1500mA 445nm diode... use an LM317t or LM338t

If input voltages are the same both chips will dissipate same amount of heat.

Why don't you use a LM1117?

It heats up way less than 317

ReNNo said:

If input voltages are the same both chips will dissipate same amount of heat.

Click to expand...

Rafa said:

Why don't you use a LM1117?

It heats up way less than 317

Click to expand...

Alrighty then, who's wrong? :undecided:

I know that voltage dropped by the regulator is converted to heat, and I also know that when regulating currents past 1A on the LM317, a heatsink is not only recommended, but required.

Renno is right.

It does not matter which regulator you use, they will run equally hot if they are in the same (TO-220) package. Just use a heatsink!

I will smell another trimmer burned, if you use that schematic for 1A .....

Try this one, instead



(it works indifferently with LM317, LM1117 or LM338)

And remember that the current set resistor take the same current of the load, in all these types of schematics, so you need that it can hold the current and dissipate the related power, or it just burn .....



EDIT: about the different chips, is not matter of dissipated power, is matter of internal dropout voltage ..... is not that LM1117 dissipate less than LM317 in the same circuit and with the same values, both of them need to dissipate the same power and amount of heat ..... is when you have a power supply voltage that is near the FV of your diode, that LM1117 gives you an advantage.

Trying to explain .....

LM317 have a typical internal dropout around 2,5V, where instead LM1117 have a typical internal dropout around 1,2V ..... if you are using, as example, a PHR diode, with a FV around 5V, with an LM317 you need, as minimum power supply voltage, the FV (5V), plus the dropout on the current sense resistor (1,25V), plus the dropout of the regulator itself (2,5V), plus at least half volt for safety, so 5+1,25+2,5+0,5=9,25V (can say 9V, but you are at limit), where instead with the LM1117, that drop 1,2V, you need 5+1,25+1,2+0,5=7,95V ..... it does not look too much difference, but with some setups, it can make the difference between a working and a non-working assembly

As matter of power dissipated, instead, if you have two identical circuits, both using a LD with 5V FV, and both powered from the same voltage, say, 12V, both the ICs need to dissipate the same power in heat, cause all the voltage that is not used from the load or dissipated from the sense resistor, is dissipated from the chip as heat ..... so it don't care that one drop 2,5V and the other 1,2V, internally, cause the difference, at the end is the same .....

You are right if you assume you somehow get to choose the supply voltage.

There can be situations where you can run off 2 cells with a 1117, but you would need 3 cells to use a 317. In such cases the 1117 provides a clear advantage, because you can lower input voltage and hence dissipation.

But if that is the goal, why not build a proper opamp driver that allows you to limit the total votage drop to about 1 volt? No need for a regulator to pass any vin-vout requirements that way.

? i do this all the times that i don't have space problems.

One of my schematics of this type of drivers is also in my album (labby type, with modulation option)

But that don't fit too good in an aixiz module

^ I enjoyed that pic in your album, of the 405nm labby. I like the diy clamp heat sink design a lot.

Hope that you get a good design Lazybeam.