LM2931-05 not giving +5V

Hi, I knocked up a setup for testing the output of a LM2931-05 low dropout 5v reg:

Schematic:
http://picasaweb.google.com/HappyRigpa/Electronics#5357916790944817874

LM2931 datasheet:
http://www.datasheetarchive.com/pdf-datasheets/Datasheets-22/DSA-424657.pdf

The circuit is powered by a 9v pp3 battery (measures just over 10v).

Am I a fool for expecting +5v from pin 1 of the 2931? What I get is 9.15v

If I remove R1, I get +3v from the reg. I have 5 of these and they all behave the same.

The 2931 is good for over 100ma. Does anyone know what is happening here?

PS The reg is socketed so it's not heat damage. The input cap is 470n 25v tantalum & the ouput cap is 100uF panasonic FM low ESR.

Thanks for helping

 

Which package are you using? They have different pinouts.

Try using a smaller value of resistance in the LED path. 10 - 20 mA should be o.k. for the LED and the value can be reduced later after the regulator is up and running. The datasheet suggests there can be stability issues with the part that require the 100 uF capacitor you have in the design. Is the output a DC value or is there an AC component to it?

 

You have the pins of the regulator connected wrong.

 

Cheers folks,

It's from National Semiconductor, TO92 package, the pins are definitely, 100% connected as per the datasheet. I have checked this about 7 times

That's why I'm using a top of the range panasonic FM cap. 40 to 70% lower ESR than FC range. Should be ok???

I've only used TO220 regs before and never encountered issues like this regardless of the cap values used. Is it a case of playing around with the load/cap values until stability is reached? Anyone had these kind of problems with TO92 regs before?

Thanks again.

 

The pin layout of the TO-92 case can trip you up.

Holding it with pins down so that you can read the printing on the front of the case, the pins (left to right) are:
Out GND In

This regulator is protected against accidental reverse insertion for a short period of time. It won't hurt to try it the other way around for a few moments.

 

I tried with pins 1 & 3 reversed and it gives 4.89 v. Fantastic, if slightly worrying as this is not how the pins are laid out in the datasheet.

Do manufacturers often wire things up differently? Please don't say I made a mistake reading the datasheet, I haven't, they are definitely reversed.

Also, since I had it the "wrong" way round to start with, shouldn't it have not passed any current at all? It is supposed to have built in reverse polarity protection.

 

If you mean does one manufacturer wire something differently from the other. Yes. Do they wire it differently than their datasheet. Not typically. The updated datasheet at National Semiconductor from June 2005 still shows the same pinout. Different manufacturers show different views of the to-92 package so as SgtWookie stated they can get mixed up easily. National uses a bottom view here.

If there is truly an error, let National Semiconductor know. I am sure they would appreciate it.

 

mistakes are usually made when we are definitely 100% sure that they are not made.

 

I have used many LM2931A-5.0 regulators in the TO-92 case.
The bottom view of the pins in the datasheet is correct.

Maybe you have some counterfeight ones.
Did you buy them at RadioShack?

 

Working is good. The pin reversal is odd. Did you buy the regulators from an authorized distributor, such as Digikey, Newark, or Mouser?

What is the exact part number on your regulators?
LM2931Z-5.0? LM2931AZ-5.0? Are you certain that they are National Semiconductor parts?

It has mirror image insertion protection for itself, not necessarily for your circuit.

What is the Vf of your LED? You should change R1 to allow around 15mA current through the LED when the output voltage is correct.
R1 >= (Vsupply - Vf(LED)) / Desired Current
R1 >= (5v - Vf(LED)) / 0.015A
Continue on from there.