 method...

So there is one thing that I did notice...

Why is everybody building their drivers with LM317???
It has voltage drop of 3V ??? WAY TOO INEFFICIENT!!!!

Ok some will say it can be deliver 1.5A current... So what?
I've been in electronics since a kid and now i repair all kinds of Mobile Phones and other electronics for living...

When using lm317 you have to use power supply +3V of the actual laser diode voltage needed to operate.. why waste so much energy when using 2 x lithium batteries at full charge you get

8.4V minus
3V (lm317) minus
~3V (red LD) minus
~0.7 (silicon diode) optimal (there are other diodes with less voltage dropout for rectifying electricity .......
===============
u actually need over 6 volts and the other is dissipated as heat... not two mention the extra space needed for 2 BATTERIES

I have build my own 2 drivers using a low dropout voltage regulator (L1117) which has the same circuit build as LM317 but with voltage dropout of MAX 1.2V at 1A :wave: and i don't use such high dropout diodes for regulating the polarity (it's not even needed that much but i like to add them :na

L1117 (adjustable) can deliver 1A max current which is WAY OVER THE MAXIMUM for a RED LASER DIODE... so why don't you use this one for the red lasers and even for all others except for those who got 1W or over lasers....

AS ALTERNATIVE - !!! LM117 can even actually deliver 1.5A !!!!!! (SO WHY LM317)


ALSO there is a super efficient voltage regulator (a little hard to find where i live but it exists) with a voltage drop of 170mV!!!! at 1A current :drool:
TPS72501 (Adjustable) a little more work on this because it has a pin terminal that acts like a switch to enable input.

when I use these voltage regulators (as current regulators) I use only one 18650 Lithium Battery and i get FULL POWER without worrying how should I get a larger host for my laser for 2 or more batteries!!!

WOW!!! way more efficient than LM317 !!!

So I thought I should share my experience :na::beer:
any comments are appreciated

IIRC, it was originally chosen because it's a standard item in stock at all RadioShack locations.

Way back in the day, a user called Daedal drew up a simple schematic for a linear constant current regulator, and he did it with parts that were all available at RadioShack so that most anybody would be able to find the parts and do it at home.

And of course, being available at RadioShack, it's still extremely easy and cheap for most anyone in the US to pick it up in their hometown, and the diagrams already exist, so people still use it.



Let me guess, you probably typed your message on a QWERTY keyboard, right? Or some similar variation like QWERTZ? Why did you do that? It's an extremely inefficient design that's over a hundred years old and is designed to make you slow! But it's still the standard simply because it was the original, and it's what everybody already used when you started typing.

yeah I guess you are right... for the past

but now parts are readily available so more efficient ones can be used

and LM117 / L1117 both are the same schematic as LM317 even the Vref is the same so easy as hell with the LM317 familiar users

I created similar linear drivers using the LM1587 and MIC29312 ICs. Currently in production at the board house. I can't wait to see my end products!

Jufran88 said:

I created similar linear drivers using the LM1587 and MIC29312 ICs. Currently in production at the board house. I can't wait to see my end products!

Click to expand...

I am glad there is someone who doesn't rely on the LM317... good job there :wave::wave:

Hello and welcome to the forum!:wave:

I have a few LM1117 adj on the way to me, and am currently experimenting
with some LD1085.
But yeah, the LM317 is available everywhere, even locally, ultra cheap, and has proved itself many times over.

So, would this one work?
http://cgi.ebay.com/TPS72501-Low-In...709?pt=LH_DefaultDomain_0&hash=item1c15722635

anselm said:

So, would this one work?
TPS72501 Low Input Voltage Linear Regulators (6) LOT - eBay (item 120618886709 end time Apr-05-11 08:12:30 PDT)

Click to expand...

of course it will... download the datasheet and you can't go wrong
it's a linear regulator with input switch (ENable pin) you should be careful with the EN pin...
the input current is typical - 0.01nA.. and max 100nA "NANO" Ampers

if you go over that rating you will damage it permanently :drool: I did that once... and i can't find another IC locally :thinking: :cryyy:
so i stick to L1117 for now

other than that it's pretty simple to make it work any capacitors on input and output are great... much like any linear regulator schmatic
the FB pin is actually the ADJ pin on LM317, LM117 etc....
Vref is 1.2 V and DOV is 170mV



good luck with those :can:

@ anselm: it's a good regulator for some applications, but it have some limitations ..... first, the input voltage cannot be more than 7V, or it blows, and also, the FB pin is referred to GND, not floating, so the RSENSE must be placed between ground and load, same as in the sink drivers, and cannot be placed like in the LM317 serie ..... also, it cannot regulate more than 5.5V at output.

