Three pin positive regulators

[eeka chu]

If I'm using something like an LM317, the ADJ pin has to pull current up through the lower resistor, R2, in the ADJ chain right?

Would this current be being pulled through the resistance by the Vref or the supply's output voltage?

I'm trying to work out a maximum resistance I can use for my ADJ chain so as to minimise current wastage, as this is very important in this application.

[richie00boy]

No current is pulled up the adj pin in the way you describe. The critical resistor is the one from the output to the adj pin, usually recommended as 240 ohms in the datasheet. All the reg does is output a voltage to make 1.25V appear at the adj pin.

You trade error/noise by making the resistor chain higher impedance. The minimum current requirement is detailed in the datasheet. Bear in mind also that the reg needs a minimum load to function as well.

[janneman]

... and, there is also a current 'coming out' of the adjust pin going to ground. That current is sourced from Vin through the internals of the regulator. It's a small current, see the data sheet, but if you make the two set resistors very large, the adjust pin current will make the output voltage deviate from the usual equation of R1 and R2. See the data sheet for the complete output voltage equation including the nominal adj current value.

Jan Didden

[eeka chu]

Thanks for the replies. A lot of the datasheets just imitate each other, so it's helpful to hear things said in a different way.

However, I'm still a bit confused about the ADJ pin's current. The datasheets refer to the current being drawn from the pin. How can the current be drawn, sourced, to the ground rail from a device that is connected to the positive rail?

Being dumb and looking at the + and - symbols on the rails, I always expected the pin to be sinking current.

I should say I, obviously, don't keep up to speed on the internal operation of regulators.

[Lisandro_P]

Originally posted by richie00boy
You trade error/noise by making the resistor chain higher impedance. The minimum current requirement is detailed in the datasheet. Bear in mind also that the reg needs a minimum load to function as well. [/B]

I never had a LM317 oscillating because of a insufficient load, but the 78xx/79xx series are notorious for this. Still, a 1k resistor from the output pin to ground is all that's needed.

[peranders]

I have made an excel file for calculate voltages with this Iadj in mind.

Maybe this can be at some help.

LM317 calculation

[peranders]

Originally posted by eeka chu
However, I'm still a bit confused about the ADJ pin's current. The datasheets refer to the current being drawn from the pin. How can the current be drawn, sourced, to the ground rail from a device that is connected to the positive rail?

If you use my Exel file you don't have to think so much. I have only used information which is available in the datasheet.

[TwoSpoons]

If current consumption is such an issue, there are plenty of other regulators to choose from with incredibly small quiescent currents.

Have a look at some of the chips from Maxim and Linear Tech.

[jackinnj]

if all else fails you can modulate the ADJ pin and use the LM317 as an amplifier (yes, it really does work!).

keep your eyes peeled for LT1963 on EBay -- they show up every once in while.

(oh, the LT1963 has 5 pins, btw)

[eeka chu]

Thanks for your help everyone. The regulator in question isn't an LM317, but it's operation is the same.

This might sound really stupid, but I just want to check. When the manufacturer quotes an ADJ pin current, that's the very minimum current that must flow down the entire length of the ADJ chain assuming the load is above that which is required for regulation?

What is confusing me is that they say the Iadj current is being sourced from the ADJ pin, which is making me think it's only passing through one of the two resistors.

I understand that the regulators need a minimum load current to maintain regulation, and that this can be drawn down the ADJ chain rather than some other current sink or through a constant load.