Question about Dual Polarity Power Supp. (Ashamed to ask)

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Hello,

I really am ashamed to ask this. But you have to realize I was taught science in America which from the K-12 education is full of wrong information.

Also keep in mind I'm only 19 and have no college education on this stuff (yet).

And trust me, I have searched on a good explanation.

Basically I want to know, what the difference between the negative and positive rails are if any. I know that an Op Amp uses both, but why exactly? I had one publication actually explain why it uses a Dual Polarity Power Supply as "Because it keeps it happy", didn't help to say the least.

I know the basics about electricity, and I've read Gainclone FAQs and building experiences for about 2 months now. But I don't want to build anything until I know exactly what's goin on.

When I see a schematic with say, a 22v + and a 22v - rail connecting to the chip, what should I be visualizing exactly?

Why is it exactky that both rails aren't 22v + ? I think what's really confusing me (and I'm sure you're all laughing as it's probably simple) is what the heck is a negative rail to begin with?

Do negative rails act the same way as positive (Should I just envision 22v +/- as 22 v a/b?) or do they do their own thing?

Or do they act like battery terminals?

I'm sorry I put you all to sleep, and I know it's pathetic that I could understand most things about amps but this one confuses me, but any help would be appreciated as I can't find any text explanations.
 
TaaJ

Voltage is always reference to Ground and ground is 0 volts. That mean ground in theory has no potential it neither positive nor negative it's 0, zero nada, nothing. However when a voltage measurement is made it is made with respect to ground with you meter. Of course + 22 vdc means it's 22vdc more positive than zero. The opposite is true for negative voltage, -22vdc is 22 vdc more " negative" than 0 volts or ground. Also, the difference between +22 vdc and -22 vdc is 44 volts.

I sure there a lot of info on basic electronics on the web, check it out., it’s very simple. Also, you should know this semiconductors technology, the transistor was developed in the USA, the Micro Processor and the IC were all developed by Texas Instruments in the US. So some of are schools must be ok.

:)
 
Kelly McDonald said:
The one thing you need to realise is that voltages are all relative to one another, for example a -22v/0v/+22v power supply isn't really any different from a 0v/+22v/+44v supply, it's all about how you measure it. Now typically we use ground as a reference point and measure from there.
I realize it's all relative to earth reference.

What's confusing me is basically, is the -22v doing the same thing as the +22. Their both voltage, both "pushing" current at the same amount (theoritically) correct?

The + rail is rectified from the positive ac (speaking from sine waves)

And the - rail is just power rectified from the negative ac. Right?
 
Instead of "voltage" think in terms of "potential difference". We take "earth" or "ground" as the universal zero point, then any other voltage can be higher or lower; if you put two 9V batteries in series and connected the middle junction to earth, the extreme ends would be +9V and -9V with respect to earth.

If you use those two batteries-in-series, centre grounded, to power the op amp, and input a music signal (which alternates positive and negative about its own grounded screen lead) the op amp can amplify both positive and negative halves of the signal. If instead one side of the op amp's supply was at ground potential (0V) then only the positive half of the signal would be amplified, and at worst the negative input would damage the IC.

There are ways of running the op amp on a single-rail supply, creating a virtual earth at half the supply voltage, but I'll leave that for now...

With your transformer supply, you need a transformer with either a centre-tap on the secondary, or two secondaries, in either case with the centre-point(s) grounded to give the zero reference point.
 
The thing to remember, is that you want to push the speaker equally in opposite directions. Since one speaker terminal is typically connected to ground, you need the output to swing both positive and negative, relative to ground. In the typical situation, positive pushes the speaker cone out and negative pushes it back.

Sheldon
 
Sheldon said:
The thing to remember, is that you want to push the speaker equally in opposite directions. Since one speaker terminal is typically connected to ground, you need the output to swing both positive and negative, relative to ground. In the typical situation, positive pushes the speaker cone out and negative pushes it back.

Sheldon
Maybe I should be reading up alot more on how the music signals work and not beat myself up on this.

I did know that speakers alternate negative and positive. I wans't aware so much about cpemma talked about negative and positive signal halves (I'm assuming the it would look like a jagged sine wave?)



