LM317 has a sister...

[Jess]

...don't know if this has been posted before or not...

A while back I needed a driver that could function with a lower supply voltage. I'm a big fan of DDL's and derivatives of same.

It turns out that the LM317 has at least one sister: the LM1086. See www.national.com for more info.

The 1086 is nearly identical to the 317, except that it has only a 1 to 1.5v dropout (the lower the current, the lower its dropout voltage).

The reference voltage is still 1.25, and the chip is otherwise pretty much identical to the 317.

So if you build a DDL circuit with the 1086, you can have lasers that operate on lower supply voltages.

Example: If your supply is three AA NimH batteries in series, or two or three 14500 batteries in parallel, your power source is 3.6 nominal (4.1 peak).

For a laser operating at ~250mA and ~2v, a 1086 driver circuit will reduce that voltage from 3.6 to 2.6 (the 1086 dropout is only ~1v at this current level). The 1N4001 diode will reportedly take another 0.7 volt, leaving you at ~1.9. The current will be whatever you establish with the resistors.

If this has already been posted, sorry for the repeat!

 

[TheMonk]

Good find. [smiley=thumbup.gif]

If this works you can power a 803T with only 2x 3.7v.

 

[Benm]

I dont know if it has been discussed here, but these lower drop out regulators exist, are pretty well available and actually work.

The issue is that they are not very good current sources either, since they still burn 1.25v in the refernce and 1 - 1.5 in the regulation, about 2.5v in total. In case of driving a blu ray diode from 2 lithiums i guess its luck, and the margin is there.

Otherwise, current sources that eat less than a volt can easily be constructed from discrete components.

 

[TheMonk]

I was too excited and forgot to add 1.25V to the 1 volt dropout 2.25V total. :-[

This is identical to LM1117T-Adj.

 

[jake21]

cool time to buy some, any one no where they sell them at?

 

[Jess]

I dont know if it has been discussed here, but these lower drop out regulators exist, are pretty well available and actually work.

The issue is that they are not very good current sources either, since they still burn 1.25v in the refernce and 1 - 1.5 in the regulation, about 2.5v in total.

Hmm.  I'm a newb so I may be wrong, but I think that when these regulator chips are used in the constant current regulation mode, the primary difference is that there is no ground for Vadj or Vout poles between the chip and the load.  The ref voltage is only a potential between Vadj and Vout, but the only place for it to get "burned" is at the load downstream.  

The chip simply supplies whatever Adj current is necessary across the resistor between VAdj and Vout to sustain a 1.25v potential across Vadj and Vout.  The Vout voltage "floats", but the reference voltage is not lost because it is actually a component of Vout.

Therefore, total dropout from 1086 = 1 to 1.5 volt.  

What differentiates 1086 from 117 & 317 is that 117 and 317 use a Darlington NPN for the internal pass transistor, which has a two-diode voltage drop.  The 1086 uses a single NPN, therefore only 1 diode voltage drop.  (I'm just quoting this verbatim from page 8 of the 1086 data sheet).  The dropout for 117 and 317 are the same, and are anywhere from 1.2 to 2.5v depending on load and temperature.  See page 6 of the 117/317 data sheet.  

Otherwise, current sources that eat less than a volt can easily be constructed from discrete components.

Link?  

 

[Jess]

cool time to buy some, any one no where they sell them at?

TO-263 - Digikey - LM1086IS-ADJ-ND

TO-220 - digikey - LM1086IT-ADJ-ND

They cost $2.42 each

 

[Bionic-Badger]

Sheesh, for that price I'd just rather pay for a Groove Driver. 1V dropout.

 

[Benm]

Link?

Something like this:

http://www.delabs-circuits.com/png_circuits/discreet_circuits/current-source.png

You can build a mirrored version witn pnp transistors if you have a case-grounded diode (most reds).

I've used the mirrored version with a BC557 for the transistor on the left and a BD140 of the one on the left. The 4.7k resistor should be lower when powering from just a few volts, i suggest 470 ohm.

Also, a capactor from the base of the transistor on the right wired to ground makes a slow-start of sorts, protecting your diode. I suggest around 10 uF though this isnt critical.