Head-Fi.org › Forums › Misc.-Category Forums › DIY (Do-It-Yourself) Discussions › Quick question about the LM317! (and other such regulators)
New Posts  All Forums:Forum Nav:

Quick question about the LM317! (and other such regulators)

post #1 of 9
2/8/11 at 8:38pm
Thread Starter 

i got one in my NAD C300 and it runs hot, damn hot, waaaay too hot to touch, so i thought i could take it off the board and heatsink it on the case, electrically insulated of course, would running 15 inches or so of some wires (nice thick solid core copper) affect it negatively somehow?

 

here is a picture of it (right above the bottom rectangular object)...look at the heatsink, it's peeling! :o

DSCF1297.JPG

 

i think it's because this amp has a class A preamp(?) or "line driver section" (those rectangular things), or something....

 

 

cheers :D

Reply
post #2 of 9
2/9/11 at 4:45am
Has the part failed from excessive heat? Is it effecting operation in any way? (beyond being too hot to touch when enclosed inside a case where no one should be able to touch it).
Reply
post #3 of 9
2/9/11 at 5:03am

If it is running too hot to the touch, it would imply that something that it is powering is having issues and drawing excessive current.

 

And yes 15" of wire would not be ideal.  The LM317 usually should have an input cap close to the regulator (as per the National datasheet):

Quote:
Normally, no capacitors are needed unless the device is situated more than 6 inches from the input filter capacitors in which case an input bypass is needed
Reply
post #4 of 9
2/9/11 at 6:58am

It could work with blocking caps on the regulator, but it's not ideal. Can you just put there bigger heatsink somehow?

Reply
post #5 of 9
2/9/11 at 7:09am

the two things that determine how hot a 317 gets are the voltage drop it is aked to do, and the amount of current being pulled through it.  Your problem could be one of 3 things, as far as I can speculate:

 

1. the V in has gotten to high

2. current output is now too high

3. needed a bigger heatsink to begin with, and it is degrading over time.

 

any or all of these should be fixed, and you could start by doing some measurements to see what is going in and out..... and also check to see what resistor values are there-- should be a 240 and one more resistor which sets the v-out.  These are easy to look up to see what V the resistor should be setting the 317 at.

 

 

Disclaimer, I am not an engineer, but have done lots of 317/337 based psus, and they are pretty straightforward.

 

hope this helps

Reply
post #6 of 9
2/9/11 at 8:54am

Too hot to touch is not necessarily bad. I just took a quick look at its data sheet and depending on the packaging the maximum temperature is either 125 degrees C or 150 degrees C. The regulator would become way too hot to touch at much lower temperatures, so you may want to get an actual reading of the temperature before you decide it is too hot. Anything under 80-90 degrees C or so is probably ok

Reply
post #7 of 9
2/9/11 at 9:10am

To agree with bcg27:

Skin burns at about 130*f=55*c

 

Thats nothing compared to the absolute max, or even "long term" max temperatures of the chip.

 

PS: you should not be poking around in electronics. Even low voltage junk can pack a wallop. Get an IR thermometer if your interested in how hot something is. "too hot to touch" ranges from totally safe to literally on fire... "enough voltage to break skin and kill instantly" (if your unlucky) starts well within the voltages commonly seen in power amps (40-60V).

Reply
post #8 of 9
2/9/11 at 9:36am
Thread Starter 

thanks alot for the replies :)

 

part hasn't failed, it seems to operate perfectly fine despite the intense heat, doesn't 'cut off' or anything so i assume it's operating within spec (i don't own a thermometer, d'oh!)

 

glad i asked about the wire, i don't want to resolve one issue and create another with bad implementation, not sure i could fit a larger heatsink in there, ideally i want the heat away from everything else too! :)

 

fzman, i am looking at the schematic and for the LM317T i see a (edit...) 226 ohm resistor and a 5K2, both specified 1% metal film...

 

it might be possible the intense heat has degraded these, or is causing their resistance to drop and make the LM317 run too hot?...i will replace them anyhow with nice shiney new ones.

 

i'll take the top off later and do some measurements and that.

 

i have also noticed a 5v rail leading to the front panel IR sensor and microcontroller for volume control/muting/power led, none of which i need, it seems to be provided by another IC (lm7995?) which runs very cool...i could probably take this 5v and mount the quietest fan i can find to cool the heatsink, there is tons of space to work with inside :)

 

edit:

 

yes i really should be using an actual thermometer!  my real cause for concern is the heatsink is peeling and there's discolouration on the PCB, not good :/ i've been using proper tools to adjust bias and dc offset (plastic screwdriver things) and i checked the heatsink for any crazy voltages (between it and ground i think) before touching it :P


Edited by googleborg - 2/9/11 at 4:00pm
Reply
post #9 of 9
3/3/11 at 5:53pm
Thread Starter 

 

finally got round to getting it away from the PCB using teflon sheathed (very) solid copper wires, less than 6 inches including traces from the caps, and changed the trimpots for much better bourns 10k 12 turn ones, so... easy to adjust now! :)

 

interesting that the temp of the LM317t directly links to the DC offset at the speaker terminals!  i blow on it and the dc offset quickly rises a few mv, then back to 0 when it returns to it's operating temp...(the one i adjusted the offsets at)

 

DSCF1434.JPG

 

next change will be directly wiring some inputs to a volume pot and then right to the preamp and then to the power amp for a 'pure path' hahaha


Edited by googleborg - 3/3/11 at 5:59pm
Reply
New Posts  All Forums:Forum Nav:
  Return Home
Head-Fi.org › Forums › Misc.-Category Forums › DIY (Do-It-Yourself) Discussions › Quick question about the LM317! (and other such regulators)