How do you get 0v from an LM317?

Status
Not open for further replies.

Fudgecake

New Member
I'm a fairly new student to Electronics and I have to build a variable voltage regulator for my college project. I've got a fairly good idea how to do it using an LM317 but I don't know how to get the LM317 to go below its reference voltage of 1.2v. There is some info on the National Semiconductors datasheet, but when I modelled it on Multisim I couldn't get it to work very effectively. Any thoughts out there?
 
Don't connect the bottom end of the 317 to zero volts, connect it to a negative reference instead.
 
Thanks Nigel. I did try connecting the adjust leg to -10v on multisim 7 but I still seemed to be getting 0.3v when the pot was fully off. Is it possible that mutisim 7 would be giving me duff info. I'm due to have a crack at building this on Thursday but I'm not confident about it at the moment.
 
DigiKey have LT 3080 for $3.75 plus $4.95 post = $8.70
Wow, the 555 timer (for the negative power supply if AC is unavailable), 741, an LM317 plus a few resistors and capacitors required to make my circuit will works out cheaper!
 
That's a lot of added complexity to save a few bucks. I prefer the KISS approach myself.
 
Kiss is good up to a point. If you look at the actual 'complexity' of the internals of the 3080 I'm sure it's a lot more than an external converter. You just don't see it, it's black boxed. If an external circuit is well designed it can often be 'simpler', if you're counting total number of active devices. And if an IC goes you're often dead in the water unless you can rework it. Discretes still have their advantages.
 
I doubt that an LM317 is any simpler than an LT3080.

And I generally view simplicity in electronic projects as the number of external parts and the number of external connections, not the number of active devices that the IC may have. Those are the biggest factors in circuit reliability and ease of construction. The number of internal active devices and connections in an IC are of interest, but only from a academic viewpoint since their number has only a small effect on reliability of the IC which in most cases is longer then the useful lifetime of the assembly anyway. (When is the last time you heard of a PC failing due to the CPU chip dying, and that's probably the most complex IC in a consumer product with more than a 100 million active devices).
 
It depends on what resources you have.

I have loads of 555 timers, :mu:A741s and LM317s in my junkbox but I have no LT3080s. It would take far less time to build the relatively complex circuit than it would for me to build the simple LT3080 circuit because I'd have to wait for the parts to arrive.
 
Certainly, availability of parts (and perhaps some degree of impatience) can affect the circuit design.

But, my preference would be to take take the time to get the parts to build a simple design. I'm just a big fan of simplicity. You can often design a complex circuit to do a job. The interesting design work it to make it as simple as possible and still work.
 
You can do its this, providing you dont mind screwing up the output voltage regulation with output current changes in the load.
And changes with temperature.
Better to take an LM4040-1.23V reference and connect it from GND to a negative voltage which sets up a very stable -1.23V reference point to tie the LM317 to.
 
LT3080 any day. It's concept is brilliantly simple. Set up a current source on the input that has good PSRR. Generate a voltage with the current source using a resistor. Then the internals of the LT3080 is just a power buffer - the output is a 'power' version of the input. You keep the amplifier inside the 3080 at constant gain (so have more control over the stability) and if you set the input resistor to 0 Ohm, you get 0V on the output. I would use the LT3080 any day. I'm a simpleton
 
Status
Not open for further replies.
Cookies are required to use this site. You must accept them to continue using the site. Learn more…