Continue to Site

Welcome to our site!

Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

  • Welcome to our site! Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

+ and - voltage regulation

Status
Not open for further replies.

solis365

New Member
quick question about lm317s and lm337s:

say I want to have a + and - regulated supply. Would I need a center tapped transformer to use as the input to both of these, with the center as ground, and then rectify the + and - separately and send them into the chips? (lm337 handling the negative voltages, obviously, and lm317 handling positive)?

or, could i just have a transformer giving + and -, reference the - as ground, rectify, and use the lm317 with input being the + rail and the ADJ going to ground, and then use the ground rail as INPUT for the LM337 and connect the ADJ to +?

would the second option give a split setup, adjustable on both sides, one negative, one positive, referenced to a common ground?

There are schematics at the bottom.


Essentially, I have two questions: Will the first circuit work (I'm pretty sure it will), and, will the second work like I think it does, or did i screw something up with the references?

and yes, the schematics are not complete as far as hooking up the LM regulators, just imagine that the proper filtering caps and bypass diodes and such are there... just trying to get you guys a picture.

oh and the signal source would be a 30V or something transformed down from the 120V mains.

the reason for two rectifiers in the first circuit is to give both positive and negative full-wave rectification. they are hooked up in parallel, but oppositely. it makes sense to me that this would work, any reasons it wouldnt?
 

Attachments

  • schem1.GIF
    schem1.GIF
    7.3 KB · Views: 175
  • schem2.GIF
    schem2.GIF
    7.3 KB · Views: 172
Last edited:
Here is what is wrong with the left diagram:

The lower bridge rectifier needs to point the other way to produce a NEGATIVE voltage with respect to ground. You left off the two filter capacitors. R2 and R4 should be deleted. There should be bypass capacitors across both positive and negative outputs.

The right diagram is really screwed up. Positive reg is not sensing the output; look at the app notes. Neg reg is totally wrong. The diode rectifiers do not produce symmetrical positive and negative unregulated voltages. No filter caps; no bypass caps.

I were giving you a grade, it would be an F
 
Last edited:
The easiest way to generate both + and - with respect to gound for regulator inputs is to use a transformer with a center tap in the secondary. Connect the center tap to ground, the other two leads to the AC input of the bridge. The + out of the bridge and ground terminal will be + voltage, and the - out and ground will be -. The transformer secondary needs to double what you would normally use.
 
The lower bridge rectifier needs to point the other way to produce a NEGATIVE voltage with respect to ground.

Yes I just flipped the connection points to + and - around on that one, should be "horizontally" flipped, not vertically as I have it.

You left off the two filter capacitors.

i said that caps and bypass diodes arent there, i know where they go, just imagine them. its not a functioning spice model anyway, just a drawing to ask a question

R2 and R4 should be deleted.

meant to be pots to enable adjustable voltage regulation, forgot to explain that, either way the resistors there would change the output voltage whether they were pots or static, just without the pots you couldnt adjust them "on the fly"

There should be bypass capacitors across both positive and negative outputs.
see above

Positive reg is not sensing the output; look at the app notes.
sorry, forgot a wire there. you could probably tell I just forgot to it since the sensing connection was there on the other three regulators i had drawn up. -.-

Neg reg is totally wrong. The diode rectifiers do not produce symmetrical positive and negative unregulated voltages.
so, grounding the one end of the bridge rectifier is wrong? would it make more sense to run that connection of the bridge, instead of to ground, to the node below R2, and delete the ground connection there?

No filter caps; no bypass caps.
again, see above

I were giving you a grade, it would be an F
i suppose that I should have just waited to post until my schematic was entirely complete, rather than trying to post a schematic that just gets the "gist" of the problem across. i suppose when a schematic isnt complete its difficult for you to tell what i know and what i dont, and thus you must point out all the omissions in order to help.


my question was, if i DIDNT have a center-tapped transformer, could I "cheat" to get a negative voltge by referencing the negative regulator with ground as "IN" and positive as "ADJ"? (and still use a positive reg. to get a + voltage rail) the way im seeing it is that if you reference it this way, OUT should end up being negative. can it work like that or is there an error in my thinking?
 
Last edited:
You can't get negative voltage using the method you describe.

You can't just connect the DC side of two bridge rectifiers in series like batteries. Try tracing the current path through the diodes and you'll se what I mean.

You can't use a negative linear regulator to generate negative voltages - it needs a negative voltage to start off with.

Here's a voltage doubler circuit which will give both a positive and negative output.
supply-gif.24303

The filter capacitors will have to be double the size of what you'd use on a bridge rectifier because each diode is acting as a half wave rectifier. With 1000µF capacitors there will be about 2V of ripple if you draw 100mA of current.

Another disadvantage is if you load one rail significantly more than the other, you can cause your transformer to overheat because there will be a net DC current in the transformer which will cause core saturation.
 
Last edited:
thanks much, it was just a thought

i dont have the rectifiers in series; they're in parallel.

is it possible to run them in parallel like that and get full-wave rectification in both the negative and positive direction, or am I stuck with half wave rectification, half positive and half negative?


and about the transformer loading, do you mean in the proposed center-tapped transformer solution, you have to load both rails approximately evenly to not saturate the core?

so if I just wanted a negative rail but didnt use it often, i'd be better off with... two separate transformers, both referencing the same ground?
 
You can't get negative voltage using the method you describe.

You can't just connect the DC side of two bridge rectifiers in series like batteries. Try tracing the current path through the diodes and you'll se what I mean...


Hero is right. The only way you can get full-wave rectification in both the positive and negative supplies is if you have a transformer with two identical windings. Then you can tie the bottom of the top bridge to the top of the bottom bridge, and call the common point ground.
 
I think this is what you intended to draw. The colored current paths for positive and negative half cycles of the input show that the bridge shorts out the transformer.
 

Attachments

  • parallel bridges annotated.PNG
    parallel bridges annotated.PNG
    111 KB · Views: 148
Status
Not open for further replies.

Latest threads

New Articles From Microcontroller Tips

Back
Top