Op-Amp Help Please

Take a look at the LM6132.

"The LM6132/34 can be driven by voltages that exceed both power supply rails, thus eliminating concerns over exceeding the common-mode voltage range. The rail-to-rail output swing capability provides the maximum possible dynamic range at the output. This is particularly important when op- erating on low supply voltages. The LM6132/34 can also drive large capacitive loads without oscillating."

Looks like the footprint is okay. One downside is that its drive capacity is pretty wimpy but I'd guess would be a good fit into a high-impedance voltmeter input.

TL082, although the datasheet values do not appear to guarantee it (what the big print giveth, the small print taketh away).

Crap -- never mind. It's a nice part but only rated for a 5 V unipolar supply.

Look again.. it's not that bad a suggestion, price aside.

Quote: "3. The magnitude of the input voltage must never exceed the magnitude of the supply voltage or 15 V, whichever is less."

But Jim, SOIC ain't microsoldering. Have them try their hand on a nice fat TSSOP, DFN or SC-75 :-)

something like the TL082? I'm not sure it is garanteed to go to V+ but I think it supposed to.

-Lasse

That's *differential* input voltage, n'est ce pas? Otherwise the input common mode specs don't make any sense on a 30V (+/-15V) supply.

Most op-amps have a limited differential input voltage, so you have to limit the voltage or sometimes just the current. In Jim's case, a sudden short-circuit could put nominal 17V in series with 5.4K across the inputs unless something limits the current.

I think TI 2372 has rail to rail inputs, but haven't checked footprint, probably matches.

This is a typical problem which is called "over the top current sensor" and there is 1001 variant of schematics of how to do that right. It is described in AOE also. BTW, if done right, there is no need for R-R opamp.

Congratulations.

Standard footprint for dual opamp.

There are not too many of R-R dual opamps which can operate from the

SOIC is macrosoldering, not microsoldering.

Vladimir Vassilevsky DSP and Mixed Signal Design Consultant

Could you use a high side current sense amplifier? That's basically an differential op amp but with the resistors built in.

The MAX4377 is a dual part available in 8 pin SOIC package with the same pin out as the 358. You would have to change some of your resistors values (mostly to zero or infinity), but I think you could do it with the same PCB layout.

See:

Jim,

At the expense of overall gain, if you merely drop the value of the "to gnd" and "out-to-input" resistors from 100K to, let's say, 10K, your circuit will work.

With those Rs equal to 10K, and no voltage across your 0.25ohm sense, both inputs to the opamp will be about 13.4V. When there's 0.25V across your sense resistor, then the output from the opamp will be at abut 0.93V. Of course, the plus input won't go any higher than 13.4V, and the minus input will always be at or below 13.4V.

With R = 20K, then the plus input will always be at 14.9V, and then the output will be 1.8V with 0.25V across the sense. So, this looks like the best you can do and stay within the opamp's input spec.

You can also increase the value of that sense resistor to recover some of the lost gain, but that will be at the expense of that resistor's power dissipation. You'll also have to watch that you don't violate the regulator's minimum Vin when the load current is at its maximum.

Bob

Linear technology have a few rail to rail input op amps that may do what you want.

Channels Packages Vs Max Rail-to-Rail In # V yes/no LT1366 2 SO-8 36 yes LT1368 2 SO-8 36 yes LT1490A 2 SO-8 44 yes LT1495 2 SO-8 36 yes LT1498 2 SO-8 36 yes LT1630 2 SO-8 36 yes LT1632 2 SO-8 36 yes LT1638 2 SO-8 44 yes LT1673 2 SO-8 36 yes LT1678 2 SO-8 36 yes

National have a few too:

LM7332 LM6152 LM6132

But the schematic (pg. 5) says it works. Which do you trust?

Cheers, James Arthur

Oopsie. Too true. That's what I get for posting with one hand while I'm grabbing the car keys with the other.

The spec. Always. But only the values in the tables that are guaranteed design limits. Kind of comes with my line of work ...

and

the

to

I'd

what

This datasheet says diff can be +/-30V:

It'll be mostly diff I assume. BUT: The table says Vicm is 11V worst case. Bummer :-(

The LM6132 is out, too, since diff abs max is 15V.

Sonny, fourteen weeks ago, the smallest soldering most of these kids had done was to patch a hole in an auto body with solder. To get them to do through hole was a major victory; doing SOIC was a triumph.

Jim

Now that's not going to drive a 1-15 volt meter (internal multiplier resistor) too well, is it?

Jim

You didn't read the entire solution. Do the math and/or some simulation and you'll see that you get very close to what you want with simple resistor changes.

If you really want help from people you'd be better served by losing that huffy attitude.

Bob