| Author |
 Topic  |
|
hillbillyboy_316
Apprentece
12 Posts |
 Posted - Dec 10 2009 : 12:31:47 AM
|
| i build a copy of aarons power supply based on the lm338k regulator and i cant get any variable control out of the 5k pot. what could be the problem? |
|
|
audioguru
Nobel Prize Winner
Canada
4218 Posts |
Posted - Dec 10 2009 : 11:45:27 AM
|
Maybe you have the pins on the LM338 regulator connected backwards.
What does your circuit do? What is its output voltage?
The LM338 is supposed to have 120 ohms (not 240 ohms which is for the more expensive LM138).
Then its pot should be 2.5k not 5k. |
 |
|
|
hillbillyboy_316
Apprentece
12 Posts |
Posted - Dec 10 2009 : 1:54:48 PM
|
| i got the wiring right on the lm338k regulator. as for as the resistor and the 5k pot i took the needed components off of the schematic from this site. the voltage goes from 1.2 to 32 volts |
|
|
|
audioguru
Nobel Prize Winner
Canada
4218 Posts |
Posted - Dec 10 2009 : 5:48:48 PM
|
The schematic from this site is wrong.
The datasheet for the LM138/LM338 from National Semiconductor shows a minimum load current of 5mA max for the more expensive LM138 (1.25V/240 ohms= 5.2mA) and has a minimum load current for the LM338 of 10mA max (1.25V/120 ohms= 10.4mA). The text says "If there is insufficient load on the output, the output (voltage) will rise.
All the LM338 circuits in the datasheet have 120 ohms for R1 and a note says it can be 240 ohms for the more expensive LM138.
What good is a regulated power supply that has its output voltage go too high without a load?? |
|
|
|
hillbillyboy_316
Apprentece
12 Posts |
Posted - Dec 10 2009 : 6:32:11 PM
|
| so i need to change the resistor to 120 ohms and the 5k pot to a 2.5k pot and everything should work? are all the other values on the schematic right? |
|
|
|
audioguru
Nobel Prize Winner
Canada
4218 Posts |
Posted - Dec 10 2009 : 8:22:06 PM
|
The max voltage will be about 27V if the resistor is 120 ohms and the pot is 2.5k ohms.
The max voltage will be about 32V if the resistor is 100 ohms and the pot is 2.5k ohms. |
|
|
|
hillbillyboy_316
Apprentece
12 Posts |
Posted - Dec 10 2009 : 8:24:16 PM
|
| everything else in the schematics right? is the 5k pot why the resistors are getting hot and burning up? and why i dont have any voltage adjustment? |
|
|
|
Aaron Cake
Administrator
Canada
6717 Posts |
Posted - Dec 13 2009 : 10:47:33 AM
|
I hate to say it, but I have built probably 50 supplies with that exact schematic and all have had excellent regulation. I don't know how I originally came upon those values as I first built this circuit 20 years ago, but it does work as pictured on this site. I will certainly, however, update the notes on the page with the official specs from the datasheet.
If you are burning up the pot or the resistors, there is a major problem somewhere else. Are you sure the regulator is connected properly? Remember, the case is live (output I think, from memory). |
|
|
|
audioguru
Nobel Prize Winner
Canada
4218 Posts |
Posted - Dec 13 2009 : 12:39:45 PM
|
I don't know the percentage of LM317, LM338 and LM350 variable regulators that need a 120 ohm resistor to prevent the output voltage from rising without any regulation when there is no load. The datasheets warn about it so I think each one should have 120 ohms.
There are hundreds of these circuits wrong on the web because the first page of each datasheet shows 240 ohms (for the more expensive version of the regulators)and many circuits never have no load.
The 120 ohm resistor dissipates (1.25V squared)/120 ohms= 0.13W. A 1/4W resistor might be slightly warm. |
|
|
|
hillbillyboy_316
Apprentece
12 Posts |
Posted - Dec 14 2009 : 12:38:52 AM
|
| the only thing different is i used a 24 volt 6 amp transformer and a 28000 uf capacitor cause thats all i could find. everything else is the same and the resistors are burning to a crisp everytime |
|
|
|
audioguru
Nobel Prize Winner
Canada
4218 Posts |
Posted - Dec 14 2009 : 10:40:35 AM
|
The 120 ohm resistor dissipates only 0.13W and will be slightly warm, not burning hot.
The 2.5k pot dissipates 0.27W and a 1/2W pot might be warm.
Look at the pins of the LM338 in the datasheet.
I hope you used the bridge rectifier and connected it correctly. |
|
|
|
hillbillyboy_316
Apprentece
12 Posts |
Posted - Dec 14 2009 : 4:09:04 PM
|
| yeah i put a bridge on it to change from ac to dc. ive triple checked the regulator wiring and its right, even measured the case to make sure cause the 2 pins are off center. |
|
|
|
hillbillyboy_316
Apprentece
12 Posts |
Posted - Dec 14 2009 : 6:19:45 PM
|
| ok i tore down my board and rebuilt the schematic and found my problem. now when i turn it on i can only get up to 9.4 volts |
|
|
|
audioguru
Nobel Prize Winner
Canada
4218 Posts |
Posted - Dec 14 2009 : 6:46:50 PM
|
Then your resistor is 120 ohms and your pot is only 782 ohms.
Or your pot is 2.5k and your resistor is 383 ohms. But the output voltage will rise without a load. |
|
|
|
hillbillyboy_316
Apprentece
12 Posts |
Posted - Dec 14 2009 : 9:40:19 PM
|
| it is a 5k pot with a 240 ohm resistor |
|
|
|
audioguru
Nobel Prize Winner
Canada
4218 Posts |
Posted - Dec 14 2009 : 10:44:41 PM
|
Without a load then the output voltage will probably rise because the LM338 needs a 120 ohm resistor or less.
The more expensive LM138 can have a 240 ohm resistor. Then the pot should be changed to 2.5k.
Remove the 240 ohm resistor and 5k pot and measure their resistance then report what they are.
Measure the DC voltage across the main filter capacitor (the input to the LM338) and tell us what voltage is there. |
|
|
|
Topic |
|
power supply
