### Author Topic: Producing a negative voltage  (Read 528 times)

0 Members and 1 Guest are viewing this topic.

#### dcbrown73

• Regular Contributor
• Posts: 77
• Country:
##### Producing a negative voltage
« on: July 30, 2019, 05:00:56 pm »
I've been studying op amps and in some cases, you need to feed the op amp a negative voltage.

This produces a few questions.

I searched Youtube and I came across this BK Precision video showing the use of two voltage sources wired together to create a negative voltage.

After looking at what they did, why is it necessary to use two power sources for this?  If you just use one voltage source, but invert the +/- you can produce the same negative voltage.  Does this have something to do with current?

EDIT:  I see the video says produce a positive AND a negative.  Though, with two power supplies, why would you need to inner connect them rather than just stand alone.

Finally, if you have an op amp in a circuit you likely aren't going to have two distinct power sources.   Do you just create difference rails and then perform the same cross connects to produce a negative voltage to power the op amp?

Thanks,
Dave
« Last Edit: July 30, 2019, 05:02:36 pm by dcbrown73 »
Why exactly do people feel I should have read their post before I responded?  As if that was necessary for me to get my point across.

#### rstofer

• Super Contributor
• Posts: 6944
• Country:
##### Re: Producing a negative voltage
« Reply #1 on: July 30, 2019, 09:49:54 pm »
You should search for w2aew's videos on op amps.  Also, Dave has done some.

If you use 2 supplies to create a + and - voltage, you have 4 wires,  Where are you going to put them?  It's obvious what you do with the intended + output and the intended - output but what about the other 2.  That's where you tie them together and also create the circuit ground.  All signals will be + or - relative to ground which we will call 0V.

You can most definitely use a dual supply op amp on just one supply.  BUT, you have to have your AC signals revolve around Vcc / 2.  There are chips called 'rail splitters' that create this voltage or you can do it with resistors and probably an op amp as a voltage follower to lower the impedance.  This is done all the time but the Vcc / 2 thing is a PITA.  It just gets in the way if you want your input and output signals to revolve around 0V.  Yes, blocking capacitors help but it's pretty easy to get dual supply voltages so why bother?

While I'm on the subject:  "Op Amps For Everyone" is a great, and free, reference and they spend a lot of time talking about single supply op amps. Google for it...

Despite what I think about Vcc / 2, single supply op amps are the most commonly used at this point.  In production, nobody wants to create two voltages when they can design around it.

I can't get around it because my only application for op amps is analog computing and some functions just naturally go below 0 and it would be truly ugly math to offset all the signals by Vcc / 2.  It could be done but it isn't conventional and it wouldn't match what is covered in the text books.
« Last Edit: July 30, 2019, 10:10:28 pm by rstofer »

#### rstofer

• Super Contributor
• Posts: 6944
• Country:
##### Re: Producing a negative voltage
« Reply #2 on: July 30, 2019, 11:57:48 pm »
Another thing to consider:  With the two supplies wired in series with the center point taken as signal ground, there are probably no issues when you hook a scope probe to 'ground'.  Presumably the inputs and outputs are all ground referenced and tying them to earth ground through the scope won't matter.

When the reference is Vcc / 2, it is a matter of some consideration where you put the scope ground lead.  It needs to be thought thru carefully and it really depends on how the power supply is (or isn't) grounded.  The alternative is to put the scope ground lead on power supply ground and understand that everything you probe is offset by Vcc / 2.  All in, a PITA.

Dave has a "How Not To Blow Up Your Scope" video that is worth watching.  It isn't directed at the op amp issue but it does deal with paying attention to where the scope ground lead is connected because, unless there has been some 'floating scope' foolery, it will be at earth ground through the power cord.

#### Jwillis

• Frequent Contributor
• Posts: 768
• Country:
##### Re: Producing a negative voltage
« Reply #3 on: July 31, 2019, 03:57:59 am »
Think of the power supplies as 2 batteries connected in series. Positive negative positive negative. Where the to batteries are connected together become common and one end  of the series will be have a positive polarity relative to the centre common and the other end will have a negative polarity relative to the centre common.

#### Brutte

• Frequent Contributor
• Posts: 614
##### Re: Producing a negative voltage
« Reply #4 on: July 31, 2019, 09:52:32 am »
None of the opamps can drive output at and beyond the supply rails. There are chips that can get closer than other but ultimately you have to leave some headroom on both sides. Everything has to happen in the middle of the range. This can be easily accomplished if you create a midpoint and call it GND. Then you typically have +12V positive voltage and -12V negative voltage, with respect to that midpoint GND.

If you do not like this convention, you can call three terminals +12V, +24V and GND and midpoint will be +12V.
Or alternatively -12V, GND and -24V and midpoint will be -12V.

#### MrAl

• Frequent Contributor
• Posts: 481
##### Re: Producing a negative voltage
« Reply #5 on: July 31, 2019, 02:26:59 pm »
Not sure if this is what you are looking for, but if you look up a buck boost inverter circuit you'll see you can get a negative supply from a positive supply with a circuit.

Another idea if you just need a small voltage and small current is to try to run a good solar cell with a white LED.  The white LED is powered from the positive power supply and the solar cell is wired to create a small negative supply voltage.  It is good for biasing.

#### rstofer

• Super Contributor
• Posts: 6944
• Country:
##### Re: Producing a negative voltage
« Reply #6 on: July 31, 2019, 02:59:05 pm »
Older op amps (like the venerable 741) were usually operated off of +-15V supplies with signal swings in the +-10V range.  That left plenty of headroom.

Today, things are different.  We are using a 3.6V supply with signals swinging around 1.8V and probably limited to +-1.5V (maybe more, check the datasheet) from that midpoint.  This is about perfect for the 3.6V ADCs on many microcontrollers.

Here's a pretty good Wiki that describes many of the terms used to describe op amps:

https://en.wikipedia.org/wiki/Operational_amplifier

« Last Edit: July 31, 2019, 03:00:57 pm by rstofer »

Smf