Single supply opamps and negative clipping
- Thread starter Ethan
- Start date
Help Support GroupDIY Audio Forum:
Yes. And that would not be a nice half-wave rectification, because the TL07X have a notion of inverting polarity when their inputs are out of range. So you may hace some kind of dual-wave rectification, with a lot of dirt.
Svart
Well-known member
it's going to depend on how you hook the inputs up really.
For a unity stage, AC coupling both the input and output would essentially get you working in this case since the opamp will have it's own virtual ground.
For a gain stage, it's essentially the same as normal only instead of referencing your shunt resistor of the divider to ground, you reference it to VCC/2.
But to answer your question directly, yes you will negatively clip the signal if all of your divider network is referenced to ground on a single supply system. similarly if you referenced it to VCC you would clip the positive side of the signal.
EDIT:
http://instruct1.cit.cornell.edu/courses/bionb440/datasheets/SingleSupply.pdf
and a single supply opamp is generally the same but they are tested for being used on a single supply. Generally they also have optimized internal biasing circuitry.
PS: when using a single supply and VCC/2, you run into problems with a dual opamp's ability to retain good PSRR. You have to make VERY sure that your VCC/2 does not move. The decoupling must be very good in this case. Use ferrites, too.
For a unity stage, AC coupling both the input and output would essentially get you working in this case since the opamp will have it's own virtual ground.
For a gain stage, it's essentially the same as normal only instead of referencing your shunt resistor of the divider to ground, you reference it to VCC/2.
But to answer your question directly, yes you will negatively clip the signal if all of your divider network is referenced to ground on a single supply system. similarly if you referenced it to VCC you would clip the positive side of the signal.
EDIT:
http://instruct1.cit.cornell.edu/courses/bionb440/datasheets/SingleSupply.pdf
and a single supply opamp is generally the same but they are tested for being used on a single supply. Generally they also have optimized internal biasing circuitry.
PS: when using a single supply and VCC/2, you run into problems with a dual opamp's ability to retain good PSRR. You have to make VERY sure that your VCC/2 does not move. The decoupling must be very good in this case. Use ferrites, too.
Svart
Well-known member
I have an interesting story about dual supply opamp that I used in a single supply setup.
I was working on a PLL/VCO design and needed to DC multiply the PLL's 0-5v output to match the VCO's 0-20v tuning range. I was also incorporating a second order loop filter into the opamp stage as well but that's not important right now.
Anyway, most designs call for an offset split rail design, something like -3V and +25V so that an opamp without rail-rail capability can be used. As an aside, most precision opamps have limited slewing ability and can't swing rail to rail either..
So we tried that at first and had trouble with the PSRR between the two rails. while we could have spent some time and money fixing it, one of our Analog Devices FAEs came to talk about some other product but happened to pull out some literature from his briefcase while he was looking for something else. I noticed it said "precision rail to rail 36V". I grabbed the note and took a look.
The conversation turned towards this part and we ordered samples.
We tried it, loved it and were able to run 0-20v within .2V of the rails with high linearity, which was surprising. In order to get the PSRR I wanted, I used tight layout, 0402 1uf and 10uf low esr tants right on VCC and on VCC/2. I used ferrites on all supplies and all regulation was done locally under one can.
It's super quiet and very reliable/repeatable. All with a dual supply opamp used on single supply.
I was working on a PLL/VCO design and needed to DC multiply the PLL's 0-5v output to match the VCO's 0-20v tuning range. I was also incorporating a second order loop filter into the opamp stage as well but that's not important right now.
Anyway, most designs call for an offset split rail design, something like -3V and +25V so that an opamp without rail-rail capability can be used. As an aside, most precision opamps have limited slewing ability and can't swing rail to rail either..
So we tried that at first and had trouble with the PSRR between the two rails. while we could have spent some time and money fixing it, one of our Analog Devices FAEs came to talk about some other product but happened to pull out some literature from his briefcase while he was looking for something else. I noticed it said "precision rail to rail 36V". I grabbed the note and took a look.
The conversation turned towards this part and we ordered samples.
We tried it, loved it and were able to run 0-20v within .2V of the rails with high linearity, which was surprising. In order to get the PSRR I wanted, I used tight layout, 0402 1uf and 10uf low esr tants right on VCC and on VCC/2. I used ferrites on all supplies and all regulation was done locally under one can.
It's super quiet and very reliable/repeatable. All with a dual supply opamp used on single supply.
![Soldering Iron Kit, 120W LED Digital Advanced Solder Iron Soldering Gun kit, 110V Welding Tools, Smart Temperature Control [356℉-932℉], Extra 5pcs Tips, Auto Sleep, Temp Calibration, Orange](https://m.media-amazon.com/images/I/51sFKu9SdeL._SL500_.jpg)
> Would a "single supply opamp" ....handle that clipping better
No.
There is NO basic difference.
"Single supply part" may pull nearly all the way to one rail. LM324 will go real close to the negative lead. And its inputs may swing a little negative of negative and still stay sane. But if you have audio biased at the negative rail (ground), you still only get half of it. (A TL071 might give you only 45% of the wave because it won't swing to the rail.)
What the chip-writers seem to mean by "single supply part" is that you can do DC systems. Say you need a 0V-10V sweep. A conventional chip may not swing within 3V of the rail. So you need a negative 3V rail to allow 0 V output. But a "single supply part" may swing to 0.01V of the rail, and you may decide that 0.01V is zero-enough for your needs. (If this was a 1KHz/Volt VCO, it would sweep 10Hz to 10KHz, which would find all buzzes/rattles in a speaker or room; you don't have zero Hz rattles and you'd rarely excite a 5Hz rattle.)
You want to stick your audio half-way between the rails. If your rails are 0V and +9V, you want to bias up at 4V to 5V. Then you can swing both ways.
No.
There is NO basic difference.
"Single supply part" may pull nearly all the way to one rail. LM324 will go real close to the negative lead. And its inputs may swing a little negative of negative and still stay sane. But if you have audio biased at the negative rail (ground), you still only get half of it. (A TL071 might give you only 45% of the wave because it won't swing to the rail.)
What the chip-writers seem to mean by "single supply part" is that you can do DC systems. Say you need a 0V-10V sweep. A conventional chip may not swing within 3V of the rail. So you need a negative 3V rail to allow 0 V output. But a "single supply part" may swing to 0.01V of the rail, and you may decide that 0.01V is zero-enough for your needs. (If this was a 1KHz/Volt VCO, it would sweep 10Hz to 10KHz, which would find all buzzes/rattles in a speaker or room; you don't have zero Hz rattles and you'd rarely excite a 5Hz rattle.)
You want to stick your audio half-way between the rails. If your rails are 0V and +9V, you want to bias up at 4V to 5V. Then you can swing both ways.
