[Radioroslyn]

Truth be told most of the times I've used the '317 is w/a 6-7v input. So the voltage drop (dissipation) across it has been minimal.  The 1st time was a copy of a circuit out of Popular Electronics back in the 80's. It used the 3A version but powered it w/a 36v transformer for a 6v output. Used a big heatsink and ran hot. I didn't really understand how it worked then and thought it needed the 36v to operate properly. This isn't the case the closer the input voltage is to the output voltage the better the IC likes it and will run cooler.

When I putting this one together I used a small transformer w/a hv winding and a 6.3 filament winding. Nowdays these transformers kind of expensive. That's why I picked the one from AES. It's close to what we need to terms of voltages and current but mostly it's easy to get. The HV side rated @ 200ma which is abt 10X what you need to power any battery set requiring A and B batteries. The divider eats up some of the current @ 6000 ohms it will consume abt 26ma. Still leaves 174ma. On the A side the voltage drop across the IC is about 8v. @ 200ma it's about 1.5w. Worst case is if you didn't heatsink it might overheat but it has a fail safe if it does it goes open till it cools then goes back to normal operation. If this is the case it's simple enough mount the IC to a scrap of aluminum or to the p/s metal chassis. Must use an insulator and some heatsink grease.