I've done very well with digital electronics since getting involved over the past two years. One thing I have more or less ignored is analog electronics, other than simple issues like regulation, rectification, voltage division, etc. Now I want to look at op-amps. I basically just want to evaluate at a few op-amps that I have using an oscilloscope to compare input and output waveforms. I will be using the inverting configuration (e.g. http://upload.wikimedia.org/wikipedia/commons/thumb/4/41/Op-Amp_Inverting_Amplifier.svg/300px-Op-Amp_Inverting_Amplifier.svg.png) and plan on providing an sinusoidal test waveform from my function generator at first and then others as well. So my understanding is that I need to create a split power supply in order to accommodate positive and negative voltages. Is it an acceptable and smart way to do this simply to use two 9 volt batteries in series to get 0V/9V/18V and then reference the 9V on the series batteries to ground on the input and output signals to effectively get -9V/9V on the op-amp? Like this?
JoeN:
Is it an acceptable and smart way to do this simply to use two 9 volt batteries in series to get 0V/9V/18V and then reference the 9V on the series batteries to ground on the input and output signals to effectively get -9V/9V on the op-amp?
Sure.
Good. Don't want to fry my function generator or anything. I will whip this up today and see how a few opamps behave at various frequencies and with different waveforms. Thanks, George.
Yes, that's probably also the simplest way - opamps shouldn't need that much
current so PP3 sized 9V batteries will be OK.
The mains powered alternative is a centre-tapped transformer and full-wave
bridge rectifier, then a couple of matched regulators like 7815 and 7915.
Either way you avoid the switching noise of a SMPS which is usually important for
analog signals.
Also, the input needs to be referenced to ground. Any DC offset in your signal generator will get amplified along with the signal.
Digital is just a special case of analog. ;')
JoeN:
I've done very well with digital electronics since getting involved over the past two years. One thing I have more or less ignored is analog electronics, other than simple issues like regulation, rectification, voltage division, etc. Now I want to look at op-amps. I basically just want to evaluate at a few op-amps that I have using an oscilloscope to compare input and output waveforms. I will be using the inverting configuration (e.g. http://upload.wikimedia.org/wikipedia/commons/thumb/4/41/Op-Amp_Inverting_Amplifier.svg/300px-Op-Amp_Inverting_Amplifier.svg.png) and plan on providing an sinusoidal test waveform from my function generator at first and then others as well. So my understanding is that I need to create a split power supply in order to accommodate positive and negative voltages. Is it an acceptable and smart way to do this simply to use two 9 volt batteries in series to get 0V/9V/18V and then reference the 9V on the series batteries to ground on the input and output signals to effectively get -9V/9V on the op-amp? Like this?
That's a perfectly acceptable and commonly used circuit to power op-amps.
MarkT:
Yes, that's probably also the simplest way - opamps shouldn't need that much
current so PP3 sized 9V batteries will be OK.The mains powered alternative is a centre-tapped transformer and full-wave
bridge rectifier, then a couple of matched regulators like 7815 and 7915.
Either way you avoid the switching noise of a SMPS which is usually important for
analog signals.
I was thinking about this and I ran across this design too earlier. The issue I have is no center-tapped transformer on hand without starting to look at old junk and hoping that one is in there somewhere. Maybe I will have to add a few to the kit if I can find a cheap supplier.
A split supply can be made from a transformer without a center tap by using half wave rectification for the negative and positive half cycles.
One has to take into account the resulting DC voltage but when no other transformer is available it works fine.
UnoDueTre:
A split supply can be made from a transformer without a center tap by using half wave rectification for the negative and positive half cycles.
One has to take into account the resulting DC voltage but when no other transformer is available it works fine.
I tried to make a variation of the 2x9V split power supply by using a 24V AC/DC power source and the L7824 and L7812 voltage regulators to ensure +/- 12V power rails to my op amps (with the output of the L7812 becoming the virtual ground for the circuit). Strangely, the L7824s would only produce 21V despite 24V being available at the output; the L7812 gave exactly 12V as expected, being driven off the output of the L7824.
I then wired a 24V and 9V DC power supply in series to get 33V, to ensure that the L7824 had plenty of voltage to work with, and still it produced only 21 V. I bought 5 from DigiKey; all 5 did this.
