Hi chaps.
I'm designing an active xover and it needs to run on a single supply. I know I could use any op-amp like TL074 and apply a bias, but with 8 separate stages in total I would need 8 voltage references, else each stage would be coupled through the bias connections.
So I think it will be easier just to use a single supply capable op-amp. Can some of you recommend a cheap one suitable for an active xover? It doesn't need to be the highest quality in the world it is for a portable speaker.
I found the TL974. This says it is rail-to-rail and suitable for battery powered kit. However I have noticed that many op-amps say rail-to-rail and they only go within about 800mv. How about the TL947 will it really work with a single supply and ground referenced input signals?
I'm designing an active xover and it needs to run on a single supply. I know I could use any op-amp like TL074 and apply a bias, but with 8 separate stages in total I would need 8 voltage references, else each stage would be coupled through the bias connections.
So I think it will be easier just to use a single supply capable op-amp. Can some of you recommend a cheap one suitable for an active xover? It doesn't need to be the highest quality in the world it is for a portable speaker.
I found the TL974. This says it is rail-to-rail and suitable for battery powered kit. However I have noticed that many op-amps say rail-to-rail and they only go within about 800mv. How about the TL947 will it really work with a single supply and ground referenced input signals?
Use two 10k resistors in series between +ve and -ve, with a 100u decoupling from the centre point (0v point) and negative rail. That is your reference point for all inputs. As no current is flowing, no problem from instability and no regulation required.
Hi,
There is no such thing as a single rail op-amp. All op-amps
can be used single rail with proper standard single rail design.
rgds, sreten.
There is no such thing as a single rail op-amp. All op-amps
can be used single rail with proper standard single rail design.
rgds, sreten.
But if I fed a bias voltage from a single source to all non-inverting inputs on each op-amp it will basically be shorting the non-inverting inputs of each op-amp together. That isn't helpful when trying to design an active xover.
So anyway, can anyone say if the TL974 will work 'as is' on a single rail supply?
I wonder if I'm not fully understanding the way a biased op-amp would work in this application actually. I imagined it would need each non-inverting input pin biased. Do I in fact just need each connection that would be grounded to be connected to the mid-supply point, like this? In this case for the low pass filters obviously the caps would block any DC actually going to the op-amp input pins.
image upload no ads
I wonder if I'm not fully understanding the way a biased op-amp would work in this application actually. I imagined it would need each non-inverting input pin biased. Do I in fact just need each connection that would be grounded to be connected to the mid-supply point, like this? In this case for the low pass filters obviously the caps would block any DC actually going to the op-amp input pins.
An externally hosted image should be here but it was not working when we last tested it.
image upload no ads
Last edited:
a supply splitter "active ground" can be used just like the dual power supply with a "natural" ground - the filter circuits won't know the difference
From the little i know the CMR will suffer with a single rail. I think that's correct, if not someone will correct me.
I did a crossover recently and went split rail, testing with two PP3 batteries. Of course, using 4 stages and LME49740, the batteries didn't last long. But then they aren't low power devices, so it wasn't a surprise.
I did a crossover recently and went split rail, testing with two PP3 batteries. Of course, using 4 stages and LME49740, the batteries didn't last long. But then they aren't low power devices, so it wasn't a surprise.
Last edited:
As sreten said, you can use any op-amp. What you need is an active supply splitter, the 10k reference will not be low enough relative to the resistor values you are connecting to virtual earth.
Thanks, so is the diagram above the correct way to apply the bias? It's just that the source resistance of the bias point needs to be low to allow current to flow back.
Well to make my life easy I have opted to use a single-rail capable op-amp.
Applying bias from a resistive divider has the problems that: A) Source impedance is high. B) It feeds the power rails directly into the op-amp inputs so there is no PSRR whatsoever.
Using a virtual ground IC like the TLE2426 overcomes these issues but that adds cost and uses PCB space.
So.. for anyone looking here in the future I've found 3 op-amps that look good for single-rail audio use. The AD824, the NJM2747 and the LMC6484.
Applying bias from a resistive divider has the problems that: A) Source impedance is high. B) It feeds the power rails directly into the op-amp inputs so there is no PSRR whatsoever.
Using a virtual ground IC like the TLE2426 overcomes these issues but that adds cost and uses PCB space.
So.. for anyone looking here in the future I've found 3 op-amps that look good for single-rail audio use. The AD824, the NJM2747 and the LMC6484.
You are missing the point, the topology of the (low-pass) filter requires a ground (midpoint bias), be it virtual or real. Simply plugging in a rail-to-rail op-amp will no difference.
Last edited:
Ok I used terminology in an unclear way. Ground is where signals are referred to, and this circuit requires ground to be a midpoint bias.
Last edited:
Hey? I don't see why I need any bias if I have an op-amp capable of swinging to the negative rail (ground).
Hi,
Only one comment Tenson and I do not understand why nobody else saw.
The middle point of the divisor MUST be very well decoupled to GND.
Only in this way you will have a low impedance reference point.
And the input and output are usually connected by capacitors.
Only one comment Tenson and I do not understand why nobody else saw.
The middle point of the divisor MUST be very well decoupled to GND.
Only in this way you will have a low impedance reference point.
And the input and output are usually connected by capacitors.
Last edited:
Hi,
Don't build single rail op-amp circuits if you don't
have a clue how to do it properly, and just want
argue about how you think works, and is wrong.
rgds, sreten.
Don't build single rail op-amp circuits if you don't
have a clue how to do it properly, and just want
argue about how you think works, and is wrong.
rgds, sreten.
OK last try. The input signal will swing negative without a DC superimposition to raise it up off the lower rail (your ground).
- Status
- Not open for further replies.