1. Home
  2. Electronics Design
  3. -12v regulator problems (LM2576-12).

-12v regulator problems (LM2576-12).

Hello, I'm building a circuit that need to have both +12v and -12v. The circuit input is 24v DC. I'm using LM7812 to get the +12v and LM2576-12 for the -12v.

The circuit for the -12v that I have used could be seen at:

formatting link
Figure 10 (I have also implemented Figure 13 for softer start)

The only different from those figures and what I have implemented is the coil: I used 100uH instead of 68uH coil.

Now, I get nice stable -12v in output. However, when putting 2k ohm load between -12v output and GND the LM2576 gets really hot and eventually shouts down.

I also noticed that the -12v draws about 70mA from power supply when the actual load is 6mA (12v / 2kohm = 6mA).

What am I doing wrong? Please advice.

Regards, Edo Reshef

Reply to
edoreshef
Loading thread data ...

That was my first idea, and then I realize that 7912 needs to have negative voltage input. It doesn't take +24v and generates -12v. It takes -24v and regulates it to -12v. I wish I had -24v input, but I don't (only have +24v).

Reply to
edoreshef

No, not at all. The system I'm connecting to has +24V DC only.

Reply to
edoreshef

Dave, The inductor max rating is 0.7A. I think it should be enough. Isn't it?

Edo

Reply to
edoreshef

If all you need is a few mA of -12, you could maybe use one of the charge-pump inverters to flip the +12 to -12. They're cheap and simple to use.

John

Reply to
John Larkin

What's the current rating of the inductor that you used? If it's saturating at low currents, that would explain the problem.

--
Dave M
MasonDG44 at comcast dot net  (Just subsitute the appropriate characters in 
the address)

Never take a laxative and a sleeping pill at the same time!!
Reply to
DaveM

Have you paid close attention to the strict layout rules for these twitchy devices? I'm talking about the top half of figure 2.

The quiescent current graph shows the expected additional supply current to operate the chip for various input voltages and a 5 volt output. It should be close for your 12 volt output, and shows less than 10 mA additional supply current that is not accounted for by the output current.

Unless you need almost as much current as this chip is capable of delivering, you will have a lot bigger inductor than you really need. This is because the current limit is not adjustable and can be as high as 6 amperes. For that current limit to work right, you need an inductor rated for at least that much current without saturating, and your supply and bypass capacitors have to be able to deliver pulses of current this large without bouncing around more than other loads can stand.

How much load current must your -12 volt supply deliver?

Reply to
John Popelish

John,

I have implemented as seen in figure 2.

I'm not sure about that. In the -12v circuit at figure 10 it says "-12v @0.7A" output. It also says that during power up it may take 5A. so what should it be?

0.7A or 5A inductor?

I need something like 120mA. I'm not sure if charge pump can supply those values.

Regards, Edo

Reply to
edoreshef

John, that will not do. I need 120mA.

Edo

Reply to
edoreshef

I'm very sure. I actually measured that there are 6mA going throw that resistor... The resistor, does get warm at all (6mA is practically nothing) but the Vin (+24v) draws 70mA. The LM2576-12 gets really hot! It seams that it absorbs most of the power.

I'm going to try 3A inductor, and see what comes up.

Edo

Reply to
edoreshef

Why don't you simply use a 7912 for the negative volts and design the psu accordingly ? Simple and reliable. Been using 79XXs for > 30 yrs and going strong !

Graham

Reply to
Pooh Bear

Which is why - since you said you were 'building a circuit' - that you provide a negative supply.

Is there no negative supply available to you ?

Graham

Reply to
Pooh Bear

In which case you have limited means of providing negative volts.

If you need only a few / maybe tens of mA then a cap based device may fit the bill. Jeez I used these a few times and forget who made them. I expect Maxim will do them now.

For more serious needs you need a buck / chuck whatever it's called that uses magnetics.

