Help with powersupply Problem.

[diegotheslinger]

Hello, lurked on here for a bit and decided to register.

My problem is with the powersupply that I recently put together. Its based off the lm117 and I have included a schematic. Problem is, it is giving me voltages it shouldn't after the regulator. Something like 0V - 22V. (The LM117 shouldn't go down to 0V right? Should go to 1.2v) Also, the biggest problem, when ever I put anykind of load on it the voltage drops considerably. Heres some of the load Voltages I got:

19.94v unloaded
19.9v w/ 100000 resistor
19.56v w/ 15000 resistor
18.7v w/ 4700 resistor
5.02v w/ 100 resistor

so it seems that with a small resistance load, there is a huge voltage drop from what I want. What would cause this? I put the values in on multisim, and it seemed to do fine. (including having a voltage range from 1.2V-22V) I have checked all my connections, and they all seem good. Doesn't seem to be any shorts. Am I missing something stupid here or is it a design flaw.

I have included the schematic and two pictures to check out. Maybe you can offer me some advice?

--- Schematic ---
**broken link removed**
---- Front ----
**broken link removed**
-----Back ----
**broken link removed**

 

[diegotheslinger]

Changed the the value of resistor R3 (260ohms) to the proper value of 120ohms for the lm317. Still same problem dropping voltage when a small load is used.

 

[aquamon]

You have a bank of 4 capacitors in parallel, what for?
Remove D5.

 

[crutschow]

You are missing the worst-case minimum load for a LM317, which is 10mA (see the data sheet). Thus for high load resistances that draw less than 10mA the voltage may rise above the set point. For a nominal 5V output, for example, the minimum load resistor (including the voltage set resistors) is 500Ω.

Your simulation model may not include that limit which is why it works while your actual circuit doesn't. You've learned one of the deficiencies of simulation, that it's only as good as its models.

 

[crutschow]

You have a bank of 4 capacitors in parallel, what for?
Remove D5.

D5 is to prevent damage to the LM317 by providing a path to discharge the output capacitance in case the input is accidentally shorted.

 

[aquamon]

ya, but his input voltage is already too low...
Make it simple, https://www.electro-tech-online.com/custompdfs/2010/05/LM117.pdf
bottom of pg. 18: 1.2V–20V Regulator with Minimum Program Current

 

[mneary]

The LM317 pin out is different from other regulators (e.g. 7805 etc.). You have switched pin 1 and 2.

 

[aquamon]

actually, looks like all pins are wrong...including pin 3. good find mneary, lol..

 

[diegotheslinger]

yep that was was it. i was looking at a wrong datasheet. switched the pins and it works great now, with no voltage drop and it is in the correct range of 1.2v to 24.32v. I spent way too long checking my connections and didn't even think to look at the orientation of the regulator. btw aquamon, the reason that have the 3 caps is cause thats all i had laying around, and i wanted a large filter value to help cutdown on ripple before the regulator. the 4th one (c4) was called for in the datasheet, but i don't think its needed. (the regulator is closer than 6in from the filter caps). I just kinda threw this together with parts i salvaged.

Thanks for pointing out the pin placement mneary!

 

[diegotheslinger]

Heres the power supply put together....the camera doesn't pick it up well but you can read the Voltage LED readout fine thru the mesh. The meter is something I found broken at a thrift store and put it all in a PC psu case. You also can't see it but the whole thing glows green thru the mesh when its on.

**broken link removed**

case on

**broken link removed**

 

[aquamon]

...and i wanted a large filter value to help cutdown on ripple before the regulator.

At 60 Hz, overkill.

 

[Hero999]

You are missing the worst-case minimum load for a LM317, which is 10mA (see the data sheet). Thus for high load resistances that draw less than 10mA the voltage may rise above the set point.

Possibly, but it depends on whether he's using the LM317 or the LM117.

In his post he stated it's the LM117 but the schematic shows the LM317.

I doubt this is the problem, as the regulator is obviously working in drop-out mode when R5 is at its maximum setting, which should give 24.4V but the input voltage is too low give that when a load is connected.

You have a bank of 4 capacitors in parallel, what for?

Presumably because he doesn't have a large 2200µF capacitor so he's make one from three smaller capacitors?

The smallest capacitor is required because it works better at high frequencies.

Remove D5.

Sorry, that's a bad idea because it helps protect the regulator against reverse voltage when large capacitors are connected to the output and the input voltage is reduced.

aquamon ,
How much current do you need?

Are you using the LM317 or the LM117?

The LM317 is limited to 1.5A, it hardly has a heat sink and the filter capacitors are too small to enable you to draw much current when the output is set to over a few volts.

 

[aquamon]

aquamon ,
How much current do you need?

Don't need any currant, maybe some celery...

 

[diegotheslinger]

it is a lm317, i miss labeled it in the picture.