Symmetrical adjustable regulator

[Olegkuskhov]

Hi i design a symmetrical adjustable voltage regulator with lm317 and lm337. I plan to boost the current capability by using transistor at the output stage, but thats another story. I fed 15 Vct supply to it and through a bridge rectifier and after the capacitor filtering i got arround 21 and -21 Vdc. I then install lm 317 and lm 337. The lm 317 is working corectly, but the lm 337 not. I use 1k and 10k resitor for adjust pin. The 1 k is used to set maximum voltage at the output and the 10 k is used the adjust the output.

The problem i faced is. The input pin and adjust pin of lm 337 shorting for no reason after i plug the transformer to mains. I know this by checking the continuity after desolder the lm 337. I thought that the protection diode i installed is wrong polarity or i short some traces in the pcb, but nothing seems shorted. I checked the bulk capacitor and it seems the capacitor is good. What can go wrong here?. I thought its just a simple things. Hope some can help me😔

[Olegkuskhov]

Additional notes. I manage to make the lm 317 and its current booster to work and nothing bad happen. With 21 Vdc. I can get arround 18.5 Vdc max at output after the current booster.

I made the pcb my self. I attach my back copper layout below.

[soldar]

You are using a single diode bridge. This forces all the current to go through both sections. This is the first thing I would change.

[Roehrenonkel]

Hi Olegkuskhov,

the resistors between "out" and "adj" are way too high.
The regulators need a minimum current, take 124 Ohm or less.

[Olegkuskhov]

Hi guys, thanks for all the reply. I will try to modify the adjustment circuit later. I'm waiting for the replacement of lm 337. So i decided to install  Im 7915 that lying arround on my workshop. I bypass the adjustment network so i goes straight to the center point and i decided to try it with the current booster and it works just fine. I got -13.8 V , about 1,2 V drop from -15 V after the output pair transistor. Below i attach some footage of the project. I hope when the replacement come, it will work as an adjustable one. 
 

[Terry Bites]

Transistor polarities are wrong so they just act like a diodes linking the regulator inputs to the their outputs.
There are no current sense resistors to turn on the power transistors.
See basic circuit attched. Adapted from datasheet applications.

As it stands the supply has no short circuit protection. You might want to consider adding that in.

[Olegkuskhov]

Thanks for your effort terry will try that soon. I plan to make full feature power supply with cc, cv and protection in the future.  Yes, the resistor you mention that in series with the transistor is indeed needed. For the current sensing shunt, i just elimalinate that feauture right now because i want the pcb dimension is limited by the heatsink underneath (just odd to see the if it's way bigger than the heatsink).

I've never thought about the short circuit protection like you just made. Is it possible to minimize the voltage drop per amp?. Maybe by using Mosfet and operates it on linear region?