How to generate a slightly negative voltage?

Go To Last Post
18 posts / 0 new
Author
Message
#1
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Howdy - I need to generate a supply that is just slightly negative. Like I only need -.25V or so. A bit more would be fine - but not much. Certainly no more than -1V. It will be sinking about 15ma. I have a 6-8V input on this board. I already have a Vin to 3.3V buck converter and a Vin to 10V boost converter on this board.

I'm thinking I might replace the inductor on the buck with a coupled inductor, and have one side of the secondary grounded, and have the other side have a diode in series with a grounded capacitor. Line up the dots so that the diode is on when the buck converter switch is on. That way I'd only have (Vin - 3.3V - 0.5V (diode)) or about 2.2-4.2V on the output. Still more than I want though.

Anybody got a better idea?

Thanks!

edit: supply needs to be fairly clean - but I can put a big fat LC filter in front of it to take care of that. The supply voltage can fluctuate - it's going to be powering an op-amp that I want to operate to the rail. My Op-amp is strange enough that I don't think I'll be able to find one with similar specs that can operate at the very rail (it can output about 15mV from the rail - but I need to get within a few uV of the rail)

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Look at the LTC1043 data sheet at linear tech web site and then scroll down and see their example:
Single 5V Supply, Ultra Precision
Instrumentation Amplifier

You not only get one microvolt or so accuracy but near zero offset and no offset adjusment, etc.

Of course it depends on the range of input you have to handle. If the above is not what you want then you might see if you have an rs232 driver onboard that will allow you to load the negative supply enough for what you need. 15 ma seems like a lot for an opamp.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Use a timer to generate a square wave on an output pin and generate the negative voltage with a couple of diodes and capacitors.

Leon Heller G1HSM

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

What about:

 +------------ VCC
_|_
 -
 |
 |
 +----|<|----- GND
 |
 | 
 +------------ V-
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

jayjay1974 wrote:
What about:

 +------------ VCC
_|_
 -
 |
 |
 +----|<|----- GND
 |
 | 
 +------------ V-


That's a really cool idea! I really hate artificial grounds - but if I can't come up with anything better I think I might go with that. Thanks!

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

leon_heller wrote:
Use a timer to generate a square wave on an output pin and generate the negative voltage with a couple of diodes and capacitors.

I'm hoping to find something stand alone.

And I think I did: http://www.national.com/ds/LM/LM...

This part looks like pretty much exactly what I need.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Nice find (LM7705)! It says 4mv noise. I probably would filter that a little more.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

nleachim wrote:
I really hate artificial grounds

Every ground is artificial :)

You can also use some 0.1 ohm or 1 ohm resistor if diode drop is too much.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Just to note that if the circuit has a MAX232 for UART use then there's a -12V pin that can be dropped to any -ve voltage you choose.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Quote:
I'm thinking I might replace the inductor on the buck with a coupled inductor, and have one side of the secondary grounded, and have the other side have a diode in series with a grounded capacitor. Line up the dots so that the diode is on when the buck converter switch is on. That way I'd only have (Vin - 3.3V - 0.5V (diode)) or about 2.2-4.2V on the output. Still more than I want though.

This will work if you do the following:

You need to arrange the coupled inductor/transformer so that the secondary inductor conducts when the switch is "OFF". This way ONLY, is when you have a fixed voltage across the main inductor of 3.3+VD

Now you have a choice, you can have equal windings and use a series resistor and schottky diode to produce about 0.235V (schottky VF) or alter the secondary turns ratio for lower output, noting that in either case the output is not tightly regulated and probably wont work well unless the 3.3V section is in continuous current mode.

I use this aproach to provide a regulated 5V rail and two floating 12V rails for H-Bridge high side bias supplies and it works great.

The charge pump methods work well, and if you use the same resistor/schottky or other diode to limit the -ve voltage will be cheaper to implement.

Another method is to capacively couple an LC filter from the switch node to 0V with a schottky diode connected from the junction of the cap and inductor to 0V, cathode to 0V. The cap can be from 1-100nF and will limit power transfer. Use a series resistor+diode etc to set your -ve output level.

Ron.

 

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

For something standalone take Leons idea and use a 555

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

For something standalone take Leon's idea and use a AVR Tiny.

JC

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Yes, I know, 200 mA drive beats 20 mA drive...
But a small FET beats tweaking R's and C's ...

JC

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

clawson wrote:
Just to note that if the circuit has a MAX232 for UART use then there's a -12V pin that can be dropped to any -ve voltage you choose.

I suggested that in my earlier post as well.

I like that National Semi part the OP found himself. It has regulation built in and supplies just enough for the negative rail of an opamp in order to swing to gnd.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Normally you would use an inverting amplifier for this. The negative feedback would keep the reference in check, and the amplifier would show very low impedance. I am not sure if those setups can actually sink current though. I would use these as references normally.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

UNiXWHoRe wrote:
Normally you would use an inverting amplifier for this. The negative feedback would keep the reference in check, and the amplifier would show very low impedance. I am not sure if those setups can actually sink current though. I would use these as references normally.

You need a negative rail in order to make use of an inverting amp.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

True. How about a charge pump? Like someone suggested before, a Max232 will work, or there are smaller chips called "switched capacitor voltage inverters" available that will take in a +5v and output a -5v from it, with about 30mA current available.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

UNiXWHoRe wrote:
True. How about a charge pump? Like someone suggested before, a Max232 will work, or there are smaller chips called "switched capacitor voltage inverters" available that will take in a +5v and output a -5v from it, with about 30mA current available.

The OP found the LM7705 at National Semiconductor and posted about it above. It does what he needs and has regulation in it as well.