Two DC DC with one Regulator

[engr_joni_ee]

I am working on a new board design in which I only have +5 V available as input power supply with sufficient input power supply current. I also need to have -5 V on the same board because I need to power an operational amplifier which require both +5 V and -5 V in addition to some other chips that also requiring -5 V on the same board.

I am thinking of using a step up or a boost DC DC converter which can accept +5 V as input and convert to higher voltages. I found TMH-0512D which output +/- 12 V with relatively lower maximum current +/- 83 mA. If I use a negative linear regulator after the DC DC converter then I can get -5 V output with -12 V input.

Here is the link to DC DC converter TMH-0512D.

TMH 0512D | Traco Power

www.tracopower.com

I have roughly estimated the maximum current I need on -5 V and found that it would be between -120 mA to -150 mA. Even if I use the negative linear regulator with maximum output current -200 mA, I don't think it will be able to deliver between -120 mA to -150 mA with only one DC DC converter TMH-0512D because of limited output current +/- 83 mA.

Therefore, I am thinking how about using two DC DC converter TMH-0512D in parallel, each with maximum current +/- 83 mA and connect the output of both DC DC converter TMH-0512D to the input of a negative linear regulator that could then deliver between -120 mA to -150 mA.

I am not sure if practically it is possible or not. Kindly give your suggestions. Thanks in advance.

 

[barry]

I am working on a new board design in which I only have +5 V available as input power supply with sufficient input power supply current. I also need to have -5 V on the same board because I need to power an operational amplifier which require both +5 V and -5 V in addition to some other chips that also requiring -5 V on the same board.

I am thinking of using a step up or a boost DC DC converter which can accept +5 V as input and convert to higher voltages. I found TMH-0512D which output +/- 12 V with relatively lower maximum current +/- 83 mA. If I use a negative linear regulator after the DC DC converter then I can get -5 V output with -12 V input.

Here is the link to DC DC converter TMH-0512D.

TMH 0512D | Traco Power

www.tracopower.com

I have roughly estimated the maximum current I need on -5 V and found that it would be between -120 mA to -150 mA. Even if I use the negative linear regulator with maximum output current -200 mA, I don't think it will be able to deliver between -120 mA to -150 mA with only one DC DC converter TMH-0512D because of limited output current +/- 83 mA.

Therefore, I am thinking how about using two DC DC converter TMH-0512D in parallel, each with maximum current +/- 83 mA and connect the output of both DC DC converter TMH-0512D to the input of a negative linear regulator that could then deliver between -120 mA to -150 mA.

I am not sure if practically it is possible or not. Kindly give your suggestions. Thanks in advance.

I wouldn't do it, especially without knowing the internals of the DC-DC. Why don't you just find a suitable DC-DC instead of trying to make two of the wrong ones work? Just look, they're out there.
This would do it:
AM1DS-0505DH30LPZ

On second thought, why do you even need a dual output DCDC. Your idea seems like an attempt to use as many parts as possible instead of just one.

All you need is 5v in to -5v out. You could get a DCDC or switching regulator.

 

[FvM]

Why don't you just find a suitable DC-DC instead of trying to make two of the wrong ones work?

Yes.

Another consideration. It's cheaper and more efficient to use separate non-isolated boost and inverting switchers instead of isolated DC/DC module.

 

[cupoftea]

With vin = 5v and vout = -5v you could use a simple Cuk Converter. (Coupled or uncoupled). Just use a cheap current mode controller, and some simple circuit to feed back the neg divider voltage to the positive referenced controller.

 

[dick_freebird]

Seems to me that

1) dead-match between pos and neg split supplies is seldom expected and even if so, should never be assumed. Unless you have it in writing.

2) an off the shelf D -DC making split rails is almost always "cross-regulating" with one actually regulated and the rest, "as close as I*R lets it".

So

3) Why not just an inverting CMOS charge pump, if your current demand is modest?
Parallel a bunch and outphase them if you want more.

 

[KlausST]

Hi,

I don´t understand your idea:
* why using DUAL output converter?
* why using a 12V one ... while you need 5V

Klaus

 

[danadakk]

If parallel operation not called out in datasheet you would have to contact vendor to see
if they will do it. "Normally" one sees a sense pin to allow circuitry to control sharing
current.