Voltage regulator problem?

[arawn]

I recently obtained anold rack mount enclosure from work and it still had a powersupply in it. The powersupply is a bipolar 15 volt output unit, so it has -15v outputs as well as +15v outputs I thought that if I added a second set of voltage regulator circuits i could chop that down to 9.4 or so volts. I used a basic lm317 circuit and put 2 on one board as I had the chips and 2 15v outputs. when I completed the voltage regulators and hooked the to the powersupply, I tested and found that I had 28 volts on the output of the regulators. This should be impossible, shouldn't it? I checked the readings 3 seperate times I still only have 15 volts coming from the powersupply, but i have 28 volts on the output side of the additional regulators.

I used the scheme from here for a fixed regulator http://www.k7mem.150m.com/Electronic_Notebook/power_supplies/lm317.html
I used his calculator set up for 10% tolerance resistors and a 180 ohm resistor for "R3" I also asked it to calculate from a 15v supply, to give me 9.4 volts.
I am just baffled! 

[punkin]

Is the power supply loaded? Some supplies will float up without a proper load. What is the input voltage with the regulator connected? The regulator you've built compares the input to the output to determine regulation. Also as you're making input/output voltage measurements, are you connecting your voltmeter to the same ground point? I guess with this thought, is the LM chip referencing the same ground as the power supply in front?

Just a few ideas.

[arawn]

I checked the readings 3 seperate times I still only have 15 volts coming from the powersupply, but i have 28 volts on the output side of the additional regulators.

the power supply is puttin out 15.1 volts i even tried loading it and it held steady. the problem is somewhere in the regulator but everything checks out against a scheme but the 317's are both showing vref of 1.62 volts and outputting 28 volts.

[punkin]

That is very odd...any chance you've checked the voltage drops as described in the configurator tool?

"R3 to be 1.25 Volts, the voltage across R4 should be 8.15 V "

[arawn]

voltage drops are the same 1.62 @r3 and 28 @r4. Checked the bypass caps and none of them are in spec. Could this be causing some of this?

[head_spaz]

You have a single power supply... that is BIPOLAR... which means it supplies a positive rail... and a negative voltage rail.

That means you have a COMMON GROUND between them.

The LM317 is a POSITIVE voltage regulator.

If you are trying to use two positive regulators... how do you plan to SEPERATE the common ground?

You're going to have to either seperate that common ground... or use one positive and one negative voltage regular to accomplish your task. (I prefer to use the LM2990 and LM2940 regulators because they are low-dropout and super quiet.)

I'm guessing here... but I think you'll find that you're now using the negative rail as your ground. If so... you'd have +15V plus +15V added together... which makes 30V.

[arawn]

that sounds perfectly logical! Doh