Hi,
I am building annother valve portable supply as a really nice period case and transformer have been sourced.
Until recently, I supplied 1.4v from an LM317 voltage reg. However, discussion elsewhere worried me as people with far more experience have told me that this is prone to failiure and of course blown fillaments.
So I have now been using 2 alkaline D cells in parallel to supply LT.
I would like to return to a transformer derived supply and would like members views on voltage regs. or other circuits and is the LM317 really a bad choice?
Many thanks,
Rob
The LM317 should be fine if properly used.
Correct capacitive loading (only 0.1uF on output) and decent heatsink.
Check the National Semiconductor's data sheet and application notes. Don't trust random Internet designs.
You can put a 200mA fuse and 3 x IN5401 in series on the output. If the LM317 goes "short" then the rectifiers blow the fuse at about 1.8V (a fresh battery can be 1.7V, briefly).
Check it with 5 x 47 Ohms in parallel as a dummy load.
Pick the two feedback resistors to have 1.4V as the nominal voltage (the battery is 1.65 down to 0.9V)
Click to see full size image
Zinc Carbon discharge curve for modern AA cell. The only difference with original cells is the duration at the approximately 12 Ohms load of a Dx96 set (125mA). A xx91 series is about 200mA I think, or about 7 Ohms load.
Thanks for that, I had thought of strapping diodes across the output as you suggest. In my original design, the LM317 dropped only about 3v at output current of 125mA. This is well within the device spec so I cannot see failiure happening.
My concern only came about after several people saying "you don't want to use a 317, they always fail..."
Rob
Maybe they had inadequate heatsink, more than 40V in (200V is OK if o/P is 170V till it gets a momentary short on o/p), driving relays, solenoids etc.
No worse than any other IC. It's purely a design/application issue, not inherent. Make sure a reputable make and not a counterfeit.
I've been using them for years and never had one fail even under/after abuse. Certainly the diode chain will save any nonsense.
- Joe
Agreed! Think it's all nonsense about 317's failing. I've been using one for years in a battery valve PSU as LT supply. I've also found them in commercial equipment that is many years old and had heavy use and still working.
Use them within their parameters and they should be fine.
SB
i have only had 1 failure that was in a tv set but it had never been bolted to its heat sink even so it had lasted 5 years
rob t
Thanks all, will continue to use my LM317 and hopefully will not be paronoid about its output!
Rob
The LM317 was designed by legendary Nat Semi engineer Bob Widlar; use with confidence. 😀
