Ground sensing single supply opamp recommendation

[VEGETA]

I think LM358 and LM324 are a good start. However, if you want modern and better parts while still direct pin-to-pin replacements... then you might look for LT1013 and LT1014 which are dual and quad op-amps respectively.

https://www.analog.com/en/products/lt1013.html
https://www.analog.com/en/products/lt1014.html

They are not so cheap thought but not expensive. LT1013 is about 2$ from Mouser, while LT1014 is about 5$.

these options are a good start, but if they work then you need to optimize for cost and so on... you won't need that if it is a hobby or personal project.

[Zero999]

@TimFox:
Albeit quite expensive, the CA3130A has been added to my shopping list. You can't have everything, I guess, so let's see how it turns out. Otherwise I will need to reconsider going dual rail. Thanks for that recommendation.
At least it may be helpful in practical understanding some of the other comments in this thread about specs and the limits of it for an opamp.

@blueskull:
I do not need to source 40mA, I would be happy with 20. And probably could even go down to 15mA. But that's about the limit. The 15V are set in stone. The 40 mA output current of the LM324 just gave an extremely confident headroom. My assumption: The larger the headroom, the higher the chances to actually meeting the spec. Sorry, if this lead to any misconception.

As I am looking for a single supply buffer, that works from ground up, I do not need to sink any current and the gain is (ideally) one. Now while I do accept, that there may be cases, where a voltage follower may need to sink current, I did not sink about this scenario. Think.

However, with the on following discussion, as far as I have been able to follow it, it was rather helpful for my understanding, how any why sinking current affects the specs. Especially in term of reading a datasheet. Or at least I hope so. Practice will show

So thanks all again for your time and the explanations

Do you want to source or sink current?

Sourcing current is easy, because the lower output transistor can be off. You also won't be sourcing the full current, at the lowest output voltage, unless you're using it to drive an extremely low load impedance. 15mA, with an output voltage of 5mV, would be a load impedance of just 300mOhm. If it has a class A stage, such as the CA3130A, it will be gauranteed to output 0V, as long as the output stage is not sinking any current, as no current will be flowing through the output stage.

Sinking current is more tricky, because the lower transistor in the output stage needs to have a low enough resistance, to get below the required on voltage. As above, sinking 15mA, with an output voltage of 5mA, would require an on resistance of 300mOhm and no small op-amp IC can do that, because the output transistors will be too small. You might be able to use a ridiculously over-rated power amplifier IC, designed for Amps, but it's likely to have a poorer offset voltage. A discrete output stage, as mentioned above, could be added, but it would be very challenging to stabilse and avoid oscillation.

[TimFox]

See Figure 17 of the CA3130 data sheet  https://www.digchip.com/datasheets/parts/datasheet/235/CA3130-pdf.php  for a circuit to boost the output current by adding external CMOS inverters alongside the internal output stage.  (Use unbuffered inverters for this circuit.)  I remember that this trick was used in General Radio DigiBridges for the current-sense amplifier.

[Zero999]

See Figure 17 of the CA3130 data sheet  https://www.digchip.com/datasheets/parts/datasheet/235/CA3130-pdf.php  for a circuit to boost the output current by adding external CMOS inverters alongside the internal output stage.  (Use unbuffered inverters for this circuit.)  I remember that this trick was used in General Radio DigiBridges for the current-sense amplifier.

There isn't any frequency compensation, so I'd be surprised if it's stable. The high gain of 255 will help somewhat, but be prepared for fun and games with oscillation.

[TimFox]

See Figure 17 of the CA3130 data sheet  https://www.digchip.com/datasheets/parts/datasheet/235/CA3130-pdf.php  for a circuit to boost the output current by adding external CMOS inverters alongside the internal output stage.  (Use unbuffered inverters for this circuit.)  I remember that this trick was used in General Radio DigiBridges for the current-sense amplifier.

There isn't any frequency compensation, so I'd be surprised if it's stable. The high gain of 255 will help somewhat, but be prepared for fun and games with oscillation.

To clarify, I was referring to the op amp configuration, not the entire high-gain amplifier circuit.  The CA3130 is not inherently unity-gain stable, and requires a compensation capacitor from pin 1 to pin 8 (the input to the CMOS inverter output stage) at unity gain.  The output stage is a 4000-series inverter, which is a transconductance near the middle of the output voltage range, crossing over to an ON resistance near either supply pin.  The circuit in Fig 17 adds three additional inverters (from a CMOS array) in parallel with the two output FETs (pin 8 to output at pin 6) to increase the transconductance and output current capability (not after the op amp, as often done with buffers and conventional op amps).  The actual circuit in Fig 17 is for a higher gain inverting amplifier.  I believe that an unbuffered 4009 CMOS hex inverter could be used, but I haven't tried it.  The output stage is interesting, since the voltage gain depends on the load resistance (high-Z output in "active" region).

[Zero999]

See Figure 17 of the CA3130 data sheet  https://www.digchip.com/datasheets/parts/datasheet/235/CA3130-pdf.php  for a circuit to boost the output current by adding external CMOS inverters alongside the internal output stage.  (Use unbuffered inverters for this circuit.)  I remember that this trick was used in General Radio DigiBridges for the current-sense amplifier.

There isn't any frequency compensation, so I'd be surprised if it's stable. The high gain of 255 will help somewhat, but be prepared for fun and games with oscillation.

To clarify, I was referring to the op amp configuration, not the entire high-gain amplifier circuit.  The CA3130 is not inherently unity-gain stable, and requires a compensation capacitor from pin 1 to pin 8 (the input to the CMOS inverter output stage) at unity gain.  The output stage is a 4000-series inverter, which is a transconductance near the middle of the output voltage range, crossing over to an ON resistance near either supply pin.  The circuit in Fig 17 adds three additional inverters (from a CMOS array) in parallel with the two output FETs (pin 8 to output at pin 6) to increase the transconductance and output current capability (not after the op amp, as often done with buffers and conventional op amps).  The actual circuit in Fig 17 is for a higher gain inverting amplifier.  I believe that an unbuffered 4009 CMOS hex inverter could be used, but I haven't tried it.  The output stage is interesting, since the voltage gain depends on the load resistance (high-Z output in "active" region).

You're right. I missed that the CMOS inverter is connected in parallel with the CA3130's internal CMOS inverter. Yes, the gain varying with the output voltage/current is an interesting feature of the CA130, which makes frequency compensation challanging.

[Terry Bites]

MAX44267? Dual op amp with built in charge pump. $5