If someone of you can get LT3080, they are better ..... or also MCP565 .....

There are plenty of posts here about using LM1117's and LD1085. The term LM317 is used loosely to mean linear regulators that are pin compatible.

foulmist said:

8.4V minus
3V (lm317) minus
~3V (red LD) minus
~0.7 (silicon diode)

Click to expand...

Red diodes are closer to 2.5V if I remember correctly, and the silicon diode is not generally placed in series, it is in parallel reverse-bias.
Ideally, no current will flow through it.

foulmist said:

(L1117) which has the same circuit build as LM317 but with voltage dropout of MAX 1.2V at 1A :wave:

Click to expand...

No, that is the MINIMUM dropout of the regulator itself. When operating in constant current mode, there is also the Vref dropout of 1.25V. Anything extra will be dropped by the regulator. If you start out with 8.4V and the laser diode requires only 2.5V, the extra voltage will be dropped by the regulator. In the case of both the lm317 and the lm1117 drivers, they drop the extra 5.9V. The advantage is that the lm1117 can stay in regulation longer, down to about 5V supply as opposed to the 6V supply of the lm317

foulmist said:

ALSO there is a super efficient voltage regulator (a little hard to find where i live but it exists) with a voltage drop of 170mV!!!! at 1A current

Click to expand...

But substantially more when used in a constant current design

Red can approach 3.8V, depending on the current you supply them with - but even LOCs can hit that 3.8V mark.

I agree with OP. It doesn't make any sense to me that people are still using LM317.

LM1117 isn't great either, because you pretty much max out it's current limit, with room left on your 445 for higher current.

My general approach is:

660nm - LM1117
445nm - 1085
405nm - 1085

A few weeks ago I found this: MIC29310 and I believe some forum members are experimenting with it now.

There's also the LD39080 with a supposed 0.2V dropout that I proposed adopting here:
http://laserpointerforums.com/f67/d...ll-linear-red-driver-ic-62025.html#post880403

The reason people use LM317s in my mind? Cheap and abundant. You can buy 10x on eBay for $3 shipped. It's hard to beat that when starting out and you don't want to place a big order, because DigiKey, Mouser, etc charge $8 shipping.

At this stage, I wouldn't buy a regulator from eBay, I'd just wait until my next digikey order and grab another dozen 1085s. But I can relate to why someone would. It's easy and cheap, and if you destroy it, it's only $0.30 down the drain. A lot of people are just getting into this when they find 317 circuits - so the idea of dropping a tiny little bit of money on the components is appealing.

These people haven't yet realized how expensive this damn hobby is

rhd said:

LM1117 isn't great either, because you pretty much max out it's current limit, with room left on your 445 for higher current.

Click to expand...

yep my post was about red laser diode, that's why I said L1117 is a good choice

i haven't tried 1085 the digikey shipping to Bulgaria is $60.... :cryyy:
not worth it but as soon as I get my hands on that I will make some drivers

You can pick whatever adjustable linear regulator you want. The reason LM317s are suggested and used is availability and cheapness. That's really it. I've got a sack of those chips just laying around and they work nicely for whatever hackish job I need.

I also suspect that it was chosen because the constant current regulator was taken from the LM317's datasheet.

At the same time, it's pretty tough to beat Mouser's 100 pcs of 1085 regulators for $33.50

If you are already looking for good efficiency ditch the linear driver and build a sync buck converter to power your diode, you can attain very high efficiencies (90%-95%).

However, if you are looking for something simple, very quick and cheap to setup, then these linear regulators are fairly nice.

Im really not a fan of linear drivers, even the LM317, but sometimes they are irreplacable when you want a very low part count and something that works in under 5 minutes. messing with switch mode supplies/converters, although much more efficient is a bigger hastle then the linear counterpart.

Is there a "DDL circuit" of non-linear driver DIYing?

IE - a fairly simple small buck circuit DIY circuit that can be easily adapted to suit different diodes, etc?