So say if the amp has a 44volt potential difference, it could potentially use 22 volts to amplify the negative signal or the positive signal.

Or to put it in consumer terms for watt hungry shoppers, we could say 56 watts (making that number up) could be used on either end (+/-) of the signal?
 
I usually stay away from the electronic forums, so forgive me if you do not like this answer. It may be because I do not understand the question.

If the question is why we use symetrical power supplies + and - rails instead of just plus or minus, it is because when no signal is applied, with symetrical supplies we have no dc offset at the output.

For a transitorized amplifier this would mean no output capacitor or transformer required.

I hope this is responsive and helps. If not, then please forgive the intrusion.

Mark
 
0V is just a reference. you could say the opamp is powered by +200V and +156V. thats a 44V difference on the opamp... it's easier to say 44V or +-22V.

having 2 supplies both referenced to ground (+20V higher then ground, and -20V lower then ground) allows the signal to change up to +20V or -20V with respect to a known voltage.

this is where "ground loops" can come in -- you assume ground is at 0v, but becuase the wire running back to the PSU is shared, well, you get a voltage at "ground" on the PCB... ground isn't ground...

rails don't need to be symetric either, but it usually makes the most sense to do so. I've had a reason for +30, -5 before, but it wasn't for an audio amp.
 
TaaJ said:
I did know that speakers alternate negative and positive. I wans't aware so much about cpemma talked about negative and positive signal halves (I'm assuming the it would look like a jagged sine wave?)

So say if the amp has a 44volt potential difference, it could potentially use 22 volts to amplify the negative signal or the positive signal.
May be easier to follow looking at the LM1875 datasheet, which has two very similar examples, a split-rail version on p1, and the single supply equivalent on p2.

Split-rail, the input capacitor is often optional, so the input signal can come straight in to pin1, get amplified and go straight to the speaker from pin4. Amplifier voltage gain is 21, so +1V input (which could be the peak of an input sine wave) is amplified to +21V , -1V input (trough of the same wave) is amplified to -21V. Providing the power rails are a bit greater than +/-21V, say +/-25V, you've a working amplifier.

Single-rail, you have to add a few more parts. The two equal resistors R1 & R2 split the supply voltage down the middle, and add half-supply voltage to the signal at pin1. Taking 50V as the supply, a +1/-1V sine wave input becomes a +26V/+24V signal and is amplified to +47V/+3V. (The reference point for this amplification is 25V, often called a "virtual earth"). Still within the supply rails, so no problem. But before sending the amplified signal on to the loudspeaker from pin4, you need to remove the 25V DC offset, which will push the speaker cone hard in one direction even with zero volts signal input, so the large capacitor C6 is added as a DC-blocking filter.

Ignoring any sound quality issues, the single-rail system needs more parts, more expense. And for the same power amplifier, touching a supply wire gives twice the shock. :hot:
 
cpemma said:

May be easier to follow looking at the LM1875 datasheet, which has two very similar examples, a split-rail version on p1, and the single supply equivalent on p2.

Split-rail, the input capacitor is often optional, so the input signal can come straight in to pin1, get amplified and go straight to the speaker from pin4. Amplifier voltage gain is 21, so +1V input (which could be the peak of an input sine wave) is amplified to +21V , -1V input (trough of the same wave) is amplified to -21V. Providing the power rails are a bit greater than +/-21V, say +/-25V, you've a working amplifier.

Single-rail, you have to add a few more parts. The two equal resistors R1 & R2 split the supply voltage down the middle, and add half-supply voltage to the signal at pin1. Taking 50V as the supply, a +1/-1V sine wave input becomes a +26V/+24V signal and is amplified to +47V/+3V. (The reference point for this amplification is 25V, often called a "virtual earth"). Still within the supply rails, so no problem. But before sending the amplified signal on to the loudspeaker from pin4, you need to remove the 25V DC offset, which will push the speaker cone hard in one direction even with zero volts signal input, so the large capacitor C6 is added as a DC-blocking filter.

Ignoring any sound quality issues, the single-rail system needs more parts, more expense. And for the same power amplifier, touching a supply wire gives twice the shock. :hot:


Great post for me, makes alot of sense.

Thanks everyone!
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.