My next thought is to use a 120V to 15VAC converter, and half-wave rectify both outputs as you described.
In the meantime, any ideas of why the L7824 is misbehaving? Thanks.
I don't think that scheme will work because the 7812 can
source current, but it cannot sink current. You should
use the 7812 &7912 as suggested.
This thread is 6 years old!!!!
Strangely, the L7824s would only produce 21V despite 24V being available at the output
Voltage regulators need extra voltage to operate. It's called "dropout voltage" and it's where the regulator drops-out of regulation.
The spec for the 78xx series says "2V typical" so I'd expect about 22V out with 24V in.
But, if you've already got 24V do you need to regulate it (or re-regulate it)?
...I've built a LOT of power supplies with transformers and linear regulators. When I start a project I usually start with the power supply even though I have a little homemade "bench power supply".
But when I used an op-amp in an automobile application I used something like [u]this[/u]. (The input & output are isolated so you can safely ground the +output to get a -12V supply.)
If I wanted something "quick and easy" for an experiment (and if I didn't have a bench supply) I'd go with batteries.
For my latest project (currently "on hold") I bought some switching power supplies similar to [u]this[/u]. The cost isn't much different than building from scratch, and it's "better".
And ever the "XY problem", it overlooks the matter that the function generator may (very likely) have been a single-supply device which did not produce a sinewave referenced to ground except via an output capacitor, so there was probably no need for a negative supply in the first place because a "virtual ground" would do the job perfectly well.
DVDdoug:
This thread is 6 years old!!!!
Voltage regulators need extra voltage to operate. It's called "dropout voltage" and it's where the regulator drops-out of regulation.The spec for the 78xx series says "2V typical" so I'd expect about 22V out with 24V in.
But, if you've already got 24V do you need to regulate it (or re-regulate it)?
...I've built a LOT of power supplies with transformers and linear regulators. When I start a project I usually start with the power supply even though I have a little homemade "bench power supply".
But when I used an op-amp in an automobile application I used something like [u]this[/u]. (The input & output are isolated so you can safely ground the +output to get a -12V supply.)
If I wanted something "quick and easy" for an experiment (and if I didn't have a bench supply) I'd go with batteries.
For my latest project (currently "on hold") I bought some switching power supplies similar to [u]this[/u]. The cost isn't much different than building from scratch, and it's "better".
-
Yes, I know the thread is old, but I thought because of the replies with split rail power supply solutions I'd add to it.
-
That's why I cobbled together with a 24V and 9V AC/DC adapter (in series) to get 33V, then pushed it into the L7824 but still got 21V out. And what's weird, the L7812, which was powered from the output of the L7824, produced a perfect 12V, not 12V minus 2V.
-
I suppose I don't have to re-regulate a 24V output from an AC/DC converter, but it just makes sense to me for the $1 in components it represents and minimal PCB real estate. Kind of belt and suspenders.
-
I've looked at DIY power supply kits and by the time everything is said and done, they're like $35...plus shipping for components, and then labor to wire it up. So I was thinking of getting a pile of AA batteries and wiring them so that I could get from 3V to 30V for testing input/output of op-amps, but 20 AA batteries cost $20 as well... I'm probably too much of a cheapskate to buy a $60 variable 30V power supply...but I think I might be shooting myself in the foot.
-
I'll look into what these Jameco things do and see if that's the answer.
DVDdoug:
But when I used an op-amp in an automobile application I used something like [u]this[/u]. (The input & output are isolated so you can safely ground the +output to get a -12V supply.)
I took a look at the data sheet; what kind of Black Magic does this device do? I can't be reading it right; it appears to say that if I push 9 to 18V into the DCW03A-12, I get +/- 12V out . How can this be?
https://www.jameco.com/Jameco/Products/ProdDS/161702.pdf
Thanks for your patience with my stupidity.
Can you not just use two 12V DC adapters? I agree about the XY problem evaluation. In the huge majority of cases, you can make an op amp circuit work single supply, by biasing it correctly. With as many as 4 amps in a package, you can also dedicate one amp to be a low impedance buffer for the 1/2Vcc bias voltage, if for some reason you need a "real" not a virtual ground.