EEVblog Electronics Community Forum
Electronics => Projects, Designs, and Technical Stuff => Topic started by: ee851 on April 02, 2013, 03:22:12 pm
-
Am thinking about making the "laboratory power supply" shown in Figure 22 of this LM317 datasheet. (http://www.onsemi.com/pub/Collateral/LM317-D.PDF)
Any suggestions on what to use for the 36V and the -10V DC mains-powered power supplies? The simpler, the better would be my first guess. Since this is a linear PSU anyway, not sure it would make any difference using linear or switch-mode power supplies. Probably not.
-
There are a LOT of different places, SMPS, a rectified transformer from mains, stacked boost converter, batteries.
However, if Vout is say... 1.25v and Vin is 36v, and your load is using a high current, your 317s will be pissing away a massive amount of power (creating heat), which you will have to remove.
One of the best solutions is a tracking pre-regulator; as the name says, it pre-regulates the voltage to a desired level to minimize power dissipation of the linear regulator, but most importantly, it is TRACKING, so it's output voltage follows the desired output of the final regulator, keeping the pre-regulator voltage at a desirable level.
Dave did some videos on it here;
EEVblog #260 - Tracking Pre-Regulator Simulation in LTspice - PSU Part 13 (https://www.youtube.com/watch?v=iTxKCQYMHbY#ws)
EEVblog #329 - Tracking Pre-Regulator LTspice Simulation Part 2 (https://www.youtube.com/watch?v=myqytW9ecww#ws)
-
There's a slight little difference between switchmode and linear ones - the linear ones are usually more stable with less ripple, but these are less effective... The switchmode ones are more effective, but there's a lot of ripple on the output and you have to cope with the EMI stuff, because of the inductors there.
I'm not sure if a LM317 is the best choice for the lab supply, but I think it should be good enough. I think it doesn't offer 0V output capability though. But if that's not a big problem, it should be ok.
I haven't designed any PSU yet, but i think designing a linear one is fairly easy.
Just use a nice transformer connected on the primary to the mains, DON'T FORGET A FUSE THERE! You never know when you goof something up, and I don't think you'd like to blow your PSU up... And on the secondary side connect it to a Graetz diode rectifier, then add some massive capacitor there, then put your LM317 there, and then add another massive capacitor to the output. Piece of cake i think.
To add a negative supply there, I might maybe choose another LM317 there which would be connected that way that the input would be connected to ground and the 'ground input' would be connected to the positive. Then I'll maybe add a switch there on the front panel to choose between the supplies...
It might be possible to use a single LM317 there, and there would be a simple switch there on the front panel for the output polarity, just some 'cross' connected dual switch there.
If you want to controll it with a single pot though, i think the LM317 is probably not enough, because the difference between your demanded 36 and -10 volts is 46, and that's 9 volts more than the maximum output rating from the datasheet. It might be also possible to cascade them up, but I don't know how much reliable that would be, and you would need a dual secondary transformer there.
If you don't really need 36 volts top, you might slice it in between, so it would be +18 to -18, and simply offset the LM317, which would be probably quite easy with three output transformer. I think I'll sketch something up and post it here.
Not sure if I was any helpful, but at least I tried.
You should really check out Dave's videos about making a power supply, he made a whole series about it!
--Richard
-
Small correction to the above post:
For a negative lm317 use the lm337(same as lm317 but it is for negative voltages)