Graham

Reply to
Pooh Bear

Not that simple.

Inductors used in such an application have to be purpose designed.

You have to consider core losses etc ... etc... Not to mention the material properties.

Graham

Reply to
Pooh Bear

A capacitive charge pump will have serious trouble at that value of current.

You should really design for a split rail supply.

Graham

Reply to
Pooh Bear

This chip is in serious trouble with a .7 amp inductor. I can't say that this is involved in your heat problem, but any later overload will give the chip fits.

If you had a control chip with appropriate current limit, you should be able to use an inductor rated for about 4 times your highest required output current. That is just based on average output current from a ramp that goes from 4x to zero in half the cycle time. If the current is continuous (ramp does not reach zero during the output part of the cycle) than the peak current will be somewhat less.

Do you have a scope to look at the stability of the 24 volt supply input to the chip? It takes pretty good quality (low ESR) capacitors on both sides of this kind of circuit to keep the ringing under control.

Reply to
John Popelish

Dunno what you are doing wrong, but I finished a product recently with this exact configuration (minus the delayed start circuit). Worked like a charm generating -12V from +12 with a few hundred mA output. I used a chunky SMT power inductor, e-mail me if you want the part number and manufacturer.

One gotcha with the basic circuit (mentioned on the datasheet) is the very high startup current spike (several amps), but that's probably not related to your problem. You might want to test without the fig.

13 stuff just to see if it's causing problems.

BTW, you probably should be using the HV version. You are very, very close to the maximum voltage rating with a 24V input-- 24+12 = 36V. You need good low-impedance capacitors CLOSE TO THE CHIP especially for Cin and also Cout. Do NOT attempt to rely on capacitors in an external supply.

Best regards, Spehro Pefhany

--
"it\'s the network..."                          "The Journey is the reward"
speff@interlog.com             Info for manufacturers: http://www.trexon.com
Embedded software/hardware/analog  Info for designers:  http://www.speff.com
Reply to
Spehro Pefhany

You have 36V across the IC that way, use the conventional bipolar switcher: View in a fixed-width font such as Courier.

. . . . >----------+--------------------+ . | | . | +-------------VIN-+ . unreg === | | . | | | . >----------+-----GND Buck PWM | . | | | . | | | . | +---FB--------SW--+ . | | | . | | | . | | | . +---/\\/\\--+--/\\/\\----|-----\\ . | | | . | SD | | . +---|>|--------------+ | . | | | . SD | C | | . +------|>|-----|-------+------||----+ | . | | | | | . | | | | | . | | +-------+ | | . | | |o |o | . | | ) || ) | . | | ) || ) | . | | LB ) || ) LA | . | | ) || ) | . | | | | | . | +---------------+ | | . | | | | . | C | C | | . +------||------+--------||----------+-----/ . | | | . | | | . Vn GND Vp . (=-Vp) . . . . .

Reply to
Fred Bloggs

** You real sure that 2kohm is not really a 200 ohm ??

Does it get warm ??

.......... Phil

Reply to
Phil Allison

the strict layout rules for these

half of figure 2.

you need almost as much current as this chip is capable of

will have a lot bigger inductor than you really need.

current limit is not adjustable and can be as

sure about that. In the -12v circuit at figure 10 it says "-12v

be?

or 5A inductor?

This chip is in serious trouble with a .7 amp inductor. I can't say that this is involved in your heat problem, but any later overload will give the chip fits.

If you had a control chip with appropriate current limit, you should be able to use an inductor rated for about 4 times your highest required output current. That is just based on average output current

from a ramp that goes from 4x to zero in half the cycle time. If the current is continuous (ramp does not reach zero during the output part

of the cycle) than the peak current will be somewhat less.

Do you have a scope to look at the stability of the 24 volt supply input to the chip? It takes pretty good quality (low ESR) capacitors on both sides of this kind of circuit to keep the ringing under control.

Reply to
John Popelish

ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.