197Views10Replies

Author Options:

Op amps? Which one is best? Answered

I have a couple op amps collected over a couple years, and I am working on a design for a linear power supply, but have stability issues, and a "down to -V" common mode input voltage for op amps is a difficult parameter to find.

Constraints:

* Inputs must work down to as close to the negative supply rail as possible 
* needs to be stable when used in a feedback loop for a linear power supply design
* low noise / offset / blah blah good good goodies?? Hopefully?

I'm actively testing these in a basic comparator circuit to see how far I can get the inputs to the -vee (or ground) before the op amp stops working as intended.I have tested the LM324, LM386, UA1458TC, LM741, and a few of the LT one's. Only the first 2 work well on a single supply w/ the inputs near ground or Vee

LM324; // slow
LM224; // ditto, but in weird thicc package
LM386; // slow
LM741; // ancient relics of the past, not rail to rail by any means, sucks IMO.
UA1458TC // ancient relics of the past, not rail to rail by any means, sucks IMO.
LM318H; // it's in a fancy metal case :D
TL072CP; // JFET inputs, limited output current and the output can't go below the noninverting input.
TL2072AC; // ditto.
OP37; // These things aren't even that fast and can't work at unity gain.
OP27; // ditto
LT1007;
LT1167;
LT1191;
LT1363; CRAZY fast, 1000V/uS!!! Maybe too fast...
LT1360;
LT1112;

Discussions

0
None
-max-iceng

Answer 2 years ago

CA3130 is not capable of 20V on the inputs. :/

0
None
-max--max-

Answer 2 years ago

20V supply voltage.**

0
None
steveastrouk

2 years ago

Just look for 'em. There are some nice TI parts, since they bought Burr-Brown, there are Analog Devices parts.....

Even the old CA3130 is worth a look.

0
None
-max-steveastrouk

Answer 2 years ago

CA3130 is 16V maximum. Ideally I need one that can handle 25V between Vcc and Vee. Otherwise my design would need to be more complicated. Although I like that it literally mentions use in linear regulator designed made to work down to 0V! That's definitly a good start.

0
None
-max-

2 years ago

I went back to the complementary pair design (as the darlington had no benefit) and it turned out that adding at least 4.7uF of output capacitance made the supply very stable, both in CV and CC mode. :D Given that it was breadboarded I wonder if it 1uF would be enough on a proper PCB...

-

But does anyone not know of any good rail to rail op amps? The LM358 is the nicest one I have, but the output goes down to only 0.6V, not good enough to achieve a 0V output.

0
None
Josehf Murchison-max-

Answer 2 years ago

I was looking at your circuit.

Do you get 0 volts in your circuit simulator program but when you build it on a breadboard you get 0.6 volts.

Because I can see that.

Although your 2222 transistors shunt to ground they have a voltage across collector to emitter of about 0.6 volts even when saturated.

0
None
-max-Josehf Murchison

Answer 2 years ago

The darlington this now outdated schematic drops between 2 and 5V. The LM358 has a minimum output of 0.6V, + 0.2V of the schottky diode (I used a 1N5817) so if I had a minimum load current capable of going to 0V, then I could achieve 0V output.

However I prefer the lower drop of a complementary pair, which is not capable of going to 0V due to the LM358's minimum output when sinking 5mA. This configuration makes the new minimum output voltage about 800mV. If I cannot get ahold of a good rail to rail op amp, then I will add a small auxiliary negative voltage rail of -5V or so. But I feel that if I can get the base of the complementy pair down to at least 0.5v then I don't need that.

0
None
Josehf Murchison-max-

Answer 2 years ago

Well an LM339 typical VOL is 150 mV I am assuming you are using a Quad Op Amp.

0
None
-max-

2 years ago

Also for some reason adding current regulation causes the output to oscillate like mad. Not exactly sure why yet. Might be a mistake in my wiring. A rail to rail op amp would be nice. Anyone know of a good one?

linear power supply design.png