Can an LT1085 be used as a constant current LED driver? There is no mention of it in anywhere in the datasheet, but since this is an LM317 replacement can I use it the same way?

 

Like the LM317, the LT1085 is a linear voltage regulator with the advantage of a low dropout voltage, so less voltage headroom is necessary. If you want to control the current in an LED just use a resistor in series with it. In your case, with a 12 volt supply a series resistor of 1K ohms should give approximately 10mA or less.

 

I see no reason why you should not be able to use the LT1085 as a current source.

 

R1 is shorted out, and the potential divider Rsense/R2 isn’t going to do much. What were you hoping to adjust?
I hope you’ve got a big heatsink. 9V loss x 2.95A =26.5watts.

 

Here's the simulation with the unneeded parts removed (you can't regulate current and voltage at the same time):
As stated by Ian0, note the high power dissipation of the regulator (red trace) from the ≈3A through the LED (yellow trace).

 

Here's the simulation with the unneeded parts removed (you can't regulate current and voltage at the same time):
As stated by Ian0, note the high power dissipation of the regulator (red trace) from the ≈3A through the LED (yellow trace).

View attachment 290323

Thank you. I'd need two regulators for constant current and constant voltage, in series? For the constant voltage part, would my original schematic work (minus Rsense)?

 

And I do have a small heatsink for the TO-220 regulator.

 

R1 is shorted out, and the potential divider Rsense/R2 isn’t going to do much. What were you hoping to adjust?
I hope you’ve got a big heatsink. 9V loss x 2.95A =26.5watts.

Trying to build a constant current/constant voltage adjustable power supply, but I may need to use two identical regulator for that.

 

And I do have a small heatsink for the TO-220 regulator.

That is unlikely to handle 29W.
For that power dissipation a heat-sink with a fan may be needed.

Trying to build a constant current/constant voltage adjustable power supply, but I may need to use two identical regulator for that.

Yes, you can put two in series for that (constant current first, feeding the constant voltage).
Note that the constant-current is only readily adjustable over only a small range with a potentiometer for the shunt resistor due to current limitations of available pots.
For a larger adjustment range, there are better circuits.

What will be your DC raw power source for the supply?

 

Either batteries or 12V from an ATX power supply.

 

That is unlikely to handle 29W.
For that power dissipation a heat-sink with a fan may be needed.
Yes, you can put two in series for that (constant current first, feeding the constant voltage).
Note that the constant-current is only readily adjustable over only a small range with a potentiometer for the shunt resistor due to current limitations of available pots.
For a larger adjustment range, there are better circuits.

What will be your DC raw power source for the supply?

Does this look right?

 

Does this look right?

Not completely.
R1 is still shorted.

Your capacitors are larger than necessary (see excerpt from data sheet below).
And the output capacitor should be a tantalum type electrolytic.

 

Dissipating over ~20-Watts of heat is just not very practical,
and will probably end-up shortening the life-expectancy of the Regulator(s).

Adding 1 or 2 additional LEDs in series will alleviate a lot of that Heat.

The best way to go with high-power LEDs is a Switching-Current-Regulator.
.
.
.

.

 

The last screenshot I posted was missing this point

But how is R1 shorted?

 

But how is R1 shorted?

How is R1 NOT shorted?

 

Yes, you can put two in series for that (constant current first, feeding the constant voltage).

Does that make much sense?

What would you expect to happen if you drive a constant voltage source with a constant current source?

Within it's compliance limits, the constant current source is going to try to FORCE the constant voltage source to consume the programmed current. But what if the load that the constant voltage source is driving either doesn't need that much current or needs more than that current?

If the constant current source succeeds, it will have done that by forcing the constant voltage source to output a different voltage -- one that results in the load accepting the programmed current. But that means that the constant voltage source has failed to hold it's programmed voltage. If the constant voltage source succeeds, it will have done so by either limiting the amount of current from the constant current source, or drawing more from it. In either case, the constant current source has failed to hold it's programmed current.

It makes no sense to try to control both the current and the voltage -- you control one and the circuit dictates the other.

 

Isn't R1 supposed to go between OUT and ADJ?
View attachment 290411View attachment 290412

Yes. But notice that neither of those diagrams have a wire that goes from one side of R1 to the other side. Yours does.

 

View attachment 290410

How is R1 NOT shorted?

oops, fixed it.

 

Does that make much sense?

What would you expect to happen if you drive a constant voltage source with a constant current source?

Within it's compliance limits, the constant current source is going to try to FORCE the constant voltage source to consume the programmed current. But what if the load that the constant voltage source is driving either doesn't need that much current or needs more than that current?

If the constant current source succeeds, it will have done that by forcing the constant voltage source to output a different voltage -- one that results in the load accepting the programmed current. But that means that the constant voltage source has failed to hold it's programmed voltage. If the constant voltage source succeeds, it will have done so by either limiting the amount of current from the constant current source, or drawing more from it. In either case, the constant current source has failed to hold it's programmed current.

It makes no sense to try to control both the current and the voltage -- you control one and the circuit dictates the other.

my goal is to build a multi-purpose adjustable power supply with constant current & voltage limit that I can use to power LEDs and laser diodes, charge batteries and power external peripherals.

 

my goal is to build a multi-purpose adjustable power supply with constant current & voltage limit that I can use to power LEDs and laser diodes, charge batteries and power external peripherals.

THIS is where the discussion needed to start.

So let's flesh this out into a reasonable set of basic specifications. THEN we can start discussing ways to design a circuit that meets them.

Let's assume that you already have black box that is an adjustable constant current power supply with a voltage limiter.

Is that REALLY going to do what you want?

It might be fine for powering some LEDs, or laser diodes, or charging batteries (at least certain kinds of batteries), as long as you don't try to do more than one of these at a time.

But that "power external peripherals" raises red flags. What kind of external peripherals are you planning to power with a constant current supply? Even if you have them, you can't power them at the same time that you are doing anything else for which you want a constant current.

So the first thing to consider is whether this thing, if built, is really going to do what you are hoping it will?

If not, then the best route is probably to decide what things it is reasonable to do with it, and take anything else off the table. The alternative is to rework your description of what you want into one that can reasonably serve all these masters. The latter will likely lead to significant complications, both in the specs and in the design. It might be better to take things off the table and plan to use a different, better-suited supply for those things.

Assuming that you still want an adjustable constant-current supply with a voltage limit, what range of currents do you want/need the current output to be adjustable over? What is the maximum output voltage that you want it to be able to output? What is the range of voltages that you want to be able to adjust the voltage limiter to?

Keep in mind that if your max current and your max voltage combine to result in a high power level, that the design gets more complicated. Quick. You might have to settle for a supply that is derated, meaning that the higher current, the less the max output voltage is at that current. Implementing this also complicates the design, unless it is done by either the user being responsible for applying the limitations, or the limitations are imposed by having the unit shut down if a temperature sensor indicates it's getting too hot.

The more power that the unit has to dissipate and/or deliver, the more attractive using a switch-mode supply becomes.

Be careful not to wish for the moon -- the more you want, the harder it's going to be to design and implement and the more it's going to cost. Be driven by an honest assessment of what you NEED, not what you would like.