LT1083 Oscillating

[Bassterror]

Hi,

I'm trying to build a regulated supply as it is described here:
http://www.shine7.com/audio/pa100.htm

When I tested it (without any load), the regulators behaved quite bad, I had oscillation....

No capacitor on the output - OK, I soldered 1µF
-> little improvement, but still bad oscillation...

Then I tryed 100µF
-> even more oscillation, LT gets slightly warm, elko gets hot, I hastily pulled the plug...

Does the 1083 need any resistive load on the output to funktion?

[Corax]

Hi Bassterror,

the LT1083 is a commonly used regulator and has usually no issues (at least to my knowledge).
However, if the regulator (starts to) oscillate some sort of positive feedback must be present.
You mentioned a (bleeding) resistor at the output you might wanna add - it could help but it's more likely NOT the reason for the oscillation.

If you're using exactly the same schematic as in the provided link, try to increase the capacitors C2 and C5 at pin 1 to GND instead of the bulk capacitors C3 and C6. You might try to add a small 100nF in addition to it (for better high frequency response).

Another often reported issue (in general) with "unstable" regulators is the PCB layout with a bad GND and/or feedback routing. A picture of your layout and what you have soldered so far might give us/me a hint what went wrong with your regulator.

For further information read also the datasheet (in particular the chapters under "Applications Information" (i.e. "Ripple Rejection", "Load Regulation", ...)

Greetings

Corax

[jackinnj]

The author of the PA100 article should have read the product folio of the LT1083 before designing the supply -- the LT1083 is a low-dropout regulator and the capacitors (C3, C6) on the output rail are part of the feedback loop. Oscillation should have been expected.

Go back and read what Linear or National Semi have to say about the care and feeding of Low Dropout Regulators.

[QSerraTico_Tico]

Originally posted by jackinnj
The author of the PA100 article should have read the product folio of the LT1083 before designing the supply -- the LT1083 is a low-dropout regulator and the capacitors (C3, C6) on the output rail are part of the feedback loop. Oscillation should have been expected.

Go back and read what Linear or National Semi have to say about the care and feeding of Low Dropout Regulators.

Actually a 100 to 150uF cap is the recommended value in the data sheet of the LT1083......
http://linear.com/pc/downloadDocumen...55,P1281,D3741

[Corax]

Hi folks,

it seems to me that are many opinions are around here of how, what, when and where is the reason for the oscillating of the LT1083.

Well, I read again the datasheet from the beginning on and found something interesting just on the first page:

"The LT1083/LT1084/LT1085 devices are pin compatible with older three-terminal regulators. A 10µF output capacitor is required on these new devices. However, this is included in most regulator designs.
Unlike PNP regulators, where up to 10% of the output current is wasted as quiescent current, the LT1083 quiescent current flows into the load, increasing efficiency."

To the first paragraph we can say the output capacitor C3/C6 has 100uF and much more as needed (10uF) and a little bit less than in other applications in the datasheet.
The second paragraph refers to the quiescent current and is always in the limit caused by R1/R2 (1.25V / 100 ohm = 12.5mA, required are, concerning to the electrical characteristics table, in worst case 10mA as the "Minimum Load Current") and R4/R5.

After all I see no abuse or misuse of the LT1083. Either there's a discrepancy between the schematic and your solder work so far or the regulator might be damaged somehow.
Speaking of solder work. Have you made some pictures of the PCB (copper and component side)?


Greetings

Corax

[Bassterror]

Hi

thanks for ur replies.
Gonna make a picture tomorrow.
Although I did it quick and dirty, I do not believe a made serious mistakes...