Negative Voltage From Positive Supply

[syntax333]

Hi, I need negative voltage from positive supply with common ground.
I feed 18V to cheap LM2596 controllable voltage output module to get 12V and I feed 12V to voltage inverter circuit given at attachment with using LM2576-12.

However when I power up the circuit, my bench power supply goes constant current mode and drops its voltage to 3.4V and LM2576-12 starts to heat up.

I couldn't figure it why it does that. Is there any other way to invert positive voltage with common ground?

[Hiemal]

yeah running those unmodified won't work; the ground is common with the input ground so you're just shorting your module out.

You'd need to break ground continuity I think.

[ejeffrey]

He has connected the "GND" to the negative output.  This generally should work with a few caveats.  The first is that the input voltage seen by the buck module is actually the input to output voltage so you have to watch for input range.  The second is that getting the enable signal and power good signals right can be annoying since they are referenced to the chip ground which is the output.  The attached figure is from the LM2576 datasheet, and it should work correctly as drawn.

I do have a couple of questions:

Is there any reason you are sending the 12V regulated output to the inverter?  It would make more sense to me to have the two converters in parallel, one taking +18 to +12, the other taking +18 to -12.

Are you sure your power supply current limit is high enough?  The problem could be as simple as hitting the current limit on the 18V supply, causing it to enter current limit mode.  Switching regulators don't like this: they try to respond by drawing more current, which draws the voltage down further.  Even if you have the current limit set high enough for the steady state the turn on transient can trip the current limit and cause it to fail.  This is a good reason to implement under voltage lockout on switching regulators.  You don't really want them turning on until the input voltage is ready.

[syntax333]

Is there any reason you are sending the 12V regulated output to the inverter?  It would make more sense to me to have the two converters in parallel, one taking +18 to +12, the other taking +18 to -12.

You are right I will change that however, I build reference circuit on breadboard and still output was the same (can't get negative voltage with same reference) my power supply can give 1A.

I set current limit to max.

[Richard Crowley]

Most of those SMPS boards are NOT isolating. They have common ground from input to output.
SOME of them are isolating so that you can use them for voltage inversion.
So, either be double-sure that it is an isolating SMPS board, or else get a board that is designed for voltage inversion.

[Jwillis]

That is not a positive to negative voltage converter.That is just an adjustable step down buck converter.
If you want a negative voltage you need something like this https://www.ebay.com/itm/5V-6V-9V-10V-12V-15V-24V-Positive-Negative-Dual-DC-DC-Boost-Buck-Converter-/123562102784

[ejeffrey]

Most of those SMPS boards are NOT isolating. They have common ground from input to output.
SOME of them are isolating so that you can use them for voltage inversion.
So, either be double-sure that it is an isolating SMPS board, or else get a board that is designed for voltage inversion.

This is not a circuit for an isoalted converter, it is a buck module used as a non-isolated inverting buck-boost.  Many buck modules can be converted to inverters simply by relabeling the terminals exactly as shown in the OPs schematic -- which is caputred from the typical circuit section of the LM2576-12 datasheet.

Application notes on this technique:

https://www.maximintegrated.com/en/app-notes/index.mvp/id/3844
http://www.ti.com/lit/an/slyt286/slyt286.pdf
https://www.analog.com/media/en/reference-design-documentation/design-notes/dn1021fa.pdf

However, I still think the problem is that this is having trouble starting up due to the current limit on the input supply.  The datasheet says (section 8.1.9)

The switch currents in this buck-boost configuration are higher than in the standard buck-mode design, thus lowering the available output current.  Also, the start-up input current of the buck- boost converter is higher than the standard buck-mode regulator, and this may overload an input power source with a current limit less than 5 A. Using a delayed turn-on or an undervoltage lockout circuit (described in Negative Boost Regulator) would allow the input voltage to rise to a high enough level before the switcher would be allowed to turn on

So try connecting a beefier supply.  Either it will work great or it will release the magic smoke

[Buriedcode]

It would be quite hard to modify the module in the original post, as the tab for that device is GND, which would have to be lifted off the board for it to work.  Or use the modules ground plane as the negative voltage, and have a separate ground wire for IO caps, and input supply ground.

[ejeffrey]

It would be quite hard to modify the module in the original post, as the tab for that device is GND, which would have to be lifted off the board for it to work.  Or use the modules ground plane as the negative voltage, and have a separate ground wire for IO caps, and input supply ground.

Yes, that is what I was assuming the OP was doing, but now that I think about it, maybe not.  You have to connect your input supply "negative/ground" wire to the output, and take the output as ground.  If you are connecting the module ground to your circuit ground that will indeed short the output.