lm324 op amp

[timmmmayyy!]

Has anyone found this problem,

When a negative voltage is connected to the input of this op amp the output
will go positive, and the other outputs will go slightly positive.

Do you know what causes this?

Thanks

 

[Tim Wescott]

Has anyone found this problem,

When a negative voltage is connected to the input of this op amp the output
will go positive, and the other outputs will go slightly positive.

Do you know what causes this?

Thanks

How negative? Are you going below the negative rail?

http://www.analog.com/library/analogDialogue/Anniversary/6.html

Search for "phase inversion".

All the parts are on one die, so if you do something that lets the
electrons escape it'll affect all the other units on the die (this can
give you fits on multiple-input ADC's; somehow the datasheet doesn't
stress what happens to the other channels when you "safely" turn on the
protection diodes).

 

[John Larkin]

Has anyone found this problem,

When a negative voltage is connected to the input of this op amp the output
will go positive, and the other outputs will go slightly positive.

Do you know what causes this?

Thanks

If you go below the negative supply, even a few tenths of a volt, a
324 (or similar-process parts, like LM339) will go crazy internally;
charge is being sprayed everywhere. There's a tiny fine-print footnote
on the original National datasheet, and some hints in the application
examples near the end.

Also: If one section of a 324 swings to the output rails, it can/will
also mess up the other amps in the same can. It's a nasty part,
overall.

LM35 and other National parts have similar problems. Pease promised me
he'd fix the LM35, maybe 10 years ago, but he hasn't so far.

John

 

[Spehro Pefhany]

Has anyone found this problem,

When a negative voltage is connected to the input of this op amp the output
will go positive, and the other outputs will go slightly positive.

Do you know what causes this?

Thanks

It's called "output phase reversal" and it's a well-known "feature" of
the LM324 and other op-amps. You will only see it if you violate the
datasheet Absolute Maximum rating of -300mV (@ 25°C) input voltage.

It's caused by current flowing from the input (forward biasing
isolation junctions) into the substrate, which b*ggers things up in
one way or another.


Best regards,
Spehro Pefhany

 

[John Fields]

Ah, give Pease a chance.

 

[Spehro Pefhany]

If you go below the negative supply, even a few tenths of a volt, a
324 (or similar-process parts, like LM339) will go crazy internally;
charge is being sprayed everywhere. There's a tiny fine-print footnote
on the original National datasheet, and some hints in the application
examples near the end.

Also: If one section of a 324 swings to the output rails, it can/will
also mess up the other amps in the same can. It's a nasty part,
overall.

LM35 and other National parts have similar problems. Pease promised me
he'd fix the LM35, maybe 10 years ago, but he hasn't so far.

John

Ah, give Pease a chance.


Best regards,
Spehro Pefhany

 

[John Larkin]

Ah, give Pease a chance.

20 points!

John

 

[Clifford Heath]

...a 324 will go crazy internally...It's a nasty part,

What ubiquitously available quad would you recommend instead?

 

[Ken Smith]

What ubiquitously available quad would you recommend instead?

If the design is finished, put a LT1014 in place of the LM324.

 

[John Larkin]

What ubiquitously available quad would you recommend instead?

Why a quad? Duals are a lot more flexible, and don't take much more
room. Decent old/cheap amps are LF353 (dual jfet) and LM1458, basicly
a dirt-cheap dual 741. But there are hundreds of nice opamps around
nowadays: cmos, rail-to-rail, micropower, precision, low-noise,
superfast, whatever.

I use a lot of LM7301's, a SOT-23 bipolar r-r amp. It's a single, so
you just plop one down wherever you need it.

John

 

[Ken Smith]

Why a quad? Duals are a lot more flexible,

Have you ever tried bending one?

LF353 (dual jfet) and LM1458,

I'll second those suggestions.

[...]

I use a lot of LM7301's, a SOT-23 bipolar r-r amp. It's a single, so
you just plop one down wherever you need it.

You can get r-r opamps and PICs in the SOT-23 package. This could make
designs even harder to copy.

 

[Spehro Pefhany]

What ubiquitously available quad would you recommend instead?

And nothing over a dime or two, please. ;-)


Best regards,
Spehro Pefhany

 

[Fred Bloggs]

If the design is finished, put a LT1014 in place of the LM324.

Make that the el cheapo Texas Instruments LT1013 family- and you are
only about 2x the National price on the LM324.

 

[John Larkin]

Have you ever tried bending one?

No, but I have stepped on both pin-up DIP8 and DIP16 packages
barefoot, and the 8-pins are much to be preferred.

John

 

[Spehro Pefhany]

Make that the el cheapo Texas Instruments LT1013 family- and you are
only about 2x the National price on the LM324.

For half the number of amplfiers, but it's a much nicer part (even the
super-relaxed-spec D version). IIRC, the dual version of the LM324
(the 358) is about the same price as the quad, maybe even more.


Best regards,
Spehro Pefhany

 

[Tim Shoppa]

How negative? Are you going below the negative rail?

http://www.analog.com/library/analogDialogue/Anniversary/6.html

Ha! That was an enjoyable little read. I especially liked

The figure was drawn with thermionic tubes to avoid any suggestion of
partiality in favour of any particular semiconductor device. Since
thermionic devices at present are not generally available in IC chip form

I'd never seen that section of their website, I gotta read the rest
now

Tim.

 

[Joerg]

Hi Fred,

Make that the el cheapo Texas Instruments LT1013 family- and you are
only about 2x the National price on the LM324.

Or just stay with the LM324 and clamp off the negative peaks using a
reverse transistor, pre-biased diode etc. This would only add a few
cents to the design.

Regards, Joerg

 

[Winfield Hill]

Joerg wrote...

Hi Fred,


Or just stay with the LM324 and clamp off the negative peaks
using a reverse transistor, pre-biased diode etc. This would
only add a few cents to the design.

If a simple diode to ground, it had better be a Schottky diode.

 

[Joerg]

Hi Winfield,

If a simple diode to ground, it had better be a Schottky diode.

A regular one would have to be pre-biased so that it starts to clamp at
less than 300mV. Schottkys are kind of pricey. I'd probably consider the
inverse transistor if the voltage swing in the other direction is always
below the breakdown.

Regards, Joerg

 

[Winfield Hill]

Joerg wrote...

Hi Winfield,


... Schottkys are kind of pricey.

Not that bad... For example, Vishay BAT82S (44348 in stock),
SD101A (16922 in stock) are both $0.03, qty 1k at Newark, $0.05
in singles, and the LS101C smd part a bit more, $0.04 for 1k...