Theory on Negative Voltages Please...

[reessi]

Hi there,

Last year I was trying to repair a laser machine that required +12v and -12v to control the servos for the mirrors. I did not fix the power supply but was trying to test it using an ATX power supply. The + side worked a bit but the negative did not, I tested the current draw through the meter on the +ve side and it was around 3 amp and so the -ve side could not pull enough since the ATX was only rated at 0.8 amp and tested at 1 amp through the meter.

Last night while building a bench power supply I had an idea and after some reading thought it should work.

From reading it is saying 0v does not exist, the multimeter is just measuring the potential difference. So I made a circuit (see pics). Which works when measuring the voltages. The +ve of the buck becomes the 0v in respect of the +12v and -12v.

Please advise if there is anything wrong with this circuit for generating a -ve rail!

If this is right I will make my bench power supply so that it can provide -ve voltages like this for future testing!

I can alter the values on the Boost and the Buck so it would equal 12v each. I just wrote down what it was set on at the moment.

I know you can also create a negative rail with a transformer with 3 taps as well.

Now how will the current draw be? Will it be shared if the voltages are equal so the currant of the boost split if I do not limit the current on the buck.

What would happen with the current draw if the voltages were set un-equally?

Can the buck be damaged if I only run something off the positive side in that circuit so the + of the buck is being used as the negative but is sinking the current???

Many thanks!

[stj]

Re: Theory on Negative Voltages Please...

0v does exist, it's your reference.
all voltages are simply relative to the reference.

as for your circuit, it probably will be a dramatic fireworks display.
most buck / boost convertors link the negative input and output.

[reessi]

Re: Theory on Negative Voltages Please...

Well after thinking about it a bit more I think it cannot work as the line I've labelled as 0v is not 0v it is +12v from the buck. All I have done is measured the difference between the terminals with the multimeter! And the -12 is not -12 it is just -12 in respect to the +12 because the probes are backwards!

But I am not really sure what negative voltages are? I tried to read about them... If anyone can explain better. I do not understand why it exists. Surly you either have voltage (+12) or you do not? Is the -12 travelling in the opposite direction? If so then why can't you just wire the plus and minus the opposite way around like to change the direction of a motor? And why are negative voltages used?



Cheers.

[reessi]

Re: Theory on Negative Voltages Please...

What about this temporary setup with 2 ATX Power Supplies as long as I dis-connect the earth wires to the mains otherwise there will be a dead short between the negative of one and the positive of the other! And I can check with meter before turning on that neither of the outputs are shorted!

I would like to make a proper PSU with negative but this could work to test my laser machine.

Can anyone advise of a method to create a negative from one ATX that is variable and at least 3 or 4 amps?

I am not finding much in google for this!

Cheers

[eccerr0r]

Re: Theory on Negative Voltages Please...

Forget power supplies and use batteries to think about them. As said, negative voltages are based on which your reference point is.

After understanding that, then go back to power supplies. Do worry that a lot of psus, especially computer/ATX PSUs, the case/ground is also a reference point and if you're not careful you'll short things out.

[momaka]

Re: Theory on Negative Voltages Please...

What about this temporary setup with 2 ATX Power Supplies as long as I dis-connect the earth wires to the mains otherwise there will be a dead short between the negative of one and the positive of the other! And I can check with meter before turning on that neither of the outputs are shorted!

That should work fine.
Indeed just make sure ground is disconnected on the "lower" PSU in that diagram.
Looks like you're getting there in terms of understanding negative voltages with that idea.

There is more than that to it, of course. But the main thing to understand here is that ground / 0V is just a relative term. One example I always give for that is to think of an apartment building with X floors above ground and Y floors below ground (garages?) If you're standing on the 3rd floor and your friend calls you from the lobby (ground), you're 3 floors above him. Likewise, if you're in the lobby and your friend is taking your car from -3rd floor below ground, you are still 3 floors above him (or you can say he is 3 floors below you, just the same as when you were on the 3rd floor and he was in the lobby.)

So what we call "ground" in circuits is really more of a return path for circuit currents and also a common point to which all voltages are reference from.

That being said, the reason why your first circuit with the boost converter probably -won't- work (or at least, it's not something I would recommend) is because what you labeled as "0V" on the output isn't actually a common return path that all circuit currents can/should go to. Why? Because on the buck converter, the points you labeled "0V" on its left side and "-12V" on its right side are one and the same connection. Therefore, the "0V" label on the right is more like 12V above the "true" circuit ground - i.e. the ground of the ATX PSU and boost converter. This creates a potential problem (pun totally intended here) in a case where your laser machine connects to a computer with a grounded PSU. What happens then is the laser machine's "0V" or supposed ground is not at the same potential as the computer that is connecting to the laser machine - the laser machine's ground is 12V higher than the computer's. So as soon as you plug in the PC to the laser machine (unless it's a laptop with a 2-prong power adapter), you will be shorting out your buck regulator's output. Also, that design is not a good idea because you generally shouldn't feed any current -into- the -output- of a buck regulator. For example, let's say the laser machine wants to use the difference between the 30V (18V) output and the supposed "0V" output from the buck regulator. Remember, you're feeding the buck regulator with 30V on the left side... and yet on the right side, you're feeding it with 30V again on it's output and somehow asking it to create a difference of 18V (by regulating to 12V above the true circuit's ground/0V reference.) If you understand how buck circuits work, then it should be clear why this won't work.

So while negative and positive voltages are indeed all about reference points, you also have to think of the current path and ask yourself, can this path provide sufficient "strength" for the return current. In the case of the boost + buck circuit in the first post, that's a clear NO. But in regards to the second circuit with the two ATX PSUs with the lower one "isolated" from true ground, that's a YES, because both of these PSUs are independently regulating their outputs to 12V.

[reessi]

Re: Theory on Negative Voltages Please...

Many thanks everybody! Especially "momaka" for the final explanation. This reply has really helped me understand what is going on and can experiment with out blowing anything up hopefully.

I will come back here if I have more query's on negative voltages...

